Azure MS Docs PDF

Contents
Windows VMs Documentation

Overview
About Virtual Machines
Quickstarts
Create VM - Portal
Create VM - PowerShell
Create VM - Azure CLI
Tutorials
1 - Create / manage a VM
2 - Create / manage disks
3 - Automate configuration
4 - Create VM images
5 - Highly available VMs
6 - Create a VM scale set
7 - Load balance VMs
8 - Manage networking
9 - Backup virtual machines
10 - Govern VMs
11 - Monitor and update VMs
12 - Manage VM security
13 - Azure Pipelines CI/CD
14 - Install a SQL\IIS\.NET stack
15 - Secure web server with SSL
Samples
PowerShell
Azure CLI
Concepts
Azure Resource Manager
Regions and availability
VM types and sizes
General purpose
B-series burstable
Compute optimized
Memory optimized
Constrained vCPUs
Storage optimized
GPU optimized
Setup GPU drivers
High performance compute
Previous generations
Azure compute units (ACU)
Benchmark scores
Maintenance and updates
Disk storage
Managed Disks
Premium SSDs
Standard SSDs
Ultra SSDs
Premium storage performance
Standard storage
Scalability targets for disks
Backup and disaster recovery for disks
Networking
Auto-scale applications
Infrastructure automation
Security and policy
States and lifecycle
Monitoring
Backup and recovery
High performance computing
Deployment considerations
Infrastructure guidelines
Infrastructure guidelines
vCPU quotas
How-to guides
Create VMs
Use C#
Use specialized disk
Portal
PowerShell
Use a template with C#
Create VM with Chef
Use Java
Use Python
Use a template
Connect to a VM
Use Azure Hybrid Benefit license
Use Multitenant Hosting Rights for Windows 10
Secure VMs
Encrypt
Use WinRM
Use access controls
Use policies
Create a Key Vault
Protect VMs
Back up a single VM
Azure portal
Azure PowerShell
Back up multiple VMs
Manage VMs
VM usage
Common PowerShell tasks
Move a VM
Change VM size
Swap the OS disk
Tag a VM
Time sync
Run scripts on a VM
Custom Script Extension
Run Command
Change temp drive letter
Change availability set
Download template
Azure VM agent
Planned maintenance
Mitigating speculative execution
Scheduled events
Monitor metadata
Enable nested virtualization
Manage costs
Prepay for VMs - Azure reservations
What are Azure reservations?
VM instance size flexibility
Use Images
Shared images (preview)
Overview
Create shared images using CLI
Troubleshoot shared images
Find and use images
Prepare VM for upload
Capture VM to image
Use generalized image
Build image with Packer
Use Windows client images
Download existing disk
Availability and scale
Virtual Machine Scale Sets
High availability
Vertically scale
Create VM in availability zone
PowerShell
Portal
Use automation tools
Chef
Publish Web App from Visual Studio
Run applications
SQL Server
MongoDB
SAP on Azure
MATLAB cluster
Visual Studio
High Performance Computing (HPC)
HPC Pack 2016 cluster
HPC Pack 2016 Azure Active Directory integration
HPC Pack 2012 R2 head node
Submit on-prem jobs to HPC Pack 2012 R2
Excel on HPC Pack
Manage storage
Disks FAQs
Add a disk
Azure Powershell
Azure portal
Detach a disk
Resize a disk
Snapshot a disk
Back up unmanaged disks
Convert disk between Standard and Premium
Migrate to Premium storage with Azure Site Recovery
Find unattached disks
Use File storage
Deploy disks with template
Enable write accelerator
Using Ultra SSD
Migrate to Managed Disks
Unmanaged VM to Managed Disks
Single Classic VM to Resource Manager
Manage networking
Create virtual network
Open ports to a VM
Azure portal
Azure PowerShell
Assign public IP address
Use multiple NICs
Use accelerated networking
Assign public DNS name
DNS resolution
Configure managed identities
Portal
CLI
PowerShell
Azure Resource Manager Template
REST
Azure SDKs
Use VM extensions
Migrate VMs
Migrate AWS and on-premises VMs
Upload on-prem VM
From Amazon Web Services (AWS)
Use Azure Site Recovery
Migrate from Classic to Azure Resource Manager
Deep Dive on migration
Plan for migration
Migrate using PowerShell
Common migration errors
Community tools for migration
FAQ
Troubleshoot
Reference
Azure CLI
Azure PowerShell
.NET
Java
Node.js
Python
Compute REST
Managed Disks REST
Resources
Author templates
Build your skills with Microsoft Learn
Azure Roadmap
Community templates
Pricing
Regional availability
Stack Overflow
Videos
FAQ
Overview of Windows virtual machines in Azure
10/5/2018 • 7 minutes to read • Edit Online
Azure Virtual Machines (VM ) is one of several types of on-demand, scalable computing resources that Azure
offers. Typically, you choose a VM when you need more control over the computing environment than the other
choices offer. This article gives you information about what you should consider before you create a VM, how you
create it, and how you manage it.
An Azure VM gives you the flexibility of virtualization without having to buy and maintain the physical hardware
that runs it. However, you still need to maintain the VM by performing tasks, such as configuring, patching, and
installing the software that runs on it.
Azure virtual machines can be used in various ways. Some examples are:
Development and test – Azure VMs offer a quick and easy way to create a computer with specific
configurations required to code and test an application.
Applications in the cloud – Because demand for your application can fluctuate, it might make economic
sense to run it on a VM in Azure. You pay for extra VMs when you need them and shut them down when you
don’t.
Extended datacenter – Virtual machines in an Azure virtual network can easily be connected to your
organization’s network.
The number of VMs that your application uses can scale up and out to whatever is required to meet your needs.
What do I need to think about before creating a VM?

There are always a multitude of design considerations when you build out an application infrastructure in Azure.
These aspects of a VM are important to think about before you start:
The names of your application resources
The location where the resources are stored
The size of the VM
The maximum number of VMs that can be created
The operating system that the VM runs
The configuration of the VM after it starts
The related resources that the VM needs
Naming
A virtual machine has a name assigned to it and it has a computer name configured as part of the operating
system. The name of a VM can be up to 15 characters.
If you use Azure to create the operating system disk, the computer name and the virtual machine name are the
same. If you upload and use your own image that contains a previously configured operating system and use it to
create a virtual machine, the names can be different. We recommend that when you upload your own image file,
you make the computer name in the operating system and the virtual machine name the same.
Locations
All resources created in Azure are distributed across multiple geographical regions around the world. Usually, the
region is called location when you create a VM. For a VM, the location specifies where the virtual hard disks are
stored.
This table shows some of the ways you can get a list of available locations.
METHOD DESCRIPTION
Azure portal Select a location from the list when you create a VM.
Azure PowerShell Use the Get-AzureRmLocation command.
REST API Use the List locations operation.
Azure CLI Use the az account list-locations operation.
VM size
The size of the VM that you use is determined by the workload that you want to run. The size that you choose then
determines factors such as processing power, memory, and storage capacity. Azure offers a wide variety of sizes to
support many types of uses.
Azure charges an hourly price based on the VM’s size and operating system. For partial hours, Azure charges only
for the minutes used. Storage is priced and charged separately.
VM Limits
Your subscription has default quota limits in place that could impact the deployment of many VMs for your
project. The current limit on a per subscription basis is 20 VMs per region. Limits can be raised by filing a support
ticket requesting an increase
Operating system disks and images
Virtual machines use virtual hard disks (VHDs) to store their operating system (OS ) and data. VHDs are also used
for the images you can choose from to install an OS.
Azure provides many marketplace images to use with various versions and types of Windows Server operating
systems. Marketplace images are identified by image publisher, offer, sku, and version (typically version is specified
as latest). Only 64-bit operating systems are supported. For more information on the supported guest operating
systems, roles, and features, see Microsoft server software support for Microsoft Azure virtual machines .
This table shows some ways that you can find the information for an image.
METHOD DESCRIPTION
Azure portal The values are automatically specified for you when you select
an image to use.
Azure PowerShell Get-AzureRMVMImagePublisher -Location location

Get-AzureRMVMImageOffer -Location location -Publisher
publisherName
Get-AzureRMVMImageSku -Location location -Publisher
publisherName -Offer offerName
REST APIs List image publishers

List image offers
List image skus
Azure CLI az vm image list-publishers --location location

az vm image list-offers --location location --publisher
publisherName
az vm image list-skus --location location --publisher
publisherName --offer offerName
You can choose to upload and use your own image and when you do, the publisher name, offer, and sku aren’t
used.
Extensions
VM extensions give your VM additional capabilities through post deployment configuration and automated tasks.
These common tasks can be accomplished using extensions:
Run custom scripts – The Custom Script Extension helps you configure workloads on the VM by running
your script when the VM is provisioned.
Deploy and manage configurations – The PowerShell Desired State Configuration (DSC ) Extension helps
you set up DSC on a VM to manage configurations and environments.
Collect diagnostics data – The Azure Diagnostics Extension helps you configure the VM to collect
diagnostics data that can be used to monitor the health of your application.
Related resources
The resources in this table are used by the VM and need to exist or be created when the VM is created.
RESOURCE REQUIRED DESCRIPTION
Resource group Yes The VM must be contained in a

resource group.
Storage account Yes The VM needs the storage account to

store its virtual hard disks.
Virtual network Yes The VM must be a member of a virtual

network.
Public IP address No The VM can have a public IP address

assigned to it to remotely access it.
Network interface Yes The VM needs the network interface to

communicate in the network.
Data disks No The VM can include data disks to

expand storage capabilities.
How do I create my first VM?

You have several choices for creating your VM. The choice that you make depends on the environment you are in.
This table provides information to get you started creating your VM.
METHOD ARTICLE
Azure portal Create a virtual machine running Windows using the portal
Templates Create a Windows virtual machine with a Resource Manager

template
Azure PowerShell Create a Windows VM using PowerShell
Client SDKs Deploy Azure Resources using C#

METHOD ARTICLE
REST APIs Create or update a VM
Azure CLI Create a VM with the Azure CLI
You hope it never happens, but occasionally something goes wrong. If this situation happens to you, look at the
information in Troubleshoot Resource Manager deployment issues with creating a Windows virtual machine in
Azure.
How do I manage the VM that I created?

VMs can be managed using a browser-based portal, command-line tools with support for scripting, or directly
through APIs. Some typical management tasks that you might perform are getting information about a VM,
logging on to a VM, managing availability, and making backups.
Get information about a VM
This table shows you some of the ways that you can get information about a VM.
METHOD DESCRIPTION
Azure portal On the hub menu, click Virtual Machines and then select the
VM from the list. On the blade for the VM, you have access to
overview information, setting values, and monitoring metrics.
Azure PowerShell For information about using PowerShell to manage VMs, see
Create and manage Windows VMs with the Azure PowerShell
module.
REST API Use the Get VM information operation to get information

about a VM.
Client SDKs For information about using C# to manage VMs, see Manage
Azure Virtual Machines using Azure Resource Manager and
C#.
Azure CLI For information about using Azure CLI to manage VMs, see
Azure CLI Reference.
Log on to the VM
You use the Connect button in the Azure portal to start a Remote Desktop (RDP ) session. Things can sometimes
go wrong when trying to use a remote connection. If this situation happens to you, check out the help information
in Troubleshoot Remote Desktop connections to an Azure virtual machine running Windows.
Manage availability
It’s important for you to understand how to ensure high availability for your application. This configuration
involves creating multiple VMs to ensure that at least one is running.
In order for your deployment to qualify for our 99.95 VM Service Level Agreement, you need to deploy two or
more VMs running your workload inside an availability set. This configuration ensures your VMs are distributed
across multiple fault domains and are deployed onto hosts with different maintenance windows. The full Azure
SLA explains the guaranteed availability of Azure as a whole.
Back up the VM
A Recovery Services vault is used to protect data and assets in both Azure Backup and Azure Site Recovery
services. You can use a Recovery Services vault to deploy and manage backups for Resource Manager-deployed
VMs using PowerShell.
Next steps
If your intent is to work with Linux VMs, look at Azure and Linux.
Learn more about the guidelines around setting up your infrastructure in the Example Azure infrastructure
walkthrough.
Quickstart: Create a Windows virtual machine in the
Azure portal
Azure virtual machines (VMs) can be created through the Azure portal. This method provides a browser-based
user interface to create VMs and their associated resources. This quickstart shows you how to use the Azure
portal to deploy a virtual machine (VM ) in Azure that runs Windows Server 2016. To see your VM in action, you
then RDP to the VM and install the IIS web server.
If you don't have an Azure subscription, create a free account before you begin.
Sign in to Azure
Sign in to the Azure portal at https://portal.azure.com.
Create virtual machine

1. Choose Create a resource in the upper left-hand corner of the Azure portal.
2. In the search box above the list of Azure Marketplace resources, search for and select Windows Server
2016 Datacenter, then choose Create.
3. In the Basics tab, under Project details, make sure the correct subscription is selected and then choose to
Create new resource group. Type the myResourceGroup for the name.
4. Under Instance details, type myVM for the Virtual machine name and choose East US for your
Location. Leave the other defaults.
5. Under Administrator account, provide a username, such as azureuser and a password. The password
must be at least 12 characters long and meet the defined complexity requirements.
6. Under Inbound port rules, choose Allow selected ports and then select RDP (3389) and HTTP from
the drop-down.
7. Leave the remaining defaults and then select the Review + create button at the bottom of the page.
Connect to virtual machine

Create a remote desktop connection to the virtual machine. These directions tell you how to connect to your VM
from a Windows computer. On a Mac, you need an RDP client such as this Remote Desktop Client from the Mac
App Store.
1. Click the Connect button on the virtual machine properties page.
2. In the Connect to virtual machine page, keep the default options to connect by DNS name over port
3389 and click Download RDP file.
3. Open the downloaded RDP file and click Connect when prompted.
4. In the Windows Security window, select More choices and then Use a different account. Type the
username as localhost\username, enter password you created for the virtual machine, and then click OK.
5. You may receive a certificate warning during the sign-in process. Click Yes or Continue to create the
connection.
Install web server

To see your VM in action, install the IIS web server. Open a PowerShell prompt on the VM and run the following
command:
Install-WindowsFeature -name Web-Server -IncludeManagementTools
When done, close the RDP connection to the VM.
View the IIS welcome page

In the portal, select the VM and in the overview of the VM, use the Click to copy button to the right of the IP
address to copy it and paste it into a browser tab. The default IIS welcome page will open, and should look like
this:
Clean up resources
When no longer needed, you can delete the resource group, virtual machine, and all related resources. To do so,
select the resource group for the virtual machine, select Delete, then confirm the name of the resource group to
delete.
Next steps
In this quickstart, you deployed a simple virtual machine, open a network port for web traffic, and installed a basic
web server. To learn more about Azure virtual machines, continue to the tutorial for Windows VMs.
Azure Windows virtual machine tutorials
Quickstart: Create a Windows virtual machine in
Azure with PowerShell
The Azure PowerShell module is used to create and manage Azure resources from the PowerShell command line
or in scripts. This quickstart shows you how to use the Azure PowerShell module to deploy a virtual machine (VM )
in Azure that runs Windows Server 2016. To see your VM in action, you then RDP to the VM and install the IIS
web server.
Launch Azure Cloud Shell

The Azure Cloud Shell is a free interactive shell that you can use to run the steps in this article. It has common
Azure tools preinstalled and configured to use with your account.
To open the Cloud Shell, just select Try it from the upper right corner of a code block. You can also launch Cloud
Shell in a separate browser tab by going to https://shell.azure.com/powershell. Select Copy to copy the blocks of
code, paste it into the Cloud Shell, and press enter to run it.
If you choose to install and use the PowerShell locally, this tutorial requires the Azure PowerShell module version
5.7.0 or later. Run Get-Module -ListAvailable AzureRM to find the version. If you need to upgrade, see Install Azure
PowerShell module. If you are running PowerShell locally, you also need to run Connect-AzureRmAccount to create a
connection with Azure.
Create resource group

Create an Azure resource group with New -AzureRmResourceGroup. A resource group is a logical container into
which Azure resources are deployed and managed.
New-AzureRmResourceGroup -Name myResourceGroup -Location EastUS

Create a VM with New -AzureRmVM. Provide names for each of the resources and the New -AzureRmVM cmdlet
creates if they don't already exist.
When prompted, provide a username and password to be used as the logon credentials for the VM:
New-AzureRmVm `
-ResourceGroupName "myResourceGroup" `
-Name "myVM" `
-Location "East US" `
-VirtualNetworkName "myVnet" `
-SubnetName "mySubnet" `
-SecurityGroupName "myNetworkSecurityGroup" `
-PublicIpAddressName "myPublicIpAddress" `
-OpenPorts 80,3389
After the deployment has completed, RDP to the VM. To see your VM in action, the IIS web server is then
installed.
To see the public IP address of the VM, use the Get-AzureRmPublicIpAddress cmdlet:
Get-AzureRmPublicIpAddress -ResourceGroupName "myResourceGroup" | Select "IpAddress"
Use the following command to create a remote desktop session from your local computer. Replace the IP address
with the public IP address of your VM.
mstsc /v:publicIpAddress
In the Windows Security window, select More choices, and then select Use a different account. Type the
username as localhost\username, enter password you created for the virtual machine, and then click OK.
You may receive a certificate warning during the sign-in process. Click Yes or Continue to create the connection
Install web server

command:
View the web server in action

With IIS installed and port 80 now open on your VM from the Internet, use a web browser of your choice to view
the default IIS welcome page. Use the public IP address of your VM obtained in a previous step. The following
example shows the default IIS web site:
Clean up resources
When no longer needed, you can use the Remove-AzureRmResourceGroup cmdlet to remove the resource group,
VM, and all related resources:
Remove-AzureRmResourceGroup -Name myResourceGroup
Next steps
Quickstart: Create a Windows virtual machine with
the Azure CLI
The Azure CLI is used to create and manage Azure resources from the command line or in scripts. This quickstart
shows you how to use the Azure CLI to deploy a virtual machine (VM ) in Azure that runs Windows Server 2016.
To see your VM in action, you then RDP to the VM and install the IIS web server.
Open Azure Cloud Shell

Azure Cloud Shell is a free, interactive shell that you can use to run the steps in this article. Common Azure tools
are preinstalled and configured in Cloud Shell for you to use with your account. Just select the Copy button to
copy the code, paste it in Cloud Shell, and then press Enter to run it. There are a few ways to open Cloud Shell:
Select Try It in the upper-right corner of a code block.
Open Cloud Shell in your browser.
Select the Cloud Shell button on the menu in the upper-right

corner of the Azure portal.
If you choose to install and use the CLI locally, this quickstart requires that you are running the Azure CLI version
2.0.30 or later. Run az --version to find the version. If you need to install or upgrade, see Install Azure CLI.
Create a resource group

Create a resource group with the az group create command. An Azure resource group is a logical container into
which Azure resources are deployed and managed. The following example creates a resource group named
myResourceGroup in the eastus location:
az group create --name myResourceGroup --location eastus

Create a VM with az vm create. The following example creates a VM named myVM. This example uses azureuser
for an administrative user name and myPassword12 as the password. Update these values to something
appropriate to your environment. These values are needed when you connect to the VM.
az vm create \
--resource-group myResourceGroup \
--name myVM \
--image win2016datacenter \
--admin-username azureuser \
--admin-password myPassword12
It takes a few minutes to create the VM and supporting resources. The following example output shows the VM
create operation was successful.
{
"fqdns": "",
"id":
"/subscriptions/<guid>/resourceGroups/myResourceGroup/providers/Microsoft.Compute/virtualMachines/myVM",
"location": "eastus",
"macAddress": "00-0D-3A-23-9A-49",
"powerState": "VM running",
"privateIpAddress": "10.0.0.4",
"publicIpAddress": "52.174.34.95",
"resourceGroup": "myResourceGroup"
}
Note your own publicIpAddress in the output from your VM. This address is used to access the VM in the next
steps.
Open port 80 for web traffic

By default, only RDP connections are opened when you create a Windows VM in Azure. Use az vm open-port to
open TCP port 80 for use with the IIS web server:
az vm open-port --port 80 --resource-group myResourceGroup --name myVM

Use the following command to create a remote desktop session from your local computer. Replace the IP address
with the public IP address of your VM. When prompted, enter the credentials used when the VM was created:
mstsc /v:publicIpAddress
Install web server

command:
View the web server in action

With IIS installed and port 80 now open on your VM from the Internet, use a web browser of your choice to view
the default IIS welcome page. Use the public IP address of your VM obtained in a previous step. The following
example shows the default IIS web site:
Clean up resources
When no longer needed, you can use the az group delete command to remove the resource group, VM, and all
related resources:
az group delete --name myResourceGroup
Next steps
Tutorial: Create and Manage Windows VMs with
Azure PowerShell
Azure virtual machines provide a fully configurable and flexible computing environment. This tutorial covers basic
Azure virtual machine deployment items such as selecting a VM size, selecting a VM image, and deploying a VM.
You learn how to:
Create and connect to a VM
Select and use VM images
View and use specific VM sizes
Resize a VM
View and understand VM state

Azure tools preinstalled and configured to use with your account. Just click the Copy to copy the code, paste it
into the Cloud Shell, and then press enter to run it. There are a few ways to launch the Cloud Shell:
Click Try It in the upper right corner of a code block.
Click the Cloud Shell button on the menu in the upper right
of the Azure portal.
PowerShell module. If you are running PowerShell locally, you also need to run Connect-AzureRmAccount to create
a connection with Azure.

Create a resource group with the New -AzureRmResourceGroup command.
An Azure resource group is a logical container into which Azure resources are deployed and managed. A resource
group must be created before a virtual machine. In the following example, a resource group named
myResourceGroupVM is created in the EastUS region:
New-AzureRmResourceGroup -ResourceGroupName "myResourceGroupVM" -Location "EastUS"
The resource group is specified when creating or modifying a VM, which can be seen throughout this tutorial.
When creating a virtual machine, several options are available such as operating system image, network
configuration, and administrative credentials. In this example, a virtual machine is created with a name of myVM
running the default latest version of Windows Server 2016 Datacenter.
Set the username and password needed for the administrator account on the virtual machine with Get-Credential:
$cred = Get-Credential
Create the virtual machine with New -AzureRmVM.
New-AzureRmVm `
-ResourceGroupName "myResourceGroupVM" `
-Name "myVM" `
-Location "EastUS" `
-Credential $cred
Connect to VM
After the deployment has completed, create a remote desktop connection with the virtual machine.
Run the following commands to return the public IP address of the virtual machine. Take note of this IP Address
so you can connect to it with your browser to test web connectivity in a future step.
Get-AzureRmPublicIpAddress -ResourceGroupName "myResourceGroupVM" | Select IpAddress
Use the following command, on your local machine, to create a remote desktop session with the virtual machine.
Replace the IP address with the publicIPAddress of your virtual machine. When prompted, enter the credentials
used when creating the virtual machine.
mstsc /v:<publicIpAddress>
In the Windows Security window, select More choices and then Use a different account. Type the username
and password you created for the virtual machine and then click OK.
Understand VM images
The Azure marketplace includes many virtual machine images that can be used to create a new virtual machine.
In the previous steps, a virtual machine was created using the Windows Server 2016 Datacenter image. In this
step, the PowerShell module is used to search the marketplace for other Windows images, which can also be used
as a base for new VMs. This process consists of finding the publisher, offer, SKU, and optionally a version number
to identify the image.
Use the Get-AzureRmVMImagePublisher command to return a list of image publishers:
Get-AzureRmVMImagePublisher -Location "EastUS"
Use the Get-AzureRmVMImageOffer to return a list of image offers. With this command, the returned list is
filtered on the specified publisher:
Get-AzureRmVMImageOffer -Location "EastUS" -PublisherName "MicrosoftWindowsServer"
Offer PublisherName Location

----- ------------- --------
Windows-HUB MicrosoftWindowsServer EastUS
WindowsServer MicrosoftWindowsServer EastUS
WindowsServer-HUB MicrosoftWindowsServer EastUS
The Get-AzureRmVMImageSku command will then filter on the publisher and offer name to return a list of
image names.
Get-AzureRmVMImageSku -Location "EastUS" -PublisherName "MicrosoftWindowsServer" -Offer "WindowsServer"
Skus Offer PublisherName Location

---- ----- ------------- --------
2008-R2-SP1 WindowsServer MicrosoftWindowsServer EastUS
2008-R2-SP1-smalldisk WindowsServer MicrosoftWindowsServer EastUS
2012-Datacenter WindowsServer MicrosoftWindowsServer EastUS
2012-Datacenter-smalldisk WindowsServer MicrosoftWindowsServer EastUS
2012-R2-Datacenter WindowsServer MicrosoftWindowsServer EastUS
2012-R2-Datacenter-smalldisk WindowsServer MicrosoftWindowsServer EastUS
2016-Datacenter WindowsServer MicrosoftWindowsServer EastUS
2016-Datacenter-Server-Core WindowsServer MicrosoftWindowsServer EastUS
2016-Datacenter-Server-Core-smalldisk WindowsServer MicrosoftWindowsServer EastUS
2016-Datacenter-smalldisk WindowsServer MicrosoftWindowsServer EastUS
2016-Datacenter-with-Containers WindowsServer MicrosoftWindowsServer EastUS
2016-Datacenter-with-Containers-smalldisk WindowsServer MicrosoftWindowsServer EastUS
2016-Datacenter-with-RDSH WindowsServer MicrosoftWindowsServer EastUS
2016-Nano-Server WindowsServer MicrosoftWindowsServer EastUS
This information can be used to deploy a VM with a specific image. This example deploys a virtual machine using
the latest version of a Windows Server 2016 with Containers image.
New-AzureRmVm `
-Name "myVM2" `
-PublicIpAddressName "myPublicIpAddress2" `
-ImageName "MicrosoftWindowsServer:WindowsServer:2016-Datacenter-with-Containers:latest" `
-Credential $cred `
-AsJob
The -AsJob parameter creates the VM as a background task, so the PowerShell prompts return to you. You can
view details of background jobs with the Get-Job cmdlet.
Understand VM sizes
A virtual machine size determines the amount of compute resources such as CPU, GPU, and memory that are
made available to the virtual machine. Virtual machines need to be created with a size appropriate for the expect
workload. If workload increases, an existing virtual machine can be resized.
VM Sizes
The following table categorizes sizes into use cases.
TYPE COMMON SIZES DESCRIPTION
General purpose Dsv3, Dv3, DSv2, Dv2, DS, D, Av2, A0- Balanced CPU-to-memory. Ideal for
7 dev / test and small to medium
applications and data solutions.
Compute optimized Fs, F High CPU-to-memory. Good for

medium traffic applications, network
appliances, and batch processes.
Memory optimized Esv3, Ev3, M, GS, G, DSv2, DS, Dv2, D High memory-to-core. Great for
relational databases, medium to large
caches, and in-memory analytics.
Storage optimized Ls High disk throughput and IO. Ideal for

Big Data, SQL, and NoSQL databases.
GPU NV, NC Specialized VMs targeted for heavy

graphic rendering and video editing.
High performance H, A8-11 Our most powerful CPU VMs with

optional high-throughput network
interfaces (RDMA).
Find available VM sizes

To see a list of VM sizes available in a particular region, use the Get-AzureRmVMSize command.
Get-AzureRmVMSize -Location "EastUS"
Resize a VM
After a VM has been deployed, it can be resized to increase or decrease resource allocation.
Before resizing a VM, check if the desired size is available on the current VM cluster. The Get-AzureRmVMSize
command returns a list of sizes.
Get-AzureRmVMSize -ResourceGroupName "myResourceGroupVM" -VMName "myVM"
If the desired size is available, the VM can be resized from a powered-on state, however it is rebooted during the
operation.
$vm = Get-AzureRmVM -ResourceGroupName "myResourceGroupVM" -VMName "myVM"

$vm.HardwareProfile.VmSize = "Standard_D4"
Update-AzureRmVM -VM $vm -ResourceGroupName "myResourceGroupVM"
If the desired size is not on the current cluster, the VM needs to be deallocated before the resize operation can
occur. Note, when the VM is powered back on, any data on the temp disk are removed, and the public IP address
change unless a static IP address is being used.
Stop-AzureRmVM -ResourceGroupName "myResourceGroupVM" -Name "myVM" -Force
$vm = Get-AzureRmVM -ResourceGroupName "myResourceGroupVM" -VMName "myVM"
$vm.HardwareProfile.VmSize = "Standard_F4s"
Update-AzureRmVM -VM $vm -ResourceGroupName "myResourceGroupVM"
Start-AzureRmVM -ResourceGroupName "myResourceGroupVM" -Name $vm.name
VM power states
An Azure VM can have one of many power states. This state represents the current state of the VM from the
standpoint of the hypervisor.
Power states
POWER STATE DESCRIPTION
Starting Indicates the virtual machine is being started.
Running Indicates that the virtual machine is running.
Stopping Indicates that the virtual machine is being stopped.
Stopped Indicates that the virtual machine is stopped. Note that

virtual machines in the stopped state still incur compute
charges.
Deallocating Indicates that the virtual machine is being deallocated.
Deallocated Indicates that the virtual machine is completely removed from

the hypervisor but still available in the control plane. Virtual
machines in the Deallocated state do not incur compute
charges.
- Indicates that the power state of the virtual machine is

unknown.
Find power state

To retrieve the state of a particular VM, use the Get-AzureRmVM command. Be sure to specify a valid name for a
virtual machine and resource group.
Get-AzureRmVM `
-Name "myVM" `
-Status | Select @{n="Status"; e={$_.Statuses[1].Code}}
Output:
Status
------
PowerState/running
Management tasks
During the lifecycle of a virtual machine, you may want to run management tasks such as starting, stopping, or
deleting a virtual machine. Additionally, you may want to create scripts to automate repetitive or complex tasks.
Using Azure PowerShell, many common management tasks can be run from the command line or in scripts.
Stop virtual machine
Stop and deallocate a virtual machine with Stop-AzureRmVM:
Stop-AzureRmVM -ResourceGroupName "myResourceGroupVM" -Name "myVM" -Force
If you want to keep the virtual machine in a provisioned state, use the -StayProvisioned parameter.
Start virtual machine
Start-AzureRmVM -ResourceGroupName "myResourceGroupVM" -Name "myVM"
Delete resource group

Deleting a resource group also deletes all resources contained within.
Remove-AzureRmResourceGroup -Name "myResourceGroupVM" -Force
Next steps
In this tutorial, you learned about basic VM creation and management such as how to:
Create and connect to a VM
Select and use VM images
View and use specific VM sizes
Resize a VM
View and understand VM state
Advance to the next tutorial to learn about VM disks.
Create and Manage VM disks
Tutorial - Manage Azure disks with Azure PowerShell
Azure virtual machines use disks to store the VMs operating system, applications, and data. When creating a VM it
is important to choose a disk size and configuration appropriate to the expected workload. This tutorial covers
deploying and managing VM disks. You learn about:
OS disks and temporary disks
Data disks
Standard and Premium disks
Disk performance
Attaching and preparing data disks

Azure tools preinstalled and configured to use with your account. Just click the Copy to copy the code, paste it into
the Cloud Shell, and then press enter to run it. There are a few ways to launch the Cloud Shell:
Default Azure disks

When an Azure virtual machine is created, two disks are automatically attached to the virtual machine.
Operating system disk - Operating system disks can be sized up to 4 terabyte, and hosts the VMs operating
system. The OS disk is assigned a drive letter of c: by default. The disk caching configuration of the OS disk is
optimized for OS performance. The OS disk should not host applications or data. For applications and data, use a
data disk, which is detailed later in this article.
Temporary disk - Temporary disks use a solid-state drive that is located on the same Azure host as the VM. Temp
disks are highly performant and may be used for operations such as temporary data processing. However, if the
VM is moved to a new host, any data stored on a temporary disk is removed. The size of the temporary disk is
determined by the VM size. Temporary disks are assigned a drive letter of d: by default.
Temporary disk sizes
TYPE COMMON SIZES MAX TEMP DISK SIZE (GIB)
General purpose A, B, and D series 1600
Compute optimized F series 576
Memory optimized D, E, G, and M series 6144
Storage optimized L series 5630
GPU N series 1440
High performance A and H series 2000
Azure data disks

Additional data disks can be added for installing applications and storing data. Data disks should be used in any
situation where durable and responsive data storage is desired. Each data disk has a maximum capacity of 4
terabytes. The size of the virtual machine determines how many data disks can be attached to a VM. For each VM
vCPU, two data disks can be attached.
Max data disks per VM
TYPE COMMON SIZES MAX DATA DISKS PER VM
General purpose A, B, and D series 64
Compute optimized F series 64
Memory optimized D, E, G, and M series 64
Storage optimized L series 64
GPU N series 64
High performance A and H series 64
VM disk types
Azure provides two types of disk.
Standard disk
Standard Storage is backed by HDDs, and delivers cost-effective storage while still being performant. Standard
disks are ideal for a cost effective dev and test workload.
Premium disk
Premium disks are backed by SSD -based high-performance, low -latency disk. Perfect for VMs running production
workload. Premium Storage supports DS -series, DSv2-series, GS -series, and FS -series VMs. Premium disks come
in five types (P10, P20, P30, P40, P50), the size of the disk determines the disk type. When selecting, a disk size the
value is rounded up to the next type. For example, if the size is below 128 GB the disk type is P10, or between 129
GB and 512 GB the disk is P20.
Premium disk performance
PREMIUM
STORAGE
DISK TYPE P4 P6 P10 P20 P30 P40 P50
Disk size 32 GB 64 GB 128 GB 512 GB 1,024 GB (1 2,048 GB (2 4,095 GB (4

(round up) TB) TB) TB)
Max IOPS 120 240 500 2,300 5,000 7,500 7,500

per disk
Throughput 25 MB/s 50 MB/s 100 MB/s 150 MB/s 200 MB/s 250 MB/s 250 MB/s
per disk
While the above table identifies max IOPS per disk, a higher level of performance can be achieved by striping
multiple data disks. For instance, 64 data disks can be attached to Standard_GS5 VM. If each of these disks are
sized as a P30, a maximum of 80,000 IOPS can be achieved. For detailed information on max IOPS per VM, see
VM types and sizes.
Create and attach disks

To complete the example in this tutorial, you must have an existing virtual machine. If needed, create a virtual
machine with the following commands.
Set the username and password needed for the administrator account on the virtual machine with Get-Credential:
New-AzureRmVm `
-ResourceGroupName "myResourceGroupDisk" `
-Name "myVM" `
-Credential $cred `
-AsJob
view details of background jobs with the Job cmdlet.
Create the initial configuration with New -AzureRmDiskConfig. The following example configures a disk that is 128
gigabytes in size.
$diskConfig = New-AzureRmDiskConfig `
-CreateOption Empty `
-DiskSizeGB 128
Create the data disk with the New -AzureRmDisk command.

$dataDisk = New-AzureRmDisk `
-ResourceGroupName "myResourceGroupDisk" `
-DiskName "myDataDisk" `
-Disk $diskConfig
Get the virtual machine that you want to add the data disk to with the Get-AzureRmVM command.
$vm = Get-AzureRmVM -ResourceGroupName "myResourceGroupDisk" -Name "myVM"
Add the data disk to the virtual machine configuration with the Add-AzureRmVMDataDisk command.
$vm = Add-AzureRmVMDataDisk `
-VM $vm `
-Name "myDataDisk" `
-CreateOption Attach `
-ManagedDiskId $dataDisk.Id `
-Lun 1
Update the virtual machine with the Update-AzureRmVM command.
Update-AzureRmVM -ResourceGroupName "myResourceGroupDisk" -VM $vm
Prepare data disks

Once a disk has been attached to the virtual machine, the operating system needs to be configured to use the disk.
The following example shows how to manually configure the first disk added to the VM. This process can also be
automated using the custom script extension.
Manual configuration
Create an RDP connection with the virtual machine. Open up PowerShell and run this script.
Get-Disk | Where partitionstyle -eq 'raw' | `

Initialize-Disk -PartitionStyle MBR -PassThru | `
New-Partition -AssignDriveLetter -UseMaximumSize | `
Format-Volume -FileSystem NTFS -NewFileSystemLabel "myDataDisk" -Confirm:$false
Next steps
In this tutorial, you learned about VM disks topics such as:
OS disks and temporary disks
Data disks
Standard and Premium disks
Disk performance
Attaching and preparing data disks
Advance to the next tutorial to learn about automating VM configuration.
Automate VM configuration
Tutorial - Deploy applications to a Windows virtual
machine in Azure with the Custom Script Extension
To configure virtual machines (VMs) in a quick and consistent manner, some form of automation is typically
desired. A common approach to customize a Windows VM is to use Custom Script Extension for Windows. In this
tutorial you learn how to:
Use the Custom Script Extension to install IIS
Create a VM that uses the Custom Script Extension
View a running IIS site after the extension is applied

Custom script extension overview

The Custom Script Extension downloads and executes scripts on Azure VMs. This extension is useful for post
deployment configuration, software installation, or any other configuration / management task. Scripts can be
downloaded from Azure storage or GitHub, or provided to the Azure portal at extension run time.
The Custom Script extension integrates with Azure Resource Manager templates, and can also be run using the
Azure CLI, PowerShell, Azure portal, or the Azure Virtual Machine REST API.
You can use the Custom Script Extension with both Windows and Linux VMs.

First, set an administrator username and password for the VM with Get-Credential:
Now you can create the VM with New -AzureRmVM. The following example creates a VM named myVM in the
EastUS location. If they do not already exist, the resource group myResourceGroupAutomate and supporting
network resources are created. To allow web traffic, the cmdlet also opens port 80.
New-AzureRmVm `
-ResourceGroupName "myResourceGroupAutomate" `
-Name "myVM" `
-OpenPorts 80 `
-Credential $cred
It takes a few minutes for the resources and VM to be created.
Automate IIS install

Use Set-AzureRmVMExtension to install the Custom Script Extension. The extension runs
powershell Add-WindowsFeature Web-Server to install the IIS webserver and then updates the Default.htm page to
show the hostname of the VM:
Set-AzureRmVMExtension -ResourceGroupName "myResourceGroupAutomate" `

-ExtensionName "IIS" `
-VMName "myVM" `
-Publisher Microsoft.Compute `
-ExtensionType CustomScriptExtension `
-TypeHandlerVersion 1.8 `
-SettingString '{"commandToExecute":"powershell Add-WindowsFeature Web-Server; powershell Add-Content -
Path \"C:\\inetpub\\wwwroot\\Default.htm\" -Value $($env:computername)"}'
Test web site

Obtain the public IP address of your load balancer with Get-AzureRmPublicIPAddress. The following example
obtains the IP address for myPublicIPAddress created earlier:
Get-AzureRmPublicIPAddress `
-ResourceGroupName "myResourceGroupAutomate" `
-Name "myPublicIPAddress" | select IpAddress
You can then enter the public IP address in to a web browser. The website is displayed, including the hostname of
the VM that the load balancer distributed traffic to as in the following example:
Next steps
In this tutorial, you automated the IIS install on a VM. You learned how to:
Use the Custom Script Extension to install IIS
Create a VM that uses the Custom Script Extension
View a running IIS site after the extension is applied
Advance to the next tutorial to learn how to create custom VM images.
Create custom VM images
Tutorial: Create a custom image of an Azure VM with
Azure PowerShell
Custom images are like marketplace images, but you create them yourself. Custom images can be used to
bootstrap configurations such as preloading applications, application configurations, and other OS configurations.
In this tutorial, you create your own custom image of an Azure virtual machine. You learn how to:
Sysprep and generalize VMs
Create a custom image
Create a VM from a custom image
List all the images in your subscription
Delete an image
Before you begin

The steps below detail how to take an existing VM and turn it into a re-usable custom image that you can use to
create new VM instances.
To complete the example in this tutorial, you must have an existing virtual machine. If needed, this script sample
can create one for you. When working through the tutorial, replace the resource group and VM names where
needed.

If you choose to install and use the PowerShell locally, this tutorial requires the AzureRM module version 5.7.0 or
later. Run Get-Module -ListAvailable AzureRM to find the version. If you need to upgrade, see Install Azure
PowerShell module.
Prepare VM
To create an image of a virtual machine, you need to prepare the VM by generalizing the VM, deallocating, and
then marking the source VM as generalized in Azure.
Generalize the Windows VM using Sysprep
Sysprep removes all your personal account information, among other things, and prepares the machine to be used
as an image. For details about Sysprep, see How to Use Sysprep: An Introduction.
1. Connect to the virtual machine.
2. Open the Command Prompt window as an administrator. Change the directory to
%windir%\system32\sysprep, and then run sysprep.exe.
3. In the System Preparation Tool dialog box, select Enter System Out-of-Box Experience (OOBE ), and make
sure that the Generalize check box is selected.
4. In Shutdown Options, select Shutdown and then click OK.
5. When Sysprep completes, it shuts down the virtual machine. Do not restart the VM.
Deallocate and mark the VM as generalized
To create an image, the VM needs to be deallocated and marked as generalized in Azure.
Deallocated the VM using Stop-AzureRmVM.
Stop-AzureRmVM -ResourceGroupName myResourceGroup -Name myVM -Force
Set the status of the virtual machine to -Generalized using Set-AzureRmVm.
Set-AzureRmVM -ResourceGroupName myResourceGroup -Name myVM -Generalized
Create the image

Now you can create an image of the VM by using New -AzureRmImageConfig and New -AzureRmImage. The
following example creates an image named myImage from a VM named myVM.
Get the virtual machine.
$vm = Get-AzureRmVM -Name myVM -ResourceGroupName myResourceGroup
Create the image configuration.
$image = New-AzureRmImageConfig -Location EastUS -SourceVirtualMachineId $vm.ID
Create the image.
New-AzureRmImage -Image $image -ImageName myImage -ResourceGroupName myResourceGroup
Create VMs from the image

Now that you have an image, you can create one or more new VMs from the image. Creating a VM from a custom
image is similar to creating a VM using a Marketplace image. When you use a Marketplace image, you have to
provide the information about the image, image provider, offer, SKU, and version. Using the simplified parameter
set for the cmdlet, you just need to provide the name of the custom image as long as it is in the same resource
N ew-AzureRMVM
group.
This example creates a VM named myVMfromImage from the myImage, in the myResourceGroup.
New-AzureRmVm `
-Name "myVMfromImage" `
-ImageName "myImage" `
-VirtualNetworkName "myImageVnet" `
-SubnetName "myImageSubnet" `
-SecurityGroupName "myImageNSG" `
-PublicIpAddressName "myImagePIP" `
-OpenPorts 3389
Image management
Here are some examples of common managed image tasks and how to complete them using PowerShell.
List all images by name.
$images = Get-AzureRMResource -ResourceType Microsoft.Compute/images

$images.name
Delete an image. This example deletes the image named myOldImage from the myResourceGroup.
Remove-AzureRmImage `
-ImageName myOldImage `
-ResourceGroupName myResourceGroup
Next steps
In this tutorial, you created a custom VM image. You learned how to:
Sysprep and generalize VMs
Create a custom image
Create a VM from a custom image
List all the images in your subscription
Delete an image
Advance to the next tutorial to learn about how highly available virtual machines.
Create highly available VMs
Tutorial: Create and deploy highly available virtual
machines with Azure PowerShell
In this tutorial, you learn how to increase the availability and reliability of your Virtual Machine solutions on Azure
using a capability called Availability Sets. Availability sets ensure that the VMs you deploy on Azure are distributed
across multiple isolated hardware nodes in a cluster. Doing this ensures that if a hardware or software failure
within Azure happens, only a subset of your VMs are impacted and that your overall solution remains available
and operational.
In this tutorial, you learn how to:
Create an availability set
Create a VM in an availability set
Check available VM sizes
Check Azure Advisor

Azure tools preinstalled and configured to use with your account. Just click the Copy to copy the code, paste it
into the Cloud Shell, and then press enter to run it. There are a few ways to launch the Cloud Shell:
Availability set overview

An Availability Set is a logical grouping capability that you can use in Azure to ensure that the VM resources you
place within it are isolated from each other when they are deployed within an Azure datacenter. Azure ensures that
the VMs you place within an Availability Set run across multiple physical servers, compute racks, storage units,
and network switches. If a hardware or Azure software failure occurs, only a subset of your VMs are impacted, and
your overall application stays up and continues to be available to your customers. Availability Sets are an essential
capability when you want to build reliable cloud solutions.
Let’s consider a typical VM -based solution where you might have four front-end web servers and 2 back-end
VMs. With Azure, you’d want to define two availability sets before you deploy your VMs: one availability set for
the web tier and one availability set for the back tier. When you create a new VM you can then specify the
availability set as a parameter to the az vm create command, and Azure automatically ensures that the VMs you
create within the available set are isolated across multiple physical hardware resources. If the physical hardware
that one of your Web Server or back-end VMs is running on has a problem, you know that the other instances of
your Web Server and back-end VMs remain running because they are on different hardware.
Use Availability Sets when you want to deploy reliable VM -based solutions in Azure.

You can create an availability set using New -AzureRmAvailabilitySet. In this example, set both the number of
update and fault domains at 2 for the availability set named myAvailabilitySet in the myResourceGroupAvailability
resource group.
Create a resource group.
New-AzureRmResourceGroup -Name myResourceGroupAvailability -Location EastUS
Create a managed availability set using New -AzureRmAvailabilitySet with the -sku aligned parameter.
New-AzureRmAvailabilitySet `
-Name "myAvailabilitySet" `
-ResourceGroupName "myResourceGroupAvailability" `
-Sku aligned `
-PlatformFaultDomainCount 2 `
-PlatformUpdateDomainCount 2
Create VMs inside an availability set

VMs must be created within the availability set to make sure they are correctly distributed across the hardware.
You can't add an existing VM to an availability set after it is created.
The hardware in a location is divided in to multiple update domains and fault domains. An update domain is a
group of VMs and underlying physical hardware that can be rebooted at the same time. VMs in the same fault
domain share common storage as well as a common power source and network switch.
When you create a VM with New -AzureRmVM, you use the -AvailabilitySetName parameter to specify the name
of the availability set.
First, set an administrator username and password for the VM with Get-Credential:
Now create two VMs with New -AzureRmVM in the availability set.
for ($i=1; $i -le 2; $i++)
{
New-AzureRmVm `
-Name "myVM$i" `
-PublicIpAddressName "myPublicIpAddress$i" `
-AvailabilitySetName "myAvailabilitySet" `
-Credential $cred
}
view details of background jobs with the Job cmdlet. It takes a few minutes to create and configure both VMs.
When finished, you have two virtual machines distributed across the underlying hardware.
If you look at the availability set in the portal by going to Resource Groups > myResourceGroupAvailability >
myAvailabilitySet, you should see how the VMs are distributed across the two fault and update domains.
Check for available VM sizes

You can add more VMs to the availability set later, but you need to know what VM sizes are available on the
hardware. Use Get-AzureRMVMSize to list all the available sizes on the hardware cluster for the availability set.
Get-AzureRmVMSize `
-AvailabilitySetName "myAvailabilitySet"
Check Azure Advisor

You can also use Azure Advisor to get more information on ways to improve the availability of your VMs. Azure
Advisor helps you follow best practices to optimize your Azure deployments. It analyzes your resource
configuration and usage telemetry and then recommends solutions that can help you improve the cost
effectiveness, performance, high availability, and security of your Azure resources.
Sign in to the Azure portal, select All services, and type Advisor. The Advisor dashboard displays personalized
recommendations for the selected subscription. For more information, see Get started with Azure Advisor.
Next steps
In this tutorial, you learned how to:
Create a VM in an availability set
Check available VM sizes
Check Azure Advisor
Advance to the next tutorial to learn about virtual machine scale sets.
Create a VM scale set
Tutorial: Create a virtual machine scale set and
deploy a highly available app on Windows with
Azure PowerShell
A virtual machine scale set allows you to deploy and manage a set of identical, auto-scaling virtual machines. You
can scale the number of VMs in the scale set manually, or define rules to autoscale based on resource usage such
as CPU, memory demand, or network traffic. In this tutorial, you deploy a virtual machine scale set in Azure. You
learn how to:
Use the Custom Script Extension to define an IIS site to scale
Create a load balancer for your scale set
Create a virtual machine scale set
Increase or decrease the number of instances in a scale set
Create autoscale rules

Scale Set overview

A virtual machine scale set allows you to deploy and manage a set of identical, auto-scaling virtual machines. VMs
in a scale set are distributed across logic fault and update domains in one or more placement groups. These are
groups of similarly configured VMs, similar to availability sets.
VMs are created as needed in a scale set. You define autoscale rules to control how and when VMs are added or
removed from the scale set. These rules can trigger based on metrics such as CPU load, memory usage, or
network traffic.
Scale sets support up to 1,000 VMs when you use an Azure platform image. For workloads with significant
installation or VM customization requirements, you may wish to Create a custom VM image. You can create up to
300 VMs in a scale set when using a custom image.
Create a scale set

Create a virtual machine scale set with New -AzureRmVmss. The following example creates a scale set named
myScaleSet that uses the Windows Server 2016 Datacenter platform image. The Azure network resources for
virtual network, public IP address, and load balancer are automatically created. When prompted, provide your own
desired administrative credentials for the VM instances in the scale set:
New-AzureRmVmss `
-ResourceGroupName "myResourceGroupScaleSet" `
-VMScaleSetName "myScaleSet" `
-PublicIpAddressName "myPublicIPAddress" `
-LoadBalancerName "myLoadBalancer" `
-UpgradePolicy "Automatic"
It takes a few minutes to create and configure all the scale set resources and VMs.
Deploy sample application

To test your scale set, install a basic web application. The Azure Custom Script Extension is used to download and
run a script that installs IIS on the VM instances. This extension is useful for post deployment configuration,
software installation, or any other configuration / management task. For more information, see the Custom Script
Extension overview.
Use the Custom Script Extension to install a basic IIS web server. Apply the Custom Script Extension that installs
IIS as follows:
# Define the script for your Custom Script Extension to run

$publicSettings = @{
"fileUris" = (,"https://raw.githubusercontent.com/Azure-Samples/compute-automation-
configurations/master/automate-iis.ps1");
"commandToExecute" = "powershell -ExecutionPolicy Unrestricted -File automate-iis.ps1"
}
# Get information about the scale set

$vmss = Get-AzureRmVmss `
-VMScaleSetName "myScaleSet"
# Use Custom Script Extension to install IIS and configure basic website
Add-AzureRmVmssExtension -VirtualMachineScaleSet $vmss `
-Name "customScript" `
-Publisher "Microsoft.Compute" `
-Type "CustomScriptExtension" `
-Setting $publicSettings
# Update the scale set and apply the Custom Script Extension to the VM instances
Update-AzureRmVmss `
-Name "myScaleSet" `
-VirtualMachineScaleSet $vmss
Test your scale set

To see your scale set in action, obtain the public IP address of your load balancer with Get-
AzureRmPublicIPAddress. The following example obtains the IP address for myPublicIP created as part of the
scale set:
-Name "myPublicIPAddress" | select IpAddress
Enter the public IP address in to a web browser. The web app is displayed, including the hostname of the VM that
the load balancer distributed traffic to:
To see the scale set in action, you can force-refresh your web browser to see the load balancer distribute traffic
across all the VMs running your app.
Management tasks
Throughout the lifecycle of the scale set, you may need to run one or more management tasks. Additionally, you
may want to create scripts that automate various lifecycle-tasks. Azure PowerShell provides a quick way to do
those tasks. Here are a few common tasks.
View VMs in a scale set
To view a list of VM instances in a scale set, use Get-AzureRmVmssVM as follows:
Get-AzureRmVmssVM -ResourceGroupName "myResourceGroupScaleSet" -VMScaleSetName "myScaleSet"
The following example output shows two VM instances in the scale set:
ResourceGroupName Name Location Sku InstanceID ProvisioningState

----------------- ---- -------- --- ---------- -----------------
MYRESOURCEGROUPSCALESET myScaleSet_0 eastus Standard_DS1_v2 0 Succeeded
MYRESOURCEGROUPSCALESET myScaleSet_1 eastus Standard_DS1_v2 1 Succeeded
To view additional information about a specific VM instance, add the -InstanceId parameter to Get-
AzureRmVmssVM. The following example views information about VM instance 1:
Get-AzureRmVmssVM -ResourceGroupName "myResourceGroupScaleSet" -VMScaleSetName "myScaleSet" -InstanceId "1"
Increase or decrease VM instances

To see the number of instances you currently have in a scale set, use Get-AzureRmVmss and query on
sku.capacity:
Get-AzureRmVmss -ResourceGroupName "myResourceGroupScaleSet" `

-VMScaleSetName "myScaleSet" | `
Select -ExpandProperty Sku
You can then manually increase or decrease the number of virtual machines in the scale set with Update-
AzureRmVmss. The following example sets the number of VMs in your scale set to 3:
# Get current scale set

$scaleset = Get-AzureRmVmss `
# Set and update the capacity of your scale set

$scaleset.sku.capacity = 3
Update-AzureRmVmss -ResourceGroupName "myResourceGroupScaleSet" `
-Name "myScaleSet" `
-VirtualMachineScaleSet $scaleset
If takes a few minutes to update the specified number of instances in your scale set.
Configure autoscale rules
Rather than manually scaling the number of instances in your scale set, you define autoscale rules. These rules
monitor the instances in your scale set and respond accordingly based on metrics and thresholds you define. The
following example scales out the number of instances by one when the average CPU load is greater than 60% over
a 5-minute period. If the average CPU load then drops below 30% over a 5-minute period, the instances are scaled
in by one instance:
# Define your scale set information
$mySubscriptionId = (Get-AzureRmSubscription)[0].Id
$myResourceGroup = "myResourceGroupScaleSet"
$myScaleSet = "myScaleSet"
$myLocation = "East US"
$myScaleSetId = (Get-AzureRmVmss -ResourceGroupName $myResourceGroup -VMScaleSetName $myScaleSet).Id
# Create a scale up rule to increase the number instances after 60% average CPU usage exceeded for a 5-minute
period
$myRuleScaleUp = New-AzureRmAutoscaleRule `
-MetricName "Percentage CPU" `
-MetricResourceId $myScaleSetId `
-Operator GreaterThan `
-MetricStatistic Average `
-Threshold 60 `
-TimeGrain 00:01:00 `
-TimeWindow 00:05:00 `
-ScaleActionCooldown 00:05:00 `
-ScaleActionDirection Increase `
-ScaleActionValue 1
# Create a scale down rule to decrease the number of instances after 30% average CPU usage over a 5-minute
period
$myRuleScaleDown = New-AzureRmAutoscaleRule `
-MetricName "Percentage CPU" `
-MetricResourceId $myScaleSetId `
-Operator LessThan `
-MetricStatistic Average `
-Threshold 30 `
-TimeGrain 00:01:00 `
-TimeWindow 00:05:00 `
-ScaleActionCooldown 00:05:00 `
-ScaleActionDirection Decrease `
-ScaleActionValue 1
# Create a scale profile with your scale up and scale down rules
$myScaleProfile = New-AzureRmAutoscaleProfile `
-DefaultCapacity 2 `
-MaximumCapacity 10 `
-MinimumCapacity 2 `
-Rule $myRuleScaleUp,$myRuleScaleDown `
-Name "autoprofile"
# Apply the autoscale rules

Add-AzureRmAutoscaleSetting `
-Location $myLocation `
-Name "autosetting" `
-ResourceGroup $myResourceGroup `
-TargetResourceId $myScaleSetId `
-AutoscaleProfile $myScaleProfile
For more design information on the use of autoscale, see autoscale best practices.
Next steps
In this tutorial, you created a virtual machine scale set. You learned how to:
Use the Custom Script Extension to define an IIS site to scale
Create a load balancer for your scale set
Increase or decrease the number of instances in a scale set
Create autoscale rules
Advance to the next tutorial to learn more about load balancing concepts for virtual machines.
Load balance virtual machines
Tutorial: Load balance Windows virtual machines in
Azure to create a highly available application with
Azure PowerShell
Load balancing provides a higher level of availability by spreading incoming requests across multiple virtual
machines. In this tutorial, you learn about the different components of the Azure load balancer that distribute
traffic and provide high availability. You learn how to:
Create an Azure load balancer
Create a load balancer health probe
Create load balancer traffic rules
Use the Custom Script Extension to create a basic IIS site
Create virtual machines and attach to a load balancer
View a load balancer in action
Add and remove VMs from a load balancer

Azure load balancer overview

An Azure load balancer is a Layer-4 (TCP, UDP ) load balancer that provides high availability by distributing
incoming traffic among healthy VMs. A load balancer health probe monitors a given port on each VM and only
distributes traffic to an operational VM.
You define a front-end IP configuration that contains one or more public IP addresses. This front-end IP
configuration allows your load balancer and applications to be accessible over the Internet.
Virtual machines connect to a load balancer using their virtual network interface card (NIC ). To distribute traffic to
the VMs, a back-end address pool contains the IP addresses of the virtual (NICs) connected to the load balancer.
To control the flow of traffic, you define load balancer rules for specific ports and protocols that map to your VMs.
Create Azure load balancer

This section details how you can create and configure each component of the load balancer. Before you can create
your load balancer, create a resource group with New -AzureRmResourceGroup. The following example creates a
resource group named myResourceGroupLoadBalancer in the EastUS location:
New-AzureRmResourceGroup `
-ResourceGroupName "myResourceGroupLoadBalancer" `
-Location "EastUS"
Create a public IP address

To access your app on the Internet, you need a public IP address for the load balancer. Create a public IP address
with New -AzureRmPublicIpAddress. The following example creates a public IP address named myPublicIP in the
myResourceGroupLoadBalancer resource group:
$publicIP = New-AzureRmPublicIpAddress `
-AllocationMethod "Static" `
-Name "myPublicIP"
Create a load balancer

Create a frontend IP pool with New -AzureRmLoadBalancerFrontendIpConfig. The following example creates a
frontend IP pool named myFrontEndPool and attaches the myPublicIP address:
$frontendIP = New-AzureRmLoadBalancerFrontendIpConfig `
-Name "myFrontEndPool" `
-PublicIpAddress $publicIP
Create a backend address pool with New -AzureRmLoadBalancerBackendAddressPoolConfig. The VMs attach to
this backend pool in the remaining steps. The following example creates a backend address pool named
myBackEndPool:
$backendPool = New-AzureRmLoadBalancerBackendAddressPoolConfig -Name "myBackEndPool"
Now, create the load balancer with New -AzureRmLoadBalancer. The following example creates a load balancer
named myLoadBalancer using the frontend and backend IP pools created in the preceding steps:
$lb = New-AzureRmLoadBalancer `
-Name "myLoadBalancer" `
-FrontendIpConfiguration $frontendIP `
-BackendAddressPool $backendPool
Create a health probe

To allow the load balancer to monitor the status of your app, you use a health probe. The health probe dynamically
adds or removes VMs from the load balancer rotation based on their response to health checks. By default, a VM
is removed from the load balancer distribution after two consecutive failures at 15-second intervals. You create a
health probe based on a protocol or a specific health check page for your app.
The following example creates a TCP probe. You can also create custom HTTP probes for more fine grained health
checks. When using a custom HTTP probe, you must create the health check page, such as healthcheck.aspx. The
probe must return an HTTP 200 OK response for the load balancer to keep the host in rotation.
To create a TCP health probe, you use Add-AzureRmLoadBalancerProbeConfig. The following example creates a
health probe named myHealthProbe that monitors each VM on TCP port 80:
Add-AzureRmLoadBalancerProbeConfig `
-Name "myHealthProbe" `
-LoadBalancer $lb `
-Protocol tcp `
-Port 80 `
-IntervalInSeconds 15 `
-ProbeCount 2
To apply the health probe, update the load balancer with Set-AzureRmLoadBalancer:
Set-AzureRmLoadBalancer -LoadBalancer $lb
Create a load balancer rule

A load balancer rule is used to define how traffic is distributed to the VMs. You define the front-end IP
configuration for the incoming traffic and the back-end IP pool to receive the traffic, along with the required
source and destination port. To make sure only healthy VMs receive traffic, you also define the health probe to use.
Create a load balancer rule with Add-AzureRmLoadBalancerRuleConfig. The following example creates a load
balancer rule named myLoadBalancerRule and balances traffic on TCP port 80:
$probe = Get-AzureRmLoadBalancerProbeConfig -LoadBalancer $lb -Name "myHealthProbe"
Add-AzureRmLoadBalancerRuleConfig `
-Name "myLoadBalancerRule" `
-LoadBalancer $lb `
-FrontendIpConfiguration $lb.FrontendIpConfigurations[0] `
-BackendAddressPool $lb.BackendAddressPools[0] `
-Protocol Tcp `
-FrontendPort 80 `
-BackendPort 80 `
-Probe $probe
Update the load balancer with Set-AzureRmLoadBalancer:
Set-AzureRmLoadBalancer -LoadBalancer $lb
Configure virtual network

Before you deploy some VMs and can test your balancer, create the supporting virtual network resources. For
more information about virtual networks, see the Manage Azure Virtual Networks tutorial.
Create network resources
Create a virtual network with New -AzureRmVirtualNetwork. The following example creates a virtual network
named myVnet with mySubnet:
# Create subnet config
$subnetConfig = New-AzureRmVirtualNetworkSubnetConfig `
-Name "mySubnet" `
-AddressPrefix 192.168.1.0/24
# Create the virtual network

$vnet = New-AzureRmVirtualNetwork `
-Name "myVnet" `
-AddressPrefix 192.168.0.0/16 `
-Subnet $subnetConfig
Virtual NICs are created with New -AzureRmNetworkInterface. The following example creates three virtual NICs.
(One virtual NIC for each VM you create for your app in the following steps). You can create additional virtual
NICs and VMs at any time and add them to the load balancer:
for ($i=1; $i -le 3; $i++)

{
New-AzureRmNetworkInterface `
-Name myVM$i `
-Subnet $vnet.Subnets[0] `
-LoadBalancerBackendAddressPool $lb.BackendAddressPools[0]
}
Create virtual machines

To improve the high availability of your app, place your VMs in an availability set.
Create an availability set with New -AzureRmAvailabilitySet. The following example creates an availability set
named myAvailabilitySet:
$availabilitySet = New-AzureRmAvailabilitySet `
-Name "myAvailabilitySet" `
-Sku aligned `
-PlatformUpdateDomainCount 2
Set an administrator username and password for the VMs with Get-Credential:
Now you can create the VMs with New -AzureRmVM. The following example creates three VMs and the required
virtual network components if they do not already exist:
for ($i=1; $i -le 3; $i++)
{
New-AzureRmVm `
-Name "myVM$i" `
-OpenPorts 80 `
-AvailabilitySetName "myAvailabilitySet" `
-Credential $cred `
-AsJob
}
view details of background jobs with the Job cmdlet. It takes a few minutes to create and configure all three VMs.
Install IIS with Custom Script Extension
In a previous tutorial on How to customize a Windows virtual machine, you learned how to automate VM
customization with the Custom Script Extension for Windows. You can use the same approach to install and
configure IIS on your VMs.
Use Set-AzureRmVMExtension to install the Custom Script Extension. The extension runs
powershell Add-WindowsFeature Web-Server to install the IIS webserver and then updates the Default.htm page to
show the hostname of the VM:
for ($i=1; $i -le 3; $i++)

{
Set-AzureRmVMExtension `
-VMName myVM$i `
-SettingString '{"commandToExecute":"powershell Add-WindowsFeature Web-Server; powershell Add-Content -
Path \"C:\\inetpub\\wwwroot\\Default.htm\" -Value $($env:computername)"}' `
-Location EastUS
}
Test load balancer

Obtain the public IP address of your load balancer with Get-AzureRmPublicIPAddress. The following example
obtains the IP address for myPublicIP created earlier:
-Name "myPublicIP" | select IpAddress
You can then enter the public IP address in to a web browser. The website is displayed, including the hostname of
the VM that the load balancer distributed traffic to as in the following example:
To see the load balancer distribute traffic across all three VMs running your app, you can force-refresh your web
browser.
Add and remove VMs

You may need to perform maintenance on the VMs running your app, such as installing OS updates. To deal with
increased traffic to your app, you may need to add additional VMs. This section shows you how to remove or add
a VM from the load balancer.
Remove a VM from the load balancer
Get the network interface card with Get-AzureRmNetworkInterface, then set the
LoadBalancerBackendAddressPools property of the virtual NIC to $null. Finally, update the virtual NIC.:
$nic = Get-AzureRmNetworkInterface `
-Name "myVM2"
$nic.Ipconfigurations[0].LoadBalancerBackendAddressPools=$null
Set-AzureRmNetworkInterface -NetworkInterface $nic
To see the load balancer distribute traffic across the remaining two VMs running your app you can force-refresh
your web browser. You can now perform maintenance on the VM, such as installing OS updates or performing a
VM reboot.
Add a VM to the load balancer
After performing VM maintenance, or if you need to expand capacity, set the LoadBalancerBackendAddressPools
property of the virtual NIC to the BackendAddressPool from Get-AzureRMLoadBalancer:
Get the load balancer:
$lb = Get-AzureRMLoadBalancer `
-ResourceGroupName myResourceGroupLoadBalancer `
-Name myLoadBalancer
$nic.IpConfigurations[0].LoadBalancerBackendAddressPools=$lb.BackendAddressPools[0]
Set-AzureRmNetworkInterface -NetworkInterface $nic
Next steps
In this tutorial, you created a load balancer and attached VMs to it. You learned how to:
Create an Azure load balancer
Create a load balancer health probe
Create load balancer traffic rules
Use the Custom Script Extension to create a basic IIS site
Create virtual machines and attach to a load balancer
View a load balancer in action
Add and remove VMs from a load balancer
Advance to the next tutorial to learn how to manage VM networking.
Manage VMs and virtual networks
Tutorial: Create and manage Azure virtual networks
for Windows virtual machines with Azure PowerShell
Azure virtual machines use Azure networking for internal and external network communication. This tutorial walks
through deploying two virtual machines and configuring Azure networking for these VMs. The examples in this
tutorial assume that the VMs are hosting a web application with a database back-end, however an application is
not deployed in the tutorial. In this tutorial, you learn how to:
Create a virtual network and subnet
Create a front-end VM
Secure network traffic
Create back-end VM

VM networking overview
Azure virtual networks enable secure network connections between virtual machines, the internet, and other Azure
services such as Azure SQL database. Virtual networks are broken down into logical segments called subnets.
Subnets are used to control network flow, and as a security boundary. When deploying a VM, it generally includes
a virtual network interface, which is attached to a subnet.
While completing this tutorial, you can see these resources created:
myVNet - The virtual network that the VMs use to communicate with each other and the internet.
myFrontendSubnet - The subnet in myVNet used by the front-end resources.
myPublicIPAddress - The public IP address used to access myFrontendVM from the internet.
myFrontendNic - The network interface used by myFrontendVM to communicate with myBackendVM.
myFrontendVM - The VM used to communicate between the internet and myBackendVM.
myBackendNSG - The network security group that controls communication between the myFrontendVM and
myBackendVM.
myBackendSubnet - The subnet associated with myBackendNSG and used by the back-end resources.
myBackendNic - The network interface used by myBackendVM to communicate with myFrontendVM.
myBackendVM - The VM that uses port 1433 to communicate with myFrontendVM.

For this tutorial, a single virtual network is created with two subnets. A front-end subnet for hosting a web
application, and a back-end subnet for hosting a database server.
Before you can create a virtual network, create a resource group using New -AzureRmResourceGroup. The
following example creates a resource group named myRGNetwork in the EastUS location:
New-AzureRmResourceGroup -ResourceGroupName myRGNetwork -Location EastUS
Create subnet configurations

Create a subnet configuration named myFrontendSubnet using New -AzureRmVirtualNetworkSubnetConfig:
$frontendSubnet = New-AzureRmVirtualNetworkSubnetConfig `
-Name myFrontendSubnet `
And, create a subnet configuration named myBackendSubnet:
$backendSubnet = New-AzureRmVirtualNetworkSubnetConfig `
-Name myBackendSubnet `
Create virtual network

Create a VNET named myVNet using myFrontendSubnet and myBackendSubnet using New -
AzureRmVirtualNetwork:
-ResourceGroupName myRGNetwork `
-Location EastUS `
-Name myVNet `
-Subnet $frontendSubnet, $backendSubnet
At this point, a network has been created and segmented into two subnets, one for front-end services, and another
for back-end services. In the next section, virtual machines are created and connected to these subnets.

A public IP address allows Azure resources to be accessible on the internet. The allocation method of the public IP
address can be configured as dynamic or static. By default, a public IP address is dynamically allocated. Dynamic
IP addresses are released when a VM is deallocated. This behavior causes the IP address to change during any
operation that includes a VM deallocation.
The allocation method can be set to static, which ensures that the IP address remains assigned to a VM, even
during a deallocated state. When using a statically allocated IP address, the IP address itself cannot be specified.
Instead, it is allocated from a pool of available addresses.
Create a public IP address named myPublicIPAddress using New -AzureRmPublicIpAddress:
$pip = New-AzureRmPublicIpAddress `
-Location EastUS `
-AllocationMethod Dynamic `
-Name myPublicIPAddress
You could change the -AllocationMethod parameter to Static to assign a static public IP address.
For a VM to communicate in a virtual network, it needs a virtual network interface (NIC ). Create a NIC using
New -AzureRmNetworkInterface:
$frontendNic = New-AzureRmNetworkInterface `
-Location EastUS `
-Name myFrontend `
-SubnetId $vnet.Subnets[0].Id `
-PublicIpAddressId $pip.Id
Set the username and password needed for the administrator account on the VM using Get-Credential. You use
these credentials to connect to the VM in additional steps:
Create the VMs using New -AzureRmVM.
New-AzureRmVM `
-Credential $cred `
-Name myFrontend `
-PublicIpAddressName myPublicIPAddress `
-Size Standard_D1 `
-SubnetName myFrontendSubnet `
-VirtualNetworkName myVNet

A network security group (NSG ) contains a list of security rules that allow or deny network traffic to resources
connected to Azure Virtual Networks (VNet). NSGs can be associated to subnets or individual network interfaces.
When an NSG is associated with a network interface, it applies only the associated VM. When an NSG is
associated to a subnet, the rules apply to all resources connected to the subnet.
Network security group rules
NSG rules define networking ports over which traffic is allowed or denied. The rules can include source and
destination IP address ranges so that traffic is controlled between specific systems or subnets. NSG rules also
include a priority (between 1—and 4096). Rules are evaluated in the order of priority. A rule with a priority of 100
is evaluated before a rule with priority 200.
All NSGs contain a set of default rules. The default rules cannot be deleted, but because they are assigned the
lowest priority, they can be overridden by the rules that you create.
Virtual network - Traffic originating and ending in a virtual network is allowed both in inbound and outbound
directions.
Internet - Outbound traffic is allowed, but inbound traffic is blocked.
Load balancer - Allow Azure’s load balancer to probe the health of your VMs and role instances. If you are not
using a load balanced set, you can override this rule.
Create network security groups
Create an inbound rule named myFrontendNSGRule to allow incoming web traffic on myFrontendVM using
New -AzureRmNetworkSecurityRuleConfig:
$nsgFrontendRule = New-AzureRmNetworkSecurityRuleConfig `
-Name myFrontendNSGRule `
-Protocol Tcp `
-Direction Inbound `
-Priority 200 `
-SourceAddressPrefix * `
-SourcePortRange * `
-DestinationAddressPrefix * `
-DestinationPortRange 80 `
-Access Allow
You can limit internal traffic to myBackendVM from only myFrontendVM by creating an NSG for the back-end
subnet. The following example creates an NSG rule named myBackendNSGRule:
$nsgBackendRule = New-AzureRmNetworkSecurityRuleConfig `
-Name myBackendNSGRule `
-Protocol Tcp `
-Priority 100 `
-SourceAddressPrefix 10.0.0.0/24 `
-DestinationPortRange 1433 `
-Access Allow
Add a network security group named myFrontendNSG using New -AzureRmNetworkSecurityGroup:
$nsgFrontend = New-AzureRmNetworkSecurityGroup `
-Location EastUS `
-Name myFrontendNSG `
-SecurityRules $nsgFrontendRule
Now, add a network security group named myBackendNSG using New -AzureRmNetworkSecurityGroup:
$nsgBackend = New-AzureRmNetworkSecurityGroup `
-Location EastUS `
-Name myBackendNSG `
-SecurityRules $nsgBackendRule
Add the network security groups to the subnets:
$vnet = Get-AzureRmVirtualNetwork `
-Name myVNet
$frontendSubnet = $vnet.Subnets[0]
$backendSubnet = $vnet.Subnets[1]
$frontendSubnetConfig = Set-AzureRmVirtualNetworkSubnetConfig `
-VirtualNetwork $vnet `
-Name myFrontendSubnet `
-AddressPrefix $frontendSubnet.AddressPrefix `
-NetworkSecurityGroup $nsgFrontend
$backendSubnetConfig = Set-AzureRmVirtualNetworkSubnetConfig `
-Name myBackendSubnet `
-AddressPrefix $backendSubnet.AddressPrefix `
-NetworkSecurityGroup $nsgBackend
Set-AzureRmVirtualNetwork -VirtualNetwork $vnet
Create a back-end VM
The easiest way to create the back-end VM for this tutorial is by using a SQL Server image. This tutorial only
creates the VM with the database server, but doesn't provide information about accessing the database.
Create myBackendNic:
$backendNic = New-AzureRmNetworkInterface `
-Location EastUS `
-Name myBackend `
-SubnetId $vnet.Subnets[1].Id
Set the username and password needed for the administrator account on the VM with Get-Credential:
Create myBackendVM.
New-AzureRmVM `
-Credential $cred `
-Name myBackend `
-ImageName "MicrosoftSQLServer:SQL2016SP1-WS2016:Enterprise:latest" `
-SubnetName MyBackendSubnet `
-VirtualNetworkName myVNet
The image that is used has SQL Server installed, but is not used in this tutorial. It is included to show you how you
can configure a VM to handle web traffic and a VM to handle database management.
Next steps
In this tutorial, you created and secured Azure networks as related to virtual machines.
Create a back-end VM
Advance to the next tutorial to learn about monitoring securing data on virtual machines using Azure backup.
Back up Windows virtual machines in Azure
Tutorial: Back up and restore files for Windows virtual
machines in Azure
You can protect your data by taking backups at regular intervals. Azure Backup creates recovery points that are
stored in geo-redundant recovery vaults. When you restore from a recovery point, you can restore the whole VM
or specific files. This article explains how to restore a single file to a VM running Windows Server and IIS. If you
don't already have a VM to use, you can create one using the Windows quickstart. In this tutorial you learn how to:
Create a backup of a VM
Schedule a daily backup
Restore a file from a backup
Backup overview
When the Azure Backup service initiates a backup job, it triggers the backup extension to take a point-in-time
snapshot. The Azure Backup service uses the VMSnapshot extension. The extension is installed during the first VM
backup if the VM is running. If the VM is not running, the Backup service takes a snapshot of the underlying
storage (since no application writes occur while the VM is stopped).
When taking a snapshot of Windows VMs, the Backup service coordinates with the Volume Shadow Copy Service
(VSS ) to get a consistent snapshot of the virtual machine's disks. Once the Azure Backup service takes the
snapshot, the data is transferred to the vault. To maximize efficiency, the service identifies and transfers only the
blocks of data that have changed since the previous backup.
When the data transfer is complete, the snapshot is removed and a recovery point is created.
Create a backup
Create a simple scheduled daily backup to a Recovery Services Vault.
1. Sign in to the Azure portal.
2. In the menu on the left, select Virtual machines.
3. From the list, select a VM to back up.
4. On the VM blade, in the Settings section, click Backup. The Enable backup blade opens.
5. In Recovery Services vault, click Create new and provide the name for the new vault. A new vault is created
in the same Resource Group and location as the virtual machine.
6. Click Backup policy. For this example, keep the defaults and click OK.
7. On the Enable backup blade, click Enable Backup. This creates a daily backup based on the default schedule.
8. To create an initial recovery point, on the Backup blade click Backup now.
9. On the Backup Now blade, click the calendar icon, use the calendar control to select the last day this recovery
point is retained, and click Backup.
10. In the Backup blade for your VM, you see the number of recovery points that are complete.
The first backup takes about 20 minutes. Proceed to the next part of this tutorial after your backup is finished.
Recover a file
If you accidentally delete or make changes to a file, you can use File Recovery to recover the file from your backup
vault. File Recovery uses a script that runs on the VM, to mount the recovery point as local drive. These drives
remain mounted for 12 hours so that you can copy files from the recovery point and restore them to the VM.
In this example, we show how to recover the image file that is used in the default web page for IIS.
1. Open a browser and connect to the IP address of the VM to show the default IIS page.
2. Connect to the VM.

3. On the VM, open File Explorer and navigate to \inetpub\wwwroot and delete the file iisstart.png.
4. On your local computer, refresh the browser to see that the image on the default IIS page is gone.
5. On your local computer, open a new tab and go the Azure portal.
6. In the menu on the left, select Virtual machines and select the VM from the list.
7. On the VM blade, in the Settings section, click Backup. The Backup blade opens.
8. In the menu at the top of the blade, select File Recovery. The File Recovery blade opens.
9. In Step 1: Select recovery point, select a recovery point from the drop-down.
10. In Step 2: Download script to browse and recover files, click the Download Executable button. Save the
file to your Downloads folder.
11. On your local computer, open File Explorer and navigate to your Downloads folder and copy the
downloaded .exe file. The filename is prefixed by your VM name.
12. On your VM (over the RDP connection) paste the .exe file to the Desktop of your VM.
13. Navigate to the desktop of your VM and double-click on the .exe. This launches a command prompt and then
mount the recovery point as a file share that you can access. When it is finished creating the share, type q to
close the command prompt.
14. On your VM, open File Explorer and navigate to the drive letter that was used for the file share.
15. Navigate to \inetpub\wwwroot and copy iisstart.png from the file share and paste it into \inetpub\wwwroot.
For example, copy F:\inetpub\wwwroot\iisstart.png and paste it into c:\inetpub\wwwroot to recover the file.
16. On your local computer, open the browser tab where you are connected to the IP address of the VM showing
the IIS default page. Press CTRL + F5 to refresh the browser page. You should now see that the image has
been restored.
17. On your local computer, go back to the browser tab for the Azure portal and in Step 3: Unmount the disks
after recovery click the Unmount Disks button. If you forget to do this step, the connection to the
mountpoint is automatically close after 12 hours. After those 12 hours, you need to download a new script to
create a new mountpoint.
Next steps
In this tutorial, you learned how to:
Create a backup of a VM
Schedule a daily backup
Restore a file from a backup
Advance to the next tutorial to learn about monitoring virtual machines.
Govern virtual machines
Tutorial: Learn about Windows virtual machine
governance with Azure PowerShell
When deploying resources to Azure, you have tremendous flexibility when deciding what types of resources to
deploy, where they are located, and how to set them up. However, that flexibility may open more options than you
would like to allow in your organization. As you consider deploying resources to Azure, you might be wondering:
How do I meet legal requirements for data sovereignty in certain countries?
How do I control costs?
How do I ensure that someone does not inadvertently change a critical system?
How do I track resource costs and bill it accurately?
This article addresses those questions. Specifically, you:
Assign users to roles and assign the roles to a scope so users have permission to perform expected actions but
not more actions.
Apply policies that prescribe conventions for resources in your subscription.
Lock resources that are critical to your system.
Tag resources so you can track them by values that make sense to your organization.
This article focuses on the tasks you take to implement governance. For a broader discussion of the concepts, see
Governance in Azure.

The examples in this article require version 6.0 or later of Azure PowerShell. If you're running PowerShell locally
and you don't have version 6.0 or later, update your version. You also need to run Connect-AzureRmAccount to create
a connection with Azure. For local installations, you must also download the Azure AD PowerShell module to
create a new Azure Active Directory group.
Understand scope
Before creating any items, let's review the concept of scope. Azure provides four levels of management:
management groups, subscription, resource group, and resource. Management groups are in a preview release.
The following image shows an example of these layers.
You apply management settings at any of these levels of scope. The level you select determines how widely the
setting is applied. Lower levels inherit settings from higher levels. When you apply a setting to the subscription,
that setting is applied to all resource groups and resources in your subscription. When you apply a setting on the
resource group, that setting is applied the resource group and all its resources. However, another resource group
does not have that setting.
Usually, it makes sense to apply critical settings at higher levels and project-specific requirements at lower levels.
For example, you might want to make sure all resources for your organization are deployed to certain regions. To
accomplish this requirement, apply a policy to the subscription that specifies the allowed locations. As other users
in your organization add new resource groups and resources, the allowed locations are automatically enforced.
In this tutorial, you apply all management settings to a resource group so you can easily remove those settings
when done.
Let's create that resource group.
New-AzureRmResourceGroup -Name myResourceGroup -Location EastUS
Currently, the resource group is empty.
Role-based access control

You want to make sure users in your organization have the right level of access to these resources. You don't want
to grant unlimited access to users, but you also need to make sure they can do their work. Role-based access
control enables you to manage which users have permission to complete specific actions at a scope.
To create and remove role assignments, users must have Microsoft.Authorization/roleAssignments/* access. This
access is granted through the Owner or User Access Administrator roles.
For managing virtual machine solutions, there are three resource-specific roles that provide commonly needed
access:
Virtual Machine Contributor
Network Contributor
Storage Account Contributor
Instead of assigning roles to individual users, it's often easier to use an Azure Active Directory group that has users
who need to take similar actions. Then, assign that group to the appropriate role. For this article, either use an
existing group for managing the virtual machine, or use the portal to create an Azure Active Directory group.
After creating a new group or finding an existing one, use the New -AzureRmRoleAssignment command to assign
the Azure Active Directory group to the Virtual Machine Contributor role for the resource group.
$adgroup = Get-AzureRmADGroup -DisplayName <your-group-name>
New-AzureRmRoleAssignment -ObjectId $adgroup.id `

-ResourceGroupName myResourceGroup `
-RoleDefinitionName "Virtual Machine Contributor"
If you receive an error stating Principal does not exist in the directory, the new group hasn't propagated
throughout Azure Active Directory. Try running the command again.
Typically, you repeat the process for Network Contributor and Storage Account Contributor to make sure users
are assigned to manage the deployed resources. In this article, you can skip those steps.
Azure Policy
Azure Policy helps you make sure all resources in subscription meet corporate standards. Your subscription already
has several policy definitions. To see the available policy definitions, use the Get-AzureRmPolicyDefinition
command:
(Get-AzureRmPolicyDefinition).Properties | Format-Table displayName, policyType
You see the existing policy definitions. The policy type is either BuiltIn or Custom. Look through the definitions
for ones that describe a condition you want assign. In this article, you assign policies that:
Limit the locations for all resources.
Limit the SKUs for virtual machines.
Audit virtual machines that don't use managed disks.
In the following example, you retrieve three policy definitions based on the display name. You use the New -
AzureRMPolicyAssignment command to assign those definitions to the resource group. For some policies, you
provide parameter values to specify the allowed values.
# Values to use for parameters
$locations ="eastus", "eastus2"
$skus = "Standard_DS1_v2", "Standard_E2s_v2"
# Get the resource group

$rg = Get-AzureRmResourceGroup -Name myResourceGroup
# Get policy definitions for allowed locations, allowed SKUs, and auditing VMs that don't use managed disks
$locationDefinition = Get-AzureRmPolicyDefinition | where-object {$_.properties.displayname -eq "Allowed
locations"}
$skuDefinition = Get-AzureRmPolicyDefinition | where-object {$_.properties.displayname -eq "Allowed virtual
machine SKUs"}
$auditDefinition = Get-AzureRmPolicyDefinition | where-object {$_.properties.displayname -eq "Audit VMs that
do not use managed disks"}
# Assign policy for allowed locations

New-AzureRMPolicyAssignment -Name "Set permitted locations" `
-Scope $rg.ResourceId `
-PolicyDefinition $locationDefinition `
-listOfAllowedLocations $locations
# Assign policy for allowed SKUs

New-AzureRMPolicyAssignment -Name "Set permitted VM SKUs" `
-PolicyDefinition $skuDefinition `
-listOfAllowedSKUs $skus
# Assign policy for auditing unmanaged disks

New-AzureRMPolicyAssignment -Name "Audit unmanaged disks" `
-PolicyDefinition $auditDefinition
Deploy the virtual machine

You have assigned roles and policies, so you're ready to deploy your solution. The default size is Standard_DS1_v2,
which is one of your allowed SKUs. When running this step, you're prompted for credentials. The values that you
enter are configured as the user name and password for the virtual machine.
New-AzureRmVm -ResourceGroupName "myResourceGroup" `

-Name "myVM" `
-OpenPorts 80,3389
After your deployment finishes, you can apply more management settings to the solution.
Lock resources
Resource locks prevent users in your organization from accidentally deleting or modifying critical resources. Unlike
role-based access control, resource locks apply a restriction across all users and roles. You can set the lock level to
CanNotDelete or ReadOnly.
To lock the virtual machine and network security group, use the New -AzureRmResourceLock command:
# Add CanNotDelete lock to the VM
New-AzureRmResourceLock -LockLevel CanNotDelete `
-LockName LockVM `
-ResourceName myVM `
-ResourceType Microsoft.Compute/virtualMachines `
# Add CanNotDelete lock to the network security group

New-AzureRmResourceLock -LockLevel CanNotDelete `
-LockName LockNSG `
-ResourceName myNetworkSecurityGroup `
-ResourceType Microsoft.Network/networkSecurityGroups `
To test the locks, try running the following command:
You see an error stating that the delete operation can't be completed because of a lock. The resource group can
only be deleted if you specifically remove the locks. That step is shown in Clean up resources.
Tag resources
You apply tags to your Azure resources to logically organize them by categories. Each tag consists of a name and a
value. For example, you can apply the name "Environment" and the value "Production" to all the resources in
production.
To add two tags to a resource group, use the Set-AzureRmResourceGroup command:
Set-AzureRmResourceGroup -Name myResourceGroup -Tag @{ Dept="IT"; Environment="Test" }
Let's suppose you want to add a third tag. Every time you apply tags to a resource or a resource group, you
overwrite the existing tags on that resource or resource group. To add a new tag without losing the existing tags,
you must retrieve the existing tags, add a new tag, and reapply the collection of tags:
# Get existing tags and add a new tag

$tags = (Get-AzureRmResourceGroup -Name myResourceGroup).Tags
$tags.Add("Project", "Documentation")
# Reapply the updated set of tags

Set-AzureRmResourceGroup -Tag $tags -Name myResourceGroup
Resources don't inherit tags from the resource group. Currently, your resource group has three tags but the
resources do not have any tags. To apply all tags from a resource group to its resources, and retain existing tags on
resources that are not duplicates, use the following script:
$group = Get-AzureRmResourceGroup myResourceGroup
if ($group.Tags -ne $null) {

# Get the resources in the resource group
$resources = Get-AzureRmResource -ResourceGroupName $group.ResourceGroupName
# Loop through each resource

foreach ($r in $resources)
{
# Get the tags for this resource
$resourcetags = (Get-AzureRmResource -ResourceId $r.ResourceId).Tags
# If the resource has existing tags, add new ones

if ($resourcetags)
{
foreach ($key in $group.Tags.Keys)
{
if (-not($resourcetags.ContainsKey($key)))
{
$resourcetags.Add($key, $group.Tags[$key])
}
}
# Reapply the updated tags to the resource

Set-AzureRmResource -Tag $resourcetags -ResourceId $r.ResourceId -Force
}
else
{
Set-AzureRmResource -Tag $group.Tags -ResourceId $r.ResourceId -Force
}
}
}
Alternatively, you can apply tags from the resource group to the resources without keeping the existing tags:

$g = Get-AzureRmResourceGroup -Name myResourceGroup
# Find all the resources in the resource group, and for each resource apply the tags from the resource group
Get-AzureRmResource -ResourceGroupName $g.ResourceGroupName | ForEach-Object {Set-AzureRmResource -ResourceId
$_.ResourceId -Tag $g.Tags -Force }
To combine several values in a single tag, use a JSON string.
Set-AzureRmResourceGroup -Name myResourceGroup -Tag @{ CostCenter="{`"Dept`":`"IT`",`"Environment`":`"Test`"}"

}
To add a new tag with several values without losing the existing tags, you must retrieve the existing tags, use a
JSON string for the new tag, and reapply the collection of tags:
# Get existing tags and add a new tag

$ResourceGroup = Get-AzureRmResourceGroup -Name myResourceGroup
$Tags = $ResourceGroup.Tags
$Tags.Add("CostCenter", "{`"Dept`":`"IT`",`"Environment`":`"Test`"}")
# Reapply the updated set of tags

$ResourceGroup | Set-AzureRmResourceGroup -Tag $Tags
To remove all tags, you pass an empty hash table.

Set-AzureRmResourceGroup -Name myResourceGroup -Tag @{ }
To apply tags to a virtual machine, use the Set-AzureRmResource command:
# Get the virtual machine

$r = Get-AzureRmResource -ResourceName myVM `
-ResourceType Microsoft.Compute/virtualMachines
# Apply tags to the virtual machine

Set-AzureRmResource -Tag @{ Dept="IT"; Environment="Test"; Project="Documentation" } -ResourceId $r.ResourceId
-Force
Find resources by tag

To find resources with a tag name and value, use the Get-AzureRmResource command:
(Get-AzureRmResource -Tag @{ Environment="Test"}).Name
You can use the returned values for management tasks like stopping all virtual machines with a tag value.
Get-AzureRmResource -Tag @{ Environment="Test"} | Where-Object {$_.ResourceType -eq

"Microsoft.Compute/virtualMachines"} | Stop-AzureRmVM
View costs by tag values

After applying tags to resources, you can view costs for resources with those tags. It takes a while for cost analysis
to show the latest usage, so you may not see the costs yet. When the costs are available, you can view costs for
resources across resource groups in your subscription. Users must have subscription level access to billing
information to see the costs.
To view costs by tag in the portal, select your subscription and select Cost Analysis.
Then, filter by the tag value, and select Apply.

You can also use the Azure Billing APIs to programmatically view costs.
Clean up resources
The locked network security group can't be deleted until the lock is removed. To remove the lock, use the Remove-
AzureRmResourceLock command:
Remove-AzureRmResourceLock -LockName LockVM `

-ResourceName myVM `
-ResourceType Microsoft.Compute/virtualMachines `
Remove-AzureRmResourceLock -LockName LockNSG `
-ResourceName myNetworkSecurityGroup `
-ResourceType Microsoft.Network/networkSecurityGroups `
When no longer needed, you can use the Remove-AzureRmResourceGroup command to remove the resource
group, VM, and all related resources.
Next steps
In this tutorial, you created a custom VM image. You learned how to:
Assign users to a role
Apply policies that enforce standards
Protect critical resources with locks
Tag resources for billing and management
Advance to the next tutorial to learn about how highly available virtual machines.
Monitor virtual machines
Tutorial: Monitor and update a Windows virtual
machine in Azure
Azure monitoring uses agents to collect boot and performance data from Azure VMs, store this data in Azure
storage, and make it accessible through portal, the Azure PowerShell module, and the Azure CLI. Update
management allows you to manage updates and patches for your Azure Windows VMs.
In this tutorial, you learn how to:
Enable boot diagnostics on a VM
View boot diagnostics
View VM host metrics
Install the diagnostics extension
View VM metrics
Create an alert
Manage Windows updates
Monitor changes and inventory
Set up advanced monitoring
This tutorial requires the Azure PowerShell module version 5.7.0 or later. Run Get-Module -ListAvailable AzureRM
to find the version. If you need to upgrade, see Install Azure PowerShell module.

To configure Azure monitoring and update management in this tutorial, you need a Windows VM in Azure. First,
set an administrator username and password for the VM with Get-Credential:
Now create the VM with New -AzureRmVM. The following example creates a VM named myVM in the EastUS
location. If they do not already exist, the resource group myResourceGroupMonitorMonitor and supporting
network resources are created:
New-AzureRmVm `
-ResourceGroupName "myResourceGroupMonitor" `
-Name "myVM" `
-Credential $cred
It takes a few minutes for the resources and VM to be created.

As Windows virtual machines boot up, the boot diagnostic agent captures screen output that can be used for
troubleshooting purpose. This capability is enabled by default. The captured screen shots are stored in an Azure
storage account, which is also created by default.
You can get the boot diagnostic data with the Get-AzureRmVMBootDiagnosticsData command. In the following
example, boot diagnostics are downloaded to the root of the *c:* drive.
Get-AzureRmVMBootDiagnosticsData -ResourceGroupName "myResourceGroupMonitor" -Name "myVM" -Windows -LocalPath

"c:\"
View host metrics

A Windows VM has a dedicated Host VM in Azure that it interacts with. Metrics are automatically collected for the
Host and can be viewed in the Azure portal.
1. In the Azure portal, click Resource Groups, select myResourceGroupMonitor, and then select myVM in the
resource list.
2. Click Metrics on the VM blade, and then select any of the Host metrics under Available metrics to see
how the Host VM is performing.
Install diagnostics extension

The basic host metrics are available, but to see more granular and VM -specific metrics, you to need to install the
Azure diagnostics extension on the VM. The Azure diagnostics extension allows additional monitoring and
diagnostics data to be retrieved from the VM. You can view these performance metrics and create alerts based on
how the VM performs. The diagnostic extension is installed through the Azure portal as follows:
resource list.
2. Click Diagnosis settings. The list shows that Boot diagnostics are already enabled from the previous section.
Click the check box for Basic metrics.
3. Click the Enable guest-level monitoring button.
View VM metrics
You can view the VM metrics in the same way that you viewed the host VM metrics:
resource list.
2. To see how the VM is performing, click Metrics on the VM blade, and then select any of the diagnostics
metrics under Available metrics.
Create alerts
You can create alerts based on specific performance metrics. Alerts can be used to notify you when average CPU
usage exceeds a certain threshold or available free disk space drops below a certain amount, for example. Alerts
are displayed in the Azure portal or can be sent via email. You can also trigger Azure Automation runbooks or
Azure Logic Apps in response to alerts being generated.
The following example creates an alert for average CPU usage.
resource list.
2. Click Alert rules on the VM blade, then click Add metric alert across the top of the alerts blade.
3. Provide a Name for your alert, such as myAlertRule
4. To trigger an alert when CPU percentage exceeds 1.0 for five minutes, leave all the other defaults selected.
5. Optionally, check the box for Email owners, contributors, and readers to send email notification. The default
action is to present a notification in the portal.
6. Click the OK button.

Update management allows you to manage updates and patches for your Azure Windows VMs. Directly from
your VM, you can quickly assess the status of available updates, schedule installation of required updates, and
review deployment results to verify updates were applied successfully to the VM.
For pricing information, see Automation pricing for Update management
Enable Update management
Enable Update management for your VM:
1. On the left-hand side of the screen, select Virtual machines.
2. From the list, select a VM.
3. On the VM screen, in the Operations section, click Update management. The Enable Update
Management screen opens.
Validation is performed to determine if Update management is enabled for this VM. The validation includes checks
for a Log Analytics workspace and linked Automation account, and if the solution is in the workspace.
A Log Analytics workspace is used to collect data that is generated by features and services such as Update
management. The workspace provides a single location to review and analyze data from multiple sources. To
perform additional actions on VMs that require updates, Azure Automation allows you to run runbooks against
VMs, such as download and apply updates.
The validation process also checks to see if the VM is provisioned with the Microsoft Monitoring Agent (MMA)
and Automation hybrid runbook worker. This agent is used to communicate with the VM and obtain information
about the update status.
Choose the Log analytics workspace and automation account and click Enable to enable the solution. The solution
takes up to 15 minutes to enable.
If any of the following prerequisites were found to be missing during onboarding, they're automatically added:
Log Analytics workspace
Automation
A Hybrid runbook worker is enabled on the VM
The Update Management screen opens. Configure the location, Log analytics workspace and Automation
account to use and click Enable. If the fields are grayed out, that means another automation solution is enabled for
the VM and the same workspace and Automation account must be used.
Enabling the solution can take up to 15 minutes. During this time, you shouldn't close the browser window. After
the solution is enabled, information about missing updates on the VM flows to Log Analytics. It can take between
30 minutes and 6 hours for the data to be available for analysis.
View update assessment
After Update management is enabled, the Update management screen appears. After the evaluation of
updates is complete, you see a list of missing updates on the Missing updates tab.
Schedule an update deployment

To install updates, schedule a deployment that follows your release schedule and service window. You can choose
which update types to include in the deployment. For example, you can include critical or security updates and
exclude update rollups.
Schedule a new Update Deployment for the VM by clicking Schedule update deployment at the top of the
Update management screen. In the New update deployment screen, specify the following information:
Name - Provide a unique name to identify the update deployment.
Update classification - Select the types of software the update deployment included in the deployment.
The classification types are:
Critical updates
Security updates
Update rollups
Feature packs
Service packs
Definition updates
Tools
Updates
Schedule settings - You can either accept the default date and time, which is 30 minutes after current time,
or specify a different time. You can also specify whether the deployment occurs once or set up a recurring
schedule. Click the Recurring option under Recurrence to set up a recurring schedule.
Maintenance window (minutes) - Specify the period of time you want the update deployment to occur
within. This helps ensure changes are performed within your defined service windows.
After you have completed configuring the schedule, click Create button and you return to the status dashboard.
Notice that the Scheduled table shows the deployment schedule you created.
WARNING
For updates that require a reboot, the VM is restarted automatically.
View results of an update deployment

After the scheduled deployment starts, you can see the status for that deployment on the Update deployments
tab on the Update management screen. If it is currently running, it's status shows as In progress. After it
completes, if successful, it changes to Succeeded. If there is a failure with one or more updates in the deployment,
the status is Partially failed. Click the completed update deployment to see the dashboard for that update
deployment.
In Update results tile is a summary of the total number of updates and deployment results on the VM. In the
table to the right is a detailed breakdown of each update and the installation results, which could be one of the
following values:
Not attempted - the update was not installed because there was insufficient time available based on the
maintenance window duration defined.
Succeeded - the update succeeded
Failed - the update failed
Click All logs to see all log entries that the deployment created.
Click the Output tile to see job stream of the runbook responsible for managing the update deployment on the
target VM.
Click Errors to see detailed information about any errors from the deployment.

You can collect and view inventory for software, files, Linux daemons, Windows Services, and Windows Registry
keys on your computers. Tracking the configurations of your machines can help you pinpoint operational issues
across your environment and better understand the state of your machines.
Enable Change and Inventory management
Enable Change and Inventory management for your VM:
1. On the left-hand side of the screen, select Virtual machines.
2. From the list, select a VM.
3. On the VM screen, in the Operations section, click Inventory or Change tracking. The Enable Change
Tracking and Inventory screen opens.
Configure the location, Log analytics workspace and Automation account to use and click Enable. If the fields are
grayed out, that means another automation solution is enabled for the VM and the same workspace and
Automation account must be used. Eventhough the solutions are separate on the menu, they are the same
solution. Enabling one enables both for your VM.
After the solution has been enabled it may take some time while inventory is being collected on the VM before
data appears.
Track changes
On your VM select Change Tracking under OPERATIONS. Click Edit Settings, the Change Tracking page is
displayed. Select the type of setting you want to track and then click + Add to configure the settings. The available
options for Windows are:
Windows Registry
Windows Files
For detailed information on Change Tracking see, Troubleshoot changes on a VM
View inventory
On your VM select Inventory under OPERATIONS. On the Software tab, there is a table list the software that
had been found. The high-level details for each software record are viewable in the table. These details include the
software name, version, publisher, last refreshed time.
Monitor Activity logs and changes

From the Change tracking page on your VM, select Manage Activity Log Connection. This task opens the
Azure Activity log page. Select Connect to connect Change tracking to the Azure activity log for your VM.
With this setting enabled, navigate to the Overview page for your VM and select Stop to stop your VM. When
prompted, select Yes to stop the VM. When it is deallocated, select Start to restart your VM.
Stopping and starting a VM logs an event in its activity log. Navigate back to the Change tracking page. Select
the Events tab at the bottom of the page. After a while, the events shown in the chart and the table. Each event can
be selected to view detailed information on the event.
The chart shows changes that have occurred over time. After you have added an Activity Log connection, the line
graph at the top displays Azure Activity Log events. Each row of bar graphs represents a different trackable
Change type. These types are Linux daemons, files, Windows Registry keys, software, and Windows services. The
change tab shows the details for the changes shown in the visualization in descending order of time that the
change occurred (most recent first).
Advanced monitoring
You can do more advanced monitoring of your VM by using the solutions like Update Management and Change
and Inventory provided by Azure Automation.
When you have access to the Log Analytics workspace, you can find the workspace key and workspace identifier
on by selecting Advanced settings under SETTINGS. Use the Set-AzureRmVMExtension command to add the
Microsoft Monitoring agent extension to the VM. Update the variable values in the below sample to reflect you
Log Analytics workspace key and workspace Id.
$workspaceId = "<Replace with your workspace Id>"

$key = "<Replace with your primary key>"
Set-AzureRmVMExtension -ResourceGroupName "myResourceGroupMonitor" `

-ExtensionName "Microsoft.EnterpriseCloud.Monitoring" `
-VMName "myVM" `
-Publisher "Microsoft.EnterpriseCloud.Monitoring" `
-ExtensionType "MicrosoftMonitoringAgent" `
-Settings @{"workspaceId" = $workspaceId} `
-ProtectedSettings @{"workspaceKey" = $key} `
-Location "East US"
After a few minutes, you should see the new VM in the Log Analytics workspace.
Next steps
In this tutorial, you configured and reviewed VMs with Azure Security Center. You learned how to:
Create a virtual network
Create a resource group and VM
Enable boot diagnostics on the VM
View host metrics
Install the diagnostics extension
View VM metrics
Create an alert
Set up advanced monitoring
Advance to the next tutorial to learn about Azure security center.
Manage VM security
Tutorial: Use Azure Security Center to monitor
Windows virtual machines
Azure Security Center can help you gain visibility into your Azure resource security practices. Security Center
offers integrated security monitoring. It can detect threats that otherwise might go unnoticed. In this tutorial, you
learn about Azure Security Center, and how to:
Set up data collection
Set up security policies
View and fix configuration health issues
Review detected threats
Security Center overview

Security Center identifies potential virtual machine (VM ) configuration issues and targeted security threats. These
might include VMs that are missing network security groups, unencrypted disks, and brute-force Remote Desktop
Protocol (RDP ) attacks. The information is shown on the Security Center dashboard in easy-to-read graphs.
To access the Security Center dashboard, in the Azure portal, on the menu, select Security Center. On the
dashboard, you can see the security health of your Azure environment, find a count of current recommendations,
and view the current state of threat alerts. You can expand each high-level chart to see more detail.
Security Center goes beyond data discovery to provide recommendations for issues that it detects. For example, if
a VM was deployed without an attached network security group, Security Center displays a recommendation, with
remediation steps you can take. You get automated remediation without leaving the context of Security Center.

Before you can get visibility into VM security configurations, you need to set up Security Center data collection.
This involves turning on data collection which automatically installs the Microsoft Monitoring Agent on all the
VMs in your subscription.
1. On the Security Center dashboard, click Security policy, and then select your subscription.
2. For Data collection, in Auto Provisioning select On.
3. For Default workspace configuration leave it as Use workspace(s) created by Security Center (default).
4. Under Security Events keep the default option of Common.
5. Click Save at the top of the page.
The Security Center data collection agent is then installed on all VMs, and data collection begins.
Set up a security policy

Security policies are used to define the items for which Security Center collects data and makes recommendations.
You can apply different security policies to different sets of Azure resources. Although by default Azure resources
are evaluated against all policy items, you can turn off individual policy items for all Azure resources or for a
resource group. For in-depth information about Security Center security policies, see Set security policies in Azure
Security Center.
To set up a security policy for an entire subscription:
1. On the Security Center dashboard, select Security policy and then select your subscription.
2. On the Security policy blade, select Security policy.
3. On the ** Security policy - Security policy ** blade, turn on or turn off policy items that you want to apply to the
subscription.
4. When you're finished selecting your settings, select Save at the top of the blade.
View VM configuration health
After you've turned on data collection and set a security policy, Security Center begins to provide alerts and
recommendations. As VMs are deployed, the data collection agent is installed. Security Center is then populated
with data for the new VMs. For in-depth information about VM configuration health, see Protect your VMs in
Security Center.
As data is collected, the resource health for each VM and related Azure resource is aggregated. The information is
shown in an easy-to-read chart.
To view resource health:
1. On the Security Center dashboard, under Prevention, select Compute.
2. On the Compute blade, select VMs and computers. This view provides a summary of the configuration status
for all your VMs.
To see all recommendations for a VM, select the VM. Recommendations and remediation are covered in more
detail in the next section of this tutorial.
Remediate configuration issues

After Security Center begins to populate with configuration data, recommendations are made based on the
security policy you set up. For instance, if a VM was set up without an associated network security group, a
recommendation is made to create one.
To see a list of all recommendations:
1. On the Security Center dashboard, select Recommendations.
2. Select a specific recommendation. A list of all resources for which the recommendation applies appears.
3. To apply a recommendation, select the resource.
4. Follow the instructions for remediation steps.
In many cases, Security Center provides actionable steps you can take to address a recommendation without
leaving Security Center. In the following example, Security Center detects a network security group that has an
unrestricted inbound rule. On the recommendation page, you can select the Edit inbound rules button. The UI
that is needed to modify the rule appears.
As recommendations are remediated, they are marked as resolved.
View detected threats

In addition to resource configuration recommendations, Security Center displays threat detection alerts. The
security alerts feature aggregates data collected from each VM, Azure networking logs, and connected partner
solutions to detect security threats against Azure resources. For in-depth information about Security Center threat
detection capabilities, see Azure Security Center detection capabilities.
The security alerts feature requires the Security Center pricing tier to be increased from Free to Standard. A 60-
day free trial is available when you move to this higher pricing tier.
To change the pricing tier:
1. On the Security Center dashboard, click Security policy, and then select your subscription.
2. Select Pricing tier.
3. Select Standard and then click Save at the top of the blade.
After you've changed the pricing tier, the security alerts graph begins to populate as security threats are detected.
Select an alert to view information. For example, you can see a description of the threat, the detection time, all
threat attempts, and the recommended remediation. In the following example, an RDP brute-force attack was
detected, with 294 failed RDP attempts. A recommended resolution is provided.
Next steps
In this tutorial, you set up Azure Security Center, and then reviewed VMs in Security Center. You learned how to:
Set up security policies
View and fix configuration health issues
Review detected threats
Advance to the next tutorial to learn how to create a CI/CD pipeline with Azure DevOps Services and a Windows
VM running IIS.
[Azure Pipelines(./tutorial-vsts-iis-cicd.md)
Tutorial: Create a continuous integration pipeline with
Azure DevOps Services and IIS
To automate the build, test, and deployment phases of application development, you can use a continuous
integration and deployment (CI/CD ) pipeline. In this tutorial, you create a CI/CD pipeline using Azure DevOps
Services and a Windows virtual machine (VM ) in Azure that runs IIS. You learn how to:
Publish an ASP.NET web application to a Azure DevOps Services project
Create a build pipeline that is triggered by code commits
Install and configure IIS on a virtual machine in Azure
Add the IIS instance to a deployment group in Azure DevOps Services
Create a release pipeline to publish new web deploy packages to IIS
Test the CI/CD pipeline
This tutorial requires the Azure PowerShell module version 5.7.0 or later. Run Get-Module -ListAvailable AzureRM
to find the version. If you need to upgrade, see Install Azure PowerShell module.
Create a project in Azure DevOps Services

Azure DevOps Services allows for easy collaboration and development without maintaining an on-premises code
management solution. Azure DevOps Services provides cloud code testing, build, and application insights. You can
choose a version control repo and IDE that best fits your code development. For this tutorial, you can use a free
organization to create a basic ASP.NET web app and CI/CD pipeline. If you do not already have a Azure DevOps
Services organization, create one.
To manage the code commit process, build pipelines, and release pipelines, create a project in Azure DevOps
Services as follows:
1. Open your Azure DevOps Services dashboard in a web browser and choose New project.
2. Enter myWebApp for the Project name. Leave all other default values to use Git version control and Agile
work item process.
3. Choose the option to Share with Team Members, then select Create.
4. Once your project has been created, choose the option to Initialize with a README or gitignore, then
Initialize.
5. Inside your new project, choose Dashboards across the top, then select Open in Visual Studio.
Create ASP.NET web application

In the previous step, you created a project in Azure DevOps Services. The final step opens your new project in
Visual Studio. You manage your code commits in the Team Explorer window. Create a local copy of your new
project, then create an ASP.NET web application from a template as follows:
1. Select Clone to create a local git repo of your Azure DevOps Services project.
2. Under Solutions, select New.
3. Select Web templates, and then choose the ASP.NET Web Application template.
a. Enter a name for your application, such as myWebApp, and uncheck the box for Create directory for
solution.
b. If the option is available, uncheck the box to Add Application Insights to project. Application Insights
requires you to authorize your web application with Azure Application Insights. To keep it simple in this
tutorial, skip this process.
c. Select OK.
4. Choose MVC from the template list.
a. Select Change Authentication, choose No Authentication, then select OK.
b. Select OK to create your solution.
5. In the Team Explorer window, choose Changes.
6. In the commit text box, enter a message such as Initial commit. Choose Commit All and Sync from the
drop-down menu.
Create build pipeline

In Azure DevOps Services, you use a build pipeline to outline how your application should be built. In this tutorial,
we create a basic pipeline that takes our source code, builds the solution, then creates web deploy package we can
use to run the web app on an IIS server.
1. In your Azure DevOps Services project, choose Build & Release across the top, then select Builds.
2. Select + New definition.
3. Choose the ASP.NET (PREVIEW ) template and select Apply.
4. Leave all the default task values. Under Get sources, ensure that the myWebApp repository and master
branch are selected.
5. On the Triggers tab, move the slider for Enable this trigger to Enabled.
6. Save the build pipeline and queue a new build by selecting Save & queue , then Save & queue again. Leave
the defaults and select Queue.
Watch as the build is scheduled on a hosted agent, then begins to build. The output is similar to the following
example:

To provide a platform to run your ASP.NET web app, you need a Windows virtual machine that runs IIS. Azure
DevOps Services uses an agent to interact with the IIS instance as you commit code and builds are triggered.
Create a Windows Server 2016 VM with New -AzureRmVM. The following example creates a VM named myVM in
the East US location. The resource group myResourceGroupAzureDevOpsServices and supporting network
resources are also created. To allow web traffic, TCP port 80 is opened to the VM. When prompted, provide a
username and password to be used as the login credentials for the VM:
# Create user object

$cred = Get-Credential -Message "Enter a username and password for the virtual machine."
# Create a virtual machine

New-AzureRmVM `
-ResourceGroupName "myResourceGroupAzureDevOpsServices" `
-Name "myVM" `
-ImageName "Win2016Datacenter" `
-PublicIpAddressName "myPublicIp" `
-Credential $cred `
-OpenPorts 80
To connect to your VM, obtain the public IP address with Get-AzureRmPublicIpAddress as follows:
Get-AzureRmPublicIpAddress -ResourceGroupName "myResourceGroup" | Select IpAddress
Create a remote desktop session to your VM:
mstsc /v:<publicIpAddress>
On the VM, open an Administrator PowerShell command prompt. Install IIS and required .NET features as
follows:
Install-WindowsFeature Web-Server,Web-Asp-Net45,NET-Framework-Features
Create deployment group

To push out the web deploy package to the IIS server, you define a deployment group in Azure DevOps Services.
This group allows you to specify which servers are the target of new builds as you commit code to Azure DevOps
Services and builds are completed.
1. In Azure DevOps Services, choose Build & Release and then select Deployment groups.
2. Choose Add Deployment group.
3. Enter a name for the group, such as myIIS, then select Create.
4. In the Register machines section, ensure Windows is selected, then check the box to Use a personal access
token in the script for authentication.
5. Select Copy script to clipboard.
Add IIS VM to the deployment group
With the deployment group created, add each IIS instance to the group. Azure DevOps Services generates a script
that downloads and configures an agent on the VM that receives new web deploy packages then applies it to IIS.
1. Back in the Administrator PowerShell session on your VM, paste and run the script copied from Azure
DevOps Services.
2. When prompted to configure tags for the agent, choose Y and enter web.
3. When prompted for the user account, press Return to accept the defaults.
4. Wait for the script to finish with a message Service vstsagent.account.computername started successfully.
5. In the Deployment groups page of the Build & Release menu, open the myIIS deployment group. On the
Machines tab, verify that your VM is listed.
Create release pipeline

To publish your builds, you create a release pipeline in Azure DevOps Services. This pipeline is triggered
automatically by a successful build of your application. You choose the deployment group to push your web deploy
package to, and define the appropriate IIS settings.
1. Choose Build & Release, then select Builds. Choose the build pipeline created in a previous step.
2. Under Recently completed, choose the most recent build, then select Release.
3. Choose Yes to create a release pipeline.
4. Choose the Empty template, then select Next.
5. Verify the project and source build pipeline are populated with your project.
6. Select the Continuous deployment check box, then select Create.
7. Select the drop-down box next to + Add tasks and choose Add a deployment group phase.
8. Choose Add next to IIS Web App Deploy(Preview), then select Close.
9. Select the Run on deployment group parent task.
a. For Deployment Group, select the deployment group you created earlier, such as myIIS.
b. In the Machine tags box, select Add and choose the web tag.
10. Select the Deploy: IIS Web App Deploy task to configure your IIS instance settings as follows:
a. For Website Name, enter Default Web Site.
b. Leave all the other default settings.
11. Choose Save, then select OK twice.
Create a release and publish

You can now push your web deploy package as a new release. This step communicates with the agent on each
instance that is part of the deployment group, pushes the web deploy package, then configures IIS to run the
updated web application.
1. In your release pipeline, select + Release, then choose Create Release.
2. Verify that the latest build is selected in the drop-down list, along with Automated deployment: After release
creation. Select Create.
3. A small banner appears across the top of your release pipeline, such as Release 'Release-1' has been created.
Select the release link.
4. Open the Logs tab to watch the release progress.
5. After the release is complete, open a web browser and enter the public IIP address of your VM. Your
ASP.NET web application is running.
Test the whole CI/CD pipeline
With your web application running on IIS, now try the whole CI/CD pipeline. After you make a change in Visual
Studio and commit your code, a build is triggered which then triggers a release of your updated web deploy
package to IIS:
1. In Visual Studio, open the Solution Explorer window.
2. Navigate to and open myWebApp | Views | Home | Index.cshtml
3. Edit line 6 to read:
<h1>ASP.NET with Azure DevOps Services and CI/CD!</h1>
4. Save the file.

5. Open the Team Explorer window, select the myWebApp project, then choose Changes.
6. Enter a commit message, such as Testing CI/CD pipeline, then choose Commit All and Sync from the drop-
down menu.
7. In Azure DevOps Services workspace, a new build is triggered from the code commit.
Choose Build & Release, then select Builds.
Choose your build pipeline, then select the Queued & running build to watch as the build progresses.
8. Once the build is successful, a new release is triggered.
Choose Build & Release, then Releases to see the web deploy package pushed to your IIS VM.
Select the Refresh icon to update the status. When the Environments column shows a green check mark,
the release has successfully deployed to IIS.
9. To see your changes applied, refresh your IIS website in a browser.
Next steps
In this tutorial, you created an ASP.NET web application in Azure DevOps Services and configured build and
release pipelines to deploy new web deploy packages to IIS on each code commit. You learned how to:
Publish an ASP.NET web application to a Azure DevOps Services project
Create a build pipeline that is triggered by code commits
Install and configure IIS on a virtual machine in Azure
Add the IIS instance to a deployment group in Azure DevOps Services
Create a release pipeline to publish new web deploy packages to IIS
Test the CI/CD pipeline
Advance to the next tutorial to learn how to install a SQL\IIS\.NET stack on a pair of Windows VMs.
SQL\IIS\.NET stack
Tutorial: Install the SQL/IIS/.NET stack in a Windows
VM with Azure PowerShell
In this tutorial, we install a SQL/IIS/.NET stack using Azure PowerShell. This stack consists of two VMs running
Windows Server 2016, one with IIS and .NET and the other with SQL Server.
Create a VM
Install IIS and the .NET Core SDK on the VM
Create a VM running SQL Server
Install the SQL Server extension

If you choose to install and use the PowerShell locally, this tutorial requires the AzureRM.Compute module version
5.7.0 or later. Run Get-Module -ListAvailable AzureRM.Compute to find the version. If you need to upgrade, see Install
Azure PowerShell module.
Create a IIS VM
In this example, we use New -AzureRMVM cmdlet in the PowerShell Cloud Shell to quickly create a Windows
Server 2016 VM and then install IIS and the .NET Framework. The IIS and SQL VMs share a resource group and
virtual network, so we create variables for those names.
$vmName = "IISVM"
$vNetName = "myIISSQLvNet"
$resourceGroup = "myIISSQLGroup"
New-AzureRmVm `
-ResourceGroupName $resourceGroup `
-Name $vmName `
-VirtualNetworkName $vNetName `
-SubnetName "myIISSubnet" `
-PublicIpAddressName "myIISPublicIpAddress" `
-OpenPorts 80,3389
Install IIS and the .NET framework using the custom script extension.
Set-AzureRmVMExtension `
-ExtensionName IIS `
-VMName $vmName `
-SettingString '{"commandToExecute":"powershell Add-WindowsFeature Web-Server,Web-Asp-Net45,NET-Framework-
Features"}' `
-Location EastUS
Create another subnet

Create a second subnet for the SQL VM. Get the vNet using [Get-AzureRmVirtualNetwork]
{/powershell/module/azurerm.network/get-azurermvirtualnetwork}.
$vNet = Get-AzureRmVirtualNetwork `
-Name $vNetName `
-ResourceGroupName $resourceGroup
Create a configuration for the subnet using Add-AzureRmVirtualNetworkSubnetConfig.
Add-AzureRmVirtualNetworkSubnetConfig `
-Name mySQLSubnet `
-VirtualNetwork $vNet `
-ServiceEndpoint Microsoft.Sql
Update the vNet with the new subnet information using Set-AzureRmVirtualNetwork
$vNet | Set-AzureRmVirtualNetwork
Azure SQL VM
Use a pre-configured Azure marketplace image of a SQL server to create the SQL VM. We first create the VM,
then we install the SQL Server Extension on the VM.
New-AzureRmVm `
-Name "mySQLVM" `
-ImageName "MicrosoftSQLServer:SQL2016SP1-WS2016:Enterprise:latest" `
-Location eastus `
-VirtualNetworkName $vNetName `
-SubnetName "mySQLSubnet" `
-PublicIpAddressName "mySQLPublicIpAddress" `
-OpenPorts 3389,1401
Use Set-AzureRmVMSqlServerExtension to add the SQL Server extension to the SQL VM.
Set-AzureRmVMSqlServerExtension `
-VMName mySQLVM `
-Name "SQLExtension" `
-Location "EastUS"
Next steps
In this tutorial, you installed a SQL\IIS\.NET stack using Azure PowerShell. You learned how to:
Create a VM
Install IIS and the .NET Core SDK on the VM
Create a VM running SQL Server
Install the SQL Server extension
Advance to the next tutorial to learn how to secure IIS web server with SSL certificates.
Secure IIS web server with SSL certificates
Tutorial: Secure a web server on a Windows virtual
machine in Azure with SSL certificates stored in Key
Vault
To secure web servers, a Secure Sockets Layer (SSL ) certificate can be used to encrypt web traffic. These SSL
certificates can be stored in Azure Key Vault, and allow secure deployments of certificates to Windows virtual
machines (VMs) in Azure. In this tutorial you learn how to:
Create an Azure Key Vault
Generate or upload a certificate to the Key Vault
Create a VM and install the IIS web server
Inject the certificate into the VM and configure IIS with an SSL binding

Overview
Azure Key Vault safeguards cryptographic keys and secrets, such certificates or passwords. Key Vault helps
streamline the certificate management process and enables you to maintain control of keys that access those
certificates. You can create a self-signed certificate inside Key Vault, or upload an existing, trusted certificate that
you already own.
Rather than using a custom VM image that includes certificates baked-in, you inject certificates into a running VM.
This process ensures that the most up-to-date certificates are installed on a web server during deployment. If you
renew or replace a certificate, you don't also have to create a new custom VM image. The latest certificates are
automatically injected as you create additional VMs. During the whole process, the certificates never leave the
Azure platform or are exposed in a script, command-line history, or template.
Before you can create a Key Vault and certificates, create a resource group with New -AzureRmResourceGroup. The
following example creates a resource group named myResourceGroupSecureWeb in the East US location:
$resourceGroup = "myResourceGroupSecureWeb"
$location = "East US"
New-AzureRmResourceGroup -ResourceGroupName $resourceGroup -Location $location
Next, create a Key Vault with New -AzureRmKeyVault. Each Key Vault requires a unique name, and should be all
lower case. Replace mykeyvault in the following example with your own unique Key Vault name:
$keyvaultName="mykeyvault"
New-AzureRmKeyVault -VaultName $keyvaultName `
-ResourceGroup $resourceGroup `
-Location $location `
-EnabledForDeployment
Generate a certificate and store in Key Vault

For production use, you should import a valid certificate signed by trusted provider with Import-AzureKeyVault
Certificate. For this tutorial, the following example shows how you can generate a self-signed certificate with Add-
AzureKeyVaultCertificate that uses the default certificate policy from New -AzureKeyVaultCertificatePolicy.
$policy = New-AzureKeyVaultCertificatePolicy `
-SubjectName "CN=www.contoso.com" `
-SecretContentType "application/x-pkcs12" `
-IssuerName Self `
-ValidityInMonths 12
Add-AzureKeyVaultCertificate `
-VaultName $keyvaultName `
-Name "mycert" `
-CertificatePolicy $policy
Create a virtual machine

Set an administrator username and password for the VM with Get-Credential:
Now you can create the VM with New -AzureRmVM. The following example creates a VM named myVM in the
EastUS location. If they do not already exist, the supporting network resources are created. To allow secure web
traffic, the cmdlet also opens port 443.
# Create a VM
New-AzureRmVm `
-Name "myVM" `
-Credential $cred `
-OpenPorts 443
# Use the Custom Script Extension to install IIS

Set-AzureRmVMExtension -ResourceGroupName $resourceGroup `
-VMName "myVM" `
-ExtensionType "CustomScriptExtension" `
-SettingString '{"commandToExecute":"powershell Add-WindowsFeature Web-Server -IncludeManagementTools"}'
It takes a few minutes for the VM to be created. The last step uses the Azure Custom Script Extension to install the
IIS web server with Set-AzureRmVmExtension.
Add a certificate to VM from Key Vault

To add the certificate from Key Vault to a VM, obtain the ID of your certificate with Get-AzureKeyVaultSecret. Add
the certificate to the VM with Add-AzureRmVMSecret:
$certURL=(Get-AzureKeyVaultSecret -VaultName $keyvaultName -Name "mycert").id
$vm=Get-AzureRmVM -ResourceGroupName $resourceGroup -Name "myVM"

$vaultId=(Get-AzureRmKeyVault -ResourceGroupName $resourceGroup -VaultName $keyVaultName).ResourceId
$vm = Add-AzureRmVMSecret -VM $vm -SourceVaultId $vaultId -CertificateStore "My" -CertificateUrl $certURL
Update-AzureRmVM -ResourceGroupName $resourceGroup -VM $vm
Configure IIS to use the certificate

Use the Custom Script Extension again with Set-AzureRmVMExtension to update the IIS configuration. This
update applies the certificate injected from Key Vault to IIS and configures the web binding:
$PublicSettings = '{
"fileUris":["https://raw.githubusercontent.com/Azure-Samples/compute-automation-
configurations/master/secure-iis.ps1"],
"commandToExecute":"powershell -ExecutionPolicy Unrestricted -File secure-iis.ps1"
}'
Set-AzureRmVMExtension -ResourceGroupName $resourceGroup `

-VMName "myVM" `
-SettingString $publicSettings
Test the secure web app

Obtain the public IP address of your VM with Get-AzureRmPublicIPAddress. The following example obtains the IP
address for myPublicIP created earlier:
Get-AzureRmPublicIPAddress -ResourceGroupName $resourceGroup -Name "myPublicIPAddress" | select "IpAddress"
Now you can open a web browser and enter https://<myPublicIP> in the address bar. To accept the security
warning if you used a self-signed certificate, select Details and then Go on to the webpage:
Your secured IIS website is then displayed as in the following example:

Next steps
In this tutorial, you secured an IIS web server with an SSL certificate stored in Azure Key Vault. You learned how
to:
Generate or upload a certificate to the Key Vault
Create a VM and install the IIS web server
Inject the certificate into the VM and configure IIS with an SSL binding
Follow this link to see pre-built virtual machine script samples.
Windows virtual machine script samples
Azure Virtual Machine PowerShell samples
The following table provides links to PowerShell script samples that create and manage Windows virtual machines
(VMs).
Quickly create a virtual machine Creates a resource group, a virtual machine, and all related
resources, with a minimum of prompts.
Create a fully configured virtual machine Creates a resource group, a virtual machine, and all related
resources.
Create highly available virtual machines Creates several virtual machines in a highly-available and
load-balanced configuration.
Create a VM and run a configuration script Creates a virtual machine and uses the Azure Custom Script
extension to install IIS.
Create a VM and run a DSC configuration Creates a virtual machine and uses the Azure Desired State
Configuration (DSC) extension to install IIS.
Upload a VHD and create VMs Uploads a local VHD file to Azure, creates an image from the
VHD, and then creates a VM from that image.
Create a VM from a managed OS disk Creates a virtual machine by attaching an existing Managed
Disk as OS disk.
Create a VM from a snapshot Creates a virtual machine from a snapshot by first creating a
managed disk from the snapshot and then attaching the new
managed disk as OS disk.
Manage storage
Create a managed disk from a VHD in the same or a different Creates a managed disk from a specialized VHD as an OS disk,
subscription or from a data VHD as a data disk, in the same or a different
subscription.
Create a managed disk from a snapshot Creates a managed disk from a snapshot.
Copy a managed disk to the same or a different subscription Copies a managed disk to the same or a different subscription
that is in the same region as the parent managed disk.
Export a snapshot as a VHD to a storage account Exports a managed snapshot as a VHD to a storage account
in a different region.
Export the VHD of a managed disk to a storage account Exports the underlying VHD of a managed disk to a storage
account in a different region.
Create a snapshot from a VHD Creates a snapshot from a VHD and then uses that snapshot
to create multiple identical managed disks quickly.
Copy a snapshot to the same or a different subscription Copies snapshot to the same or a different subscription that is
in the same region as the parent snapshot.
Secure virtual machines
Encrypt a VM and its data disks Creates an Azure key vault, an encryption key, and a service
principal, and then encrypts a VM.
Monitor a VM with Log Analytics Creates a virtual machine, installs the Azure Log Analytics
agent, and enrolls the VM in a Log Analytics workspace.
Azure CLI Samples for Windows virtual machines
The following table includes links to bash scripts built using the Azure CLI that deploy Windows virtual machines.
Create a virtual machine Creates a Windows virtual machine with minimal

configuration.
Create a fully configured virtual machine Creates a resource group, virtual machine, and all related
resources.
Create highly available virtual machines Creates several virtual machines in a highly available and load
balanced configuration.
Create a VM and run configuration script Creates a virtual machine and uses the Azure Custom Script
extension to install IIS.
Create a VM and run DSC configuration Creates a virtual machine and uses the Azure Desired State
Configuration (DSC) extension to install IIS.
Network virtual machines
Secure network traffic between virtual machines Creates two virtual machines, all related resources, and an
internal and external network security groups (NSG).
Secure virtual machines
Encrypt a VM and data disks Creates an Azure Key Vault, encryption key, and service
principal, then encrypts a VM.
Monitor a VM with Log Analytics Creates a virtual machine, installs the Log Analytics agent, and
enrolls the VM in a Log Analytics workspace.
Azure Resource Manager overview
The infrastructure for your application is typically made up of many components – maybe a virtual machine,
storage account, and virtual network, or a web app, database, database server, and third-party services. You don't
see these components as separate entities, instead you see them as related and interdependent parts of a single
entity. You want to deploy, manage, and monitor them as a group. Azure Resource Manager enables you to work
with the resources in your solution as a group. You can deploy, update, or delete all the resources for your
solution in a single, coordinated operation. You use a template for deployment and that template can work for
different environments such as testing, staging, and production. Resource Manager provides security, auditing,
and tagging features to help you manage your resources after deployment.
Consistent management layer

Resource Manager provides a consistent management layer to perform tasks through Azure PowerShell, Azure
CLI, Azure portal, REST API, and client SDKs. All capabilities that are available in the Azure portal are also
available through Azure PowerShell, Azure CLI, the Azure REST APIs, and client SDKs. Functionality initially
released through APIs will be represented in the portal within 180 days of initial release.
Choose the tools and APIs that work best for you - they have the same capability and provide consistent results.
The following image shows how all the tools interact with the same Azure Resource Manager API. The API
passes requests to the Resource Manager service, which authenticates and authorizes the requests. Resource
Manager then routes the requests to the appropriate resource providers.
Terminology
If you're new to Azure Resource Manager, there are some terms you might not be familiar with.
resource - A manageable item that is available through Azure. Some common resources are a virtual
machine, storage account, web app, database, and virtual network, but there are many more.
resource group - A container that holds related resources for an Azure solution. The resource group can
include all the resources for the solution, or only those resources that you want to manage as a group. You
decide how you want to allocate resources to resource groups based on what makes the most sense for your
organization. See Resource groups.
resource provider - A service that supplies the resources you can deploy and manage through Resource
Manager. Each resource provider offers operations for working with the resources that are deployed. Some
common resource providers are Microsoft.Compute, which supplies the virtual machine resource,
Microsoft.Storage, which supplies the storage account resource, and Microsoft.Web, which supplies resources
related to web apps. See Resource providers.
Resource Manager template - A JavaScript Object Notation (JSON ) file that defines one or more resources
to deploy to a resource group. It also defines the dependencies between the deployed resources. The template
can be used to deploy the resources consistently and repeatedly. See Template deployment.
declarative syntax - Syntax that lets you state "Here is what I intend to create" without having to write the
sequence of programming commands to create it. The Resource Manager template is an example of
declarative syntax. In the file, you define the properties for the infrastructure to deploy to Azure.
The benefits of using Resource Manager

Resource Manager provides several benefits:
You can deploy, manage, and monitor all the resources for your solution as a group, rather than handling
these resources individually.
You can repeatedly deploy your solution throughout the development lifecycle and have confidence your
resources are deployed in a consistent state.
You can manage your infrastructure through declarative templates rather than scripts.
You can define the dependencies between resources so they're deployed in the correct order.
You can apply access control to all services in your resource group because Role-Based Access Control
(RBAC ) is natively integrated into the management platform.
You can apply tags to resources to logically organize all the resources in your subscription.
You can clarify your organization's billing by viewing costs for a group of resources sharing the same tag.
Guidance
The following suggestions help you take full advantage of Resource Manager when working with your solutions.
1. Define and deploy your infrastructure through the declarative syntax in Resource Manager templates, rather
than through imperative commands.
2. Define all deployment and configuration steps in the template. You should have no manual steps for setting
up your solution.
3. Run imperative commands to manage your resources, such as to start or stop an app or machine.
4. Arrange resources with the same lifecycle in a resource group. Use tags for all other organizing of resources.
For guidance on how enterprises can use Resource Manager to effectively manage subscriptions, see Azure
enterprise scaffold - prescriptive subscription governance.
For recommendations on creating Resource Manager templates that you can use across global Azure, Azure
sovereign clouds, and Azure Stack, see Develop Azure Resource Manager templates for cloud consistency.
Quickstarts and tutorials

Use the following quickstarts and tutorials to learn how to develop resource manager templates:
Quickstarts
TITLE DESCRIPTION
Use the Azure portal Generate a template using the portal, and the process of
editing and deploying the template.
Use Visual Studio Code Use Visual Studio Code to create and edit templates, and
how to use the Azure Cloud shell to deploy templates.
Use Visual Studio Use Visual Studio to create, edit, and deploy templates.
Tutorials
TITLE DESCRIPTION
Utilize template reference Utilize the template reference documentation to develop

templates. In the tutorial, you find the storage account
schema, and use the information to create an encrypted
storage account.
Create multiple instances Create multiple instances of Azure resources. In the

tutorial, you create multiple instances of storage account.
Set resource deployment order Define resource dependencies. In the tutorial, you create
a virtual network, a virtual machine, and the dependent
Azure resources. You learn how the dependencies are
defined.
Use conditions Deploy resources based on some parameter values. In

the tutorial, you define a template to create a new
storage account or use an existing storage account based
on the value of a parameter.
Integrate key vault Retrieve secrets/passwords from Azure Key Vault. In the
tutorial, you create a virtual machine. The virtual machine
administrator password is retrieved from a Key Vault.
Create linked templates Modularize templates, and call other templates from a
template. In the tutorial, you create a virtual network, a
virtual machine, and the dependent resources. The
dependent storage account is defined in a linked
template.
Use safe deployment practices Use Azure Deployment manager.
Resource groups
There are some important factors to consider when defining your resource group:
1. All the resources in your group should share the same lifecycle. You deploy, update, and delete them together.
If one resource, such as a database server, needs to exist on a different deployment cycle it should be in
another resource group.
2. Each resource can only exist in one resource group.
3. You can add or remove a resource to a resource group at any time.
4. You can move a resource from one resource group to another group. For more information, see Move
resources to new resource group or subscription.
5. A resource group can contain resources that reside in different regions.
6. A resource group can be used to scope access control for administrative actions.
7. A resource can interact with resources in other resource groups. This interaction is common when the two
resources are related but don't share the same lifecycle (for example, web apps connecting to a database).
When creating a resource group, you need to provide a location for that resource group. You may be wondering,
"Why does a resource group need a location? And, if the resources can have different locations than the resource
group, why does the resource group location matter at all?" The resource group stores metadata about the
resources. Therefore, when you specify a location for the resource group, you're specifying where that metadata
is stored. For compliance reasons, you may need to ensure that your data is stored in a particular region.
Resource providers
Each resource provider offers a set of resources and operations for working with an Azure service. For example,
if you want to store keys and secrets, you work with the Microsoft.KeyVault resource provider. This resource
provider offers a resource type called vaults for creating the key vault.
The name of a resource type is in the format: {resource-provider}/{resource-type}. For example, the key vault
type is Microsoft.KeyVault/vaults.
Before getting started with deploying your resources, you should gain an understanding of the available resource
providers. Knowing the names of resource providers and resources helps you define resources you want to
deploy to Azure. Also, you need to know the valid locations and API versions for each resource type. For more
information, see Resource providers and types.
Template deployment
With Resource Manager, you can create a template (in JSON format) that defines the infrastructure and
configuration of your Azure solution. By using a template, you can repeatedly deploy your solution throughout
its lifecycle and have confidence your resources are deployed in a consistent state. When you create a solution
from the portal, the solution automatically includes a deployment template. You don't have to create your
template from scratch because you can start with the template for your solution and customize it to meet your
specific needs. For a sample, see Quickstart: Create and deploy Azure Resource Manager templates by using the
Azure portal. You can also retrieve a template for an existing resource group by either exporting the current state
of the resource group, or viewing the template used for a particular deployment. Viewing the exported template
is a helpful way to learn about the template syntax.
To learn about the format of the template and how you construct it, see Quickstart: Create and deploy Azure
Resource Manager templates by using the Azure portal. To view the JSON syntax for resources types, see Define
resources in Azure Resource Manager templates.
Resource Manager processes the template like any other request (see the image for Consistent management
layer). It parses the template and converts its syntax into REST API operations for the appropriate resource
providers. For example, when Resource Manager receives a template with the following resource definition:
"resources": [
{
"apiVersion": "2016-01-01",
"type": "Microsoft.Storage/storageAccounts",
"name": "mystorageaccount",
"location": "westus",
"sku": {
"name": "Standard_LRS"
},
"kind": "Storage",
"properties": {
}
}
]
It converts the definition to the following REST API operation, which is sent to the Microsoft.Storage resource
provider:
PUT
https://management.azure.com/subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Micr
osoft.Storage/storageAccounts/mystorageaccount?api-version=2016-01-01
REQUEST BODY
{
"properties": {
}
"sku": {
},
"kind": "Storage"
}
How you define templates and resource groups is entirely up to you and how you want to manage your solution.
For example, you can deploy your three tier application through a single template to a single resource group.
But, you don't have to define your entire infrastructure in a single template. Often, it makes sense to divide your
deployment requirements into a set of targeted, purpose-specific templates. You can easily reuse these templates
for different solutions. To deploy a particular solution, you create a master template that links all the required
templates. The following image shows how to deploy a three tier solution through a parent template that
includes three nested templates.
If you envision your tiers having separate lifecycles, you can deploy your three tiers to separate resource groups.
Notice the resources can still be linked to resources in other resource groups.
For information about nested templates, see Using linked templates with Azure Resource Manager.
Azure Resource Manager analyzes dependencies to ensure resources are created in the correct order. If one
resource relies on a value from another resource (such as a virtual machine needing a storage account for disks),
you set a dependency. For more information, see Defining dependencies in Azure Resource Manager templates.
You can also use the template for updates to the infrastructure. For example, you can add a resource to your
solution and add configuration rules for the resources that are already deployed. If the template specifies creating
a resource but that resource already exists, Azure Resource Manager performs an update instead of creating a
new asset. Azure Resource Manager updates the existing asset to the same state as it would be as new.
Resource Manager provides extensions for scenarios when you need additional operations such as installing
particular software that isn't included in the setup. If you're already using a configuration management service,
like DSC, Chef or Puppet, you can continue working with that service by using extensions. For information about
virtual machine extensions, see About virtual machine extensions and features.
Finally, the template becomes part of the source code for your app. You can check it in to your source code
repository and update it as your app evolves. You can edit the template through Visual Studio.
After defining your template, you're ready to deploy the resources to Azure. For the commands to deploy the
resources, see:
Deploy resources with Resource Manager templates and Azure PowerShell
Deploy resources with Resource Manager templates and Azure CLI
Deploy resources with Resource Manager templates and Azure portal
Deploy resources with Resource Manager templates and Resource Manager REST API
Safe deployment practices

When deploying a complex service to Azure, you might need to deploy your service to multiple regions, and
check its health before proceeding to the next step. Use Azure Deployment Manager to coordinate a staged
rollout of the service. By staging the rollout of your service, you can find potential problems before it has been
deployed to all regions. If you don't need these precautions, the deployment operations in the preceding section
are the better option.
Deployment Manager is currently in private preview.
Tags
Resource Manager provides a tagging feature that enables you to categorize resources according to your
requirements for managing or billing. Use tags when you have a complex collection of resource groups and
resources, and need to visualize those assets in the way that makes the most sense to you. For example, you
could tag resources that serve a similar role in your organization or belong to the same department. Without
tags, users in your organization can create multiple resources that may be difficult to later identify and manage.
For example, you may wish to delete all the resources for a particular project. If those resources aren't tagged for
the project, you have to manually find them. Tagging can be an important way for you to reduce unnecessary
costs in your subscription.
Resources do not need to reside in the same resource group to share a tag. You can create your own tag
taxonomy to ensure that all users in your organization use common tags rather than users inadvertently applying
slightly different tags (such as "dept" instead of "department").
The following example shows a tag applied to a virtual machine.
"resources": [
{
"type": "Microsoft.Compute/virtualMachines",
"apiVersion": "2015-06-15",
"name": "SimpleWindowsVM",
"location": "[resourceGroup().location]",
"tags": {
"costCenter": "Finance"
},
...
}
]
The usage report for your subscription includes tag names and values, which enables you to break out costs by
tags. For more information about tags, see Using tags to organize your Azure resources.
Access control
Resource Manager enables you to control who has access to specific actions for your organization. It natively
integrates role-based access control (RBAC ) into the management platform and applies that access control to all
services in your resource group.
There are two main concepts to understand when working with role-based access control:
Role definitions - describe a set of permissions and can be used in many assignments.
Role assignments - associate a definition with an identity (user or group) for a particular scope (subscription,
resource group, or resource). The assignment is inherited by lower scopes.
You can add users to pre-defined platform and resource-specific roles. For example, you can take advantage of
the pre-defined role called Reader that permits users to view resources but not change them. You add users in
your organization that need this type of access to the Reader role and apply the role to the subscription, resource
group, or resource.
Azure provides the following four platform roles:
1. Owner - can manage everything, including access
2. Contributor - can manage everything except access
3. Reader - can view everything, but can't make changes
4. User Access Administrator - can manage user access to Azure resources
Azure also provides several resource-specific roles. Some common ones are:
1. Virtual Machine Contributor - can manage virtual machines but not grant access to them, and can't manage
the virtual network or storage account to which they're connected
2. Network Contributor - can manage all network resources, but not grant access to them
3. Storage Account Contributor - Can manage storage accounts, but not grant access to them
4. SQL Server Contributor - Can manage SQL servers and databases, but not their security-related policies
5. Website Contributor - Can manage websites, but not the web plans to which they're connected
For the full list of roles and permitted actions, see RBAC: Built in Roles. For more information about role-based
access control, see Azure Role-based Access Control.
In some cases, you want to run code or script that accesses resources, but you don't want to run it under a user’s
credentials. Instead, you want to create an identity called a service principal for the application and assign the
appropriate role for the service principal. Resource Manager enables you to create credentials for the application
and programmatically authenticate the application. To learn about creating service principals, see one of
following topics:
Use Azure PowerShell to create a service principal to access resources
Use Azure CLI to create a service principal to access resources
Use portal to create Azure Active Directory application and service principal that can access resources
You can also explicitly lock critical resources to prevent users from deleting or modifying them. For more
information, see Lock resources with Azure Resource Manager.
Customized policies
Resource Manager enables you to create customized policies for managing your resources. The types of policies
you create can include diverse scenarios. You can enforce a naming convention on resources, limit which types
and instances of resources can be deployed, or limit which regions can host a type of resource. You can require a
tag value on resources to organize billing by departments. You create policies to help reduce costs and maintain
consistency in your subscription.
There are many more types of policies you can create. For more information, see What is Azure Policy?.
SDKs
Azure SDKs are available for multiple languages and platforms. Each of these language implementations is
available through its ecosystem package manager and GitHub.
Here are the Open Source SDK repositories.
Azure SDK for .NET
Azure Management Libraries for Java
Azure SDK for Node.js
Azure SDK for PHP
Azure SDK for Python
Azure SDK for Ruby
For information about using these languages with your resources, see:
Azure for .NET developers
Azure for Java developers
Azure for Node.js developers
Azure for Python developers
NOTE
If the SDK doesn't provide the required functionality, you can also call to the Azure REST API directly.
Next steps
In this article, you learned how to use Azure Resource Manager for deployment, management, and access control
of resources on Azure. Proceed to the next article to learn how to create your first Azure Resource Manager
template.
Quickstart: Create and deploy Azure Resource Manager templates by using the Azure portal
Regions and availability for virtual machines in Azure
Azure operates in multiple datacenters around the world. These datacenters are grouped in to geographic regions,
giving you flexibility in choosing where to build your applications. It is important to understand how and where
your virtual machines (VMs) operate in Azure, along with your options to maximize performance, availability, and
redundancy. This article provides you with an overview of the availability and redundancy features of Azure.
What are Azure regions?

You create Azure resources in defined geographic regions like 'West US', 'North Europe', or 'Southeast Asia'. You
can review the list of regions and their locations. Within each region, multiple datacenters exist to provide for
redundancy and availability. This approach gives you flexibility as you design applications to create VMs closest to
your users and to meet any legal, compliance, or tax purposes.
Special Azure regions

Azure has some special regions that you may wish to use when building out your applications for compliance or
legal purposes. These special regions include:
US Gov Virginia and US Gov Iowa
A physical and logical network-isolated instance of Azure for US government agencies and partners,
operated by screened US persons. Includes additional compliance certifications such as FedRAMP and
DISA. Read more about Azure Government.
China East and China North
These regions are available through a unique partnership between Microsoft and 21Vianet, whereby
Microsoft does not directly maintain the datacenters. See more about Microsoft Azure in China.
Germany Central and Germany Northeast
These regions are available via a data trustee model whereby customer data remains in Germany under
control of T-Systems, a Deutsche Telekom company, acting as the German data trustee.
Region pairs
Each Azure region is paired with another region within the same geography (such as US, Europe, or Asia). This
approach allows for the replication of resources, such as VM storage, across a geography that should reduce the
likelihood of natural disasters, civil unrest, power outages, or physical network outages affecting both regions at
once. Additional advantages of region pairs include:
In the event of a wider Azure outage, one region is prioritized out of every pair to help reduce the time to
restore for applications.
Planned Azure updates are rolled out to paired regions one at a time to minimize downtime and risk of
application outage.
Data continues to reside within the same geography as its pair (except for Brazil South) for tax and law
enforcement jurisdiction purposes.
Examples of region pairs include:
PRIMARY SECONDARY
West US East US
North Europe West Europe
Southeast Asia East Asia
You can see the full list of regional pairs here.
Feature availability
Some services or VM features are only available in certain regions, such as specific VM sizes or storage types.
There are also some global Azure services that do not require you to select a particular region, such as Azure
Active Directory, Traffic Manager, or Azure DNS. To assist you in designing your application environment, you can
check the availability of Azure services across each region. You can also programmatically query the supported VM
sizes and restrictions in each region.
Storage availability
Understanding Azure regions and geographies becomes important when you consider the available storage
replication options. Depending on the storage type, you have different replication options.
Azure Managed Disks
Locally redundant storage (LRS )
Replicates your data three times within the region in which you created your storage account.
Storage account-based disks
Locally redundant storage (LRS )
Replicates your data three times within the region in which you created your storage account.
Zone redundant storage (ZRS )
Replicates your data three times across two to three facilities, either within a single region or across two
regions.
Geo-redundant storage (GRS )
Replicates your data to a secondary region that is hundreds of miles away from the primary region.
Read-access geo-redundant storage (RA-GRS )
Replicates your data to a secondary region, as with GRS, but also then provides read-only access to the
data in the secondary location.
The following table provides a quick overview of the differences between the storage replication types:
REPLICATION
STRATEGY LRS ZRS GRS RA-GRS
Data is replicated No Yes Yes Yes

across multiple
facilities.
Data can be read No No No Yes

from the secondary
location and from the
primary location.
REPLICATION
STRATEGY LRS ZRS GRS RA-GRS
Number of copies of 3 3 6 6
data maintained on
separate nodes.
You can read more about Azure Storage replication options here. For more information about managed disks, see
Azure Managed Disks overview.
Storage costs
Prices vary depending on the storage type and availability that you select.
Azure Managed Disks
Premium Managed Disks are backed by Solid-State Drives (SSDs) and Standard Managed Disks are backed by
regular spinning disks. Both Premium and Standard Managed Disks are charged based on the provisioned
capacity for the disk.
Unmanaged disks
Premium storage is backed by Solid-State Drives (SSDs) and is charged based on the capacity of the disk.
Standard storage is backed by regular spinning disks and is charged based on the in-use capacity and desired
storage availability.
For RA-GRS, there is an additional Geo-Replication Data Transfer charge for the bandwidth of
replicating that data to another Azure region.
See Azure Storage Pricing for pricing information on the different storage types and availability options.
Availability sets
An availability set is a logical grouping of VMs within a datacenter that allows Azure to understand how your
application is built to provide for redundancy and availability. We recommended that two or more VMs are created
within an availability set to provide for a highly available application and to meet the 99.95% Azure SLA. There is
no cost for the Availability Set itself, you only pay for each VM instance that you create. When a single VM is using
Azure Premium Storage, the Azure SLA applies for unplanned maintenance events.
An availability set is composed of two additional groupings that protect against hardware failures and allow
updates to safely be applied - fault domains (FDs) and update domains (UDs). You can read more about how to
manage the availability of Linux VMs or Windows VMs.
Fault domains
A fault domain is a logical group of underlying hardware that share a common power source and network switch,
similar to a rack within an on-premises datacenter. As you create VMs within an availability set, the Azure platform
automatically distributes your VMs across these fault domains. This approach limits the impact of potential
physical hardware failures, network outages, or power interruptions.
Update domains
An update domain is a logical group of underlying hardware that can undergo maintenance or be rebooted at the
same time. As you create VMs within an availability set, the Azure platform automatically distributes your VMs
across these update domains. This approach ensures that at least one instance of your application always remains
running as the Azure platform undergoes periodic maintenance. The order of update domains being rebooted may
not proceed sequentially during planned maintenance, but only one update domain is rebooted at a time.
Managed Disk fault domains
For VMs using Azure Managed Disks, VMs are aligned with managed disk fault domains when using a managed
availability set. This alignment ensures that all the managed disks attached to a VM are within the same managed
disk fault domain. Only VMs with managed disks can be created in a managed availability set. The number of
managed disk fault domains varies by region - either two or three managed disk fault domains per region. You can
read more about these managed disk fault domains for Linux VMs or Windows VMs.
Availability zones
Availability zones, an alternative to availability sets, expand the level of control you have to maintain the availability
of the applications and data on your VMs. An Availability Zone is a physically separate zone within an Azure
region. There are three Availability Zones per supported Azure region. Each Availability Zone has a distinct power
source, network, and cooling. By architecting your solutions to use replicated VMs in zones, you can protect your
apps and data from the loss of a datacenter. If one zone is compromised, then replicated apps and data are instantly
available in another zone.
Learn more about deploying a Windows or Linux VM in an Availability Zone.
Next steps
You can now start to use these availability and redundancy features to build your Azure environment. For best
practices information, see Azure availability best practices.
Sizes for Windows virtual machines in Azure
This article describes the available sizes and options for the Azure virtual machines you can use to run your
Windows apps and workloads. It also provides deployment considerations to be aware of when you're planning
to use these resources. This article is also available for Linux virtual machines.
TYPE SIZES DESCRIPTION
General purpose B, Dsv3, Dv3, DSv2, Dv2, Av2, DC Balanced CPU-to-memory ratio. Ideal
for testing and development, small to
medium databases, and low to
medium traffic web servers.
Compute optimized Fsv2, Fs, F High CPU-to-memory ratio. Good for

medium traffic web servers, network
appliances, batch processes, and
application servers.
Memory optimized Esv3, Ev3, M, GS, G, DSv2, Dv2 High memory-to-CPU ratio. Great for
relational database servers, medium to
large caches, and in-memory analytics.
Storage optimized Ls High disk throughput and IO. Ideal for

Big Data, SQL, and NoSQL databases.
GPU NV, NVv2, NC, NCv2, NCv3, ND Specialized virtual machines targeted
for heavy graphic rendering and video
editing, as well as model training and
inferencing (ND) with deep learning.
Available with single or multiple GPUs.
High performance compute H Our fastest and most powerful CPU

virtual machines with optional high-
throughput network interfaces
(RDMA).
For information about pricing of the various sizes, see Virtual Machines Pricing.
To see general limits on Azure VMs, see Azure subscription and service limits, quotas, and constraints.
Storage costs are calculated separately based on used pages in the storage account. For details, Azure
Storage Pricing.
Learn more about how Azure compute units (ACU ) can help you compare compute performance across
Azure SKUs.
REST API
For information on using the REST API to query for VM sizes, see the following:
List available virtual machine sizes for resizing
List available virtual machine sizes for a subscription
List available virtual machine sizes in an availability set
ACU
Learn more about how Azure compute units (ACU ) can help you compare compute performance across Azure
SKUs.
Benchmark scores
Learn more about compute performance for Windows VMs using the CoreMark benchmark scores.
Next steps
Learn more about the different VM sizes that are available:
General purpose
Compute optimized
Memory optimized
Storage optimized
GPU optimized
Check the Previous generation page for A Standard, Dv1 (D1-4 and D11-14 v1), and A8-A11 series
General purpose virtual machine sizes
General purpose VM sizes provide balanced CPU -to-memory ratio. Ideal for testing and development, small to
medium databases, and low to medium traffic web servers. This article provides information about the number
of vCPUs, data disks and NICs as well as storage throughput for sizes in this grouping.
The DC -series is a new family of virtual machines in Azure that can help protect the confidentiality and
integrity of your data and code while it’s processed in the public cloud. These machines are backed by the
latest generation of 3.7GHz Intel XEON E -2176G Processor with SGX technology. With the Intel Turbo
Boost Technology these machines can go up to 4.7GHz. DC series instances enable customers to build
secure enclave-based applications to protect their code and data while it’s in use.
The Av2-series VMs can be deployed on a variety of hardware types and processors. A-series VMs have
CPU performance and memory configurations best suited for entry level workloads like development
and test. The size is throttled, based upon the hardware, to offer consistent processor performance for the
running instance, regardless of the hardware it is deployed on. To determine the physical hardware on
which this size is deployed, query the virtual hardware from within the Virtual Machine.
Example use cases include development and test servers, low traffic web servers, small to medium
databases, proof-of-concepts, and code repositories.
Dv2-series, a follow -on to the original D -series, features a more powerful CPU and optimal CPU -to-
memory configuration making them suitable for most production workloads. The Dv2-series CPU is
about 35% faster than the D -series CPU. It is based on the latest generation Intel Xeon® E5-2673 v3 2.4
GHz (Haswell) or E5-2673 v4 2.3 GHz (Broadwell) processors, and with the Intel Turbo Boost Technology
2.0, can go up to 3.1 GHz. The Dv2-series has the same memory and disk configurations as the D -series.
The Dv3-series features the 2.4 GHz Intel Xeon® E5-2673 v3 (Haswell) processor or the latest 2.3 GHz
Intel XEON ® E5-2673 v4 (Broadwell) processor in a hyper-threaded configuration, providing a better
value proposition for most general purpose workloads. Memory has been expanded (from ~3.5
GiB/vCPU to 4 GiB/vCPU ) while disk and network limits have been adjusted on a per core basis to align
with the move to hyperthreading. The Dv3 no longer has the high memory VM sizes of the D/Dv2
families, those have been moved to the new Ev3 family.
Example D -series use cases include enterprise-grade applications, relational databases, in-memory
caching, and analytics.
B-series
Premium Storage: Supported
Premium Storage Caching: Not Supported
The B -series burstable VMs are ideal for workloads that do not need the full performance of the CPU
continuously, like web servers, small databases and development and test environments. These workloads
typically have burstable performance requirements. The B -Series provides these customers the ability to
purchase a VM size with a price conscious baseline performance that allows the VM instance to build up credits
when the VM is utilizing less than its base performance. When the VM has accumulated credit, the VM can
burst above the VM’s baseline using up to 100% of the CPU when your application requires the higher CPU
performance.
Example use cases include development and test servers, low -traffic web servers, small databases, micro
services, servers for proof-of-concepts, build servers.
MAX
MAX UNCAC
CREDIT MAX LOCAL HED
S BANKE DISK DISK
LOCAL BASE BANKE D MAX PERF: PERF:
MEMOR SSD: PERF OF D/ CREDIT DATA IOPS / IOPS / MAX
SIZE VCPU Y: GIB GIB A CORE HOUR S DISKS MBPS MBPS NICS
Standa 1 1 4 10% 6 144 2 400 / 320 / 2

rd_B1s 10 10
Standa 1 2 4 20% 12 288 2 800 / 640 / 2

rd_B1 10 10
ms
Standa 2 4 8 40% 24 576 4 1600 / 1280 / 3

rd_B2s 15 15
Standa 2 8 16 60% 36 864 4 2400 / 1920 / 3

rd_B2 22.5 22.5
ms
Standa 4 16 32 90% 54 1296 8 3600 / 2880 / 4

rd_B4 35 35
ms
Standa 8 32 64 135% 81 1944 16 4320 / 4320 / 4

rd_B8 50 50
ms
Dsv3-series 1
ACU: 160-190
Premium Storage Caching: Supported
Dsv3-series sizes are based on the 2.4 GHz Intel Xeon® E5-2673 v3 (Haswell) processor or the latest 2.3 GHz
Intel XEON ® E5-2673 v4 (Broadwell) processor that can achieve 3.5GHz with Intel Turbo Boost Technology
2.0 and use premium storage. The Dsv3-series sizes offer a combination of vCPU, memory, and temporary
storage for most production workloads.
MAX
CACHED
AND TEMP
STORAGE MAX
THROUGHP UNCACHED MAX NICS /
UT: IOPS / DISK EXPECTED
TEMP MBPS THROUGHP NETWORK
MEMORY: STORAGE MAX DATA (CACHE SIZE UT: IOPS / BANDWIDT
SIZE VCPU GIB (SSD) GIB DISKS IN GIB) MBPS H (MBPS)
Standard_D 2 8 16 4 4,000 / 32 3,200 / 48 2 / 1,000

2s_v3 (50)
Standard_D 4 16 32 8 8,000 / 64 6,400 / 96 2 / 2,000

4s_v3 (100)
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_D 8 32 64 16 16,000 / 12,800 / 4 / 4,000

8s_v3 128 (200) 192
Standard_D 16 64 128 32 32,000 / 25,600 / 8 / 8,000

16s_v3 256 (400) 384
Standard_D 32 128 256 32 64,000 / 51,200 / 8 / 16,000

32s_v3 512 (800) 768
Standard_D 64 256 512 32 128,000 / 80,000 / 8 / 30,000

64s_v3 1024 1200
(1600)
1 Dsv3 -series VM’s feature Intel® Hyper-Threading Technology
Dv3-series 1
ACU: 160-190
Premium Storage: Not Supported
Dv3-series sizes are based on the 2.4 GHz Intel Xeon® E5-2673 v3 (Haswell) processor or 2.3 GHz Intel XEON
® E5-2673 v4 (Broadwell) processor that can achieve 3.5GHz with Intel Turbo Boost Technology 2.0. The Dv3-
series sizes offer a combination of vCPU, memory, and temporary storage for most production workloads.
Data disk storage is billed separately from virtual machines. To use premium storage disks, use the Dsv3 sizes.
The pricing and billing meters for Dsv3 sizes are the same as Dv3-series.
MAX TEMP
STORAGE
THROUGHPUT:
TEMP IOPS / READ MAX NICS /
STORAGE MAX DATA MBPS / WRITE NETWORK
SIZE VCPU MEMORY: GIB (SSD) GIB DISKS MBPS BANDWIDTH
Standard_D2_ 2 8 50 4 3000/46/23 2 / 1,000

v3
Standard_D4_ 4 16 100 8 6000/93/46 2 / 2,000

v3
Standard_D8_ 8 32 200 16 12000/187/9 4 / 4,000

v3 3
Standard_D1 16 64 400 32 24000/375/1 8 / 8,000

6_v3 87
MAX TEMP
STORAGE
THROUGHPUT:
TEMP IOPS / READ MAX NICS /
STORAGE MAX DATA MBPS / WRITE NETWORK
Standard_D3 32 128 800 32 48000/750/3 8 / 16,000

2_v3 75
Standard_D6 64 256 1600 32 96000/1000/ 8 / 30,000

4_v3 500
1 Dv3 -series VM’s feature Intel® Hyper-Threading Technology
DSv2-series
ACU: 210-250
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_D 1 3.5 7 4 4,000 / 32 3,200 / 48 2 / 750

S1_v2 (43)
Standard_D 2 7 14 8 8,000 / 64 6,400 / 96 2 / 1500

S2_v2 (86)
Standard_D 4 14 28 16 16,000 / 12,800 / 4 / 3000

S3_v2 128 (172) 192
Standard_D 8 28 56 32 32,000 / 25,600 / 8 / 6000

S4_v2 256 (344) 384
Standard_D 16 56 112 64 64,000 / 51,200 / 8 / 12000

S5_v2 512 (688) 768
Dv2-series
ACU: 210-250
MAX TEMP
STORAGE
THROUGHP MAX NICS /
UT: IOPS / EXPECTED
TEMP READ MBPS NETWORK
MEMORY: STORAGE / WRITE MAX DATA THROUGHP BANDWIDT
SIZE VCPU GIB (SSD) GIB MBPS DISKS UT: IOPS H (MBPS)
Standard_D 1 3.5 50 3000 / 46 / 4 4x500 2 / 750

1_v2 23
Standard_D 2 7 100 6000 / 93 / 8 8x500 2 / 1500

2_v2 46
Standard_D 4 14 200 12000 / 16 16x500 4 / 3000

3_v2 187 / 93
Standard_D 8 28 400 24000 / 32 32x500 8 / 6000

4_v2 375 / 187
Standard_D 16 56 800 48000 / 64 64x500 8 / 12000

5_v2 750 / 375
Av2-series
ACU: 100
MAX TEMP
STORAGE MAX NICS /
THROUGHPUT: MAX DATA EXPECTED
TEMP IOPS / READ DISKS / NETWORK
STORAGE MBPS / WRITE THROUGHPUT: BANDWIDTH
SIZE VCPU MEMORY: GIB (SSD) GIB MBPS IOPS (MBPS)
Standard_A1_ 1 2 10 1000 / 20 / 2 / 2x500 2 / 250

v2 10
Standard_A2_ 2 4 20 2000 / 40 / 4 / 4x500 2 / 500

v2 20
Standard_A4_ 4 8 40 4000 / 80 / 8 / 8x500 4 / 1000

v2 40
Standard_A8_ 8 16 80 8000 / 160 / 16 / 16x500 8 / 2000

v2 80
Standard_A2 2 16 20 2000 / 40 / 4 / 4x500 2 / 500

m_v2 20
Standard_A4 4 32 40 4000 / 80 / 8 / 8x500 4 / 1000

m_v2 40
Standard_A8 8 64 80 8000 / 160 / 16 / 16x500 8 / 2000

m_v2 80
DC-series
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_D 2 8 100 2 4000 / 32 3200 /48 2 / 1500

C2s (43)
Standard_D 4 16 200 4 8000 / 64 6400 /96 2 / 3000

C4s (86)
Size table definitions

Storage capacity is shown in units of GiB or 1024^3 bytes. When comparing disks measured in GB (1000^3
bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may appear smaller.
For example, 1023 GiB = 1098.4 GB
Disk throughput is measured in input/output operations per second (IOPS ) and MBps where MBps = 10^6
bytes/sec.
Data disks can operate in cached or uncached modes. For cached data disk operation, the host cache mode is
set to ReadOnly or ReadWrite. For uncached data disk operation, the host cache mode is set to None.
If you want to get the best performance for your VMs, you should limit the number of data disks to 2 disks
per vCPU.
Expected network bandwidth is the maximum aggregated bandwidth allocated per VM type across all
NICs, for all destinations. Upper limits are not guaranteed, but are intended to provide guidance for selecting
the right VM type for the intended application. Actual network performance will depend on a variety of
factors including network congestion, application loads, and network settings. For information on optimizing
network throughput, see Optimizing network throughput for Windows and Linux. To achieve the expected
network performance on Linux or Windows, it may be necessary to select a specific version or optimize your
VM. For more information, see How to reliably test for virtual machine throughput.
Other sizes
Compute optimized
Memory optimized
Storage optimized
GPU optimized
Next steps
SKUs.
B-series burstable virtual machine sizes
The B -series VM family allows you to choose which VM size provides you the necessary base level performance
for your workload, with the ability to burst CPU performance up to 100% of an Intel® Broadwell E5-2673 v4 2.3
GHz, or an Intel® Haswell 2.4 GHz E5-2673 v3 processor vCPU.
The B -series VMs are ideal for workloads that do not need the full performance of the CPU continuously, like web
servers, proof of concepts, small databases and development build environments. These workloads typically have
burstable performance requirements. The B -series provides you with the ability to purchase a VM size with
baseline performance and the VM instance builds up credits when it is using less than its baseline. When the VM
has accumulated credit, the VM can burst above the baseline using up to 100% of the vCPU when your application
requires higher CPU performance.
The B -series comes in the following six VM sizes:
TEMP CREDITS MAX

MEMORY: STORAGE BASE CPU MAX CPU BANKED / BANKED
SIZE VCPU'S GIB (SSD) GIB PERF OF VM PERF OF VM HOUR CREDITS
Standard_B 1 1 4 10% 100% 6 144

1s
Standard_B 1 2 4 20% 100% 12 288

1ms
Standard_B 2 4 8 40% 200% 24 576

2s
Standard_B 2 8 16 60% 200% 36 864

2ms
Standard_B 4 16 32 90% 400% 54 1296

4ms
Standard_B 8 32 64 135% 800% 81 1944

8ms
Q&A
Q: How do you get 135% baseline performance from a VM?
A: The 135% is shared amongst the 8 vCPU’s that make up the VM size. For example, if your application uses 4 of
the 8 cores working on batch processing and each of those 4 vCPU’s are running at 30% utilization the total
amount of VM CPU performance would equal 120%. Meaning that your VM would be building credit time based
on the 15% delta from your baseline performance. But it also means that when you have credits available that
same VM can use 100% of all 8 vCPU’s giving that VM a Max CPU performance of 800%.
Q: How can I monitor my credit balance and consumption
A: We will be introducing 2 new metrics in the coming weeks, the Credit metric will allow you to view how many
credits your VM has banked and the ConsumedCredit metric will show how many CPU credits your VM has
consumed from the bank. You will be able to view these metrics from the metrics pane in the portal or
programmatically through the Azure Monitor APIs.
For more information on how to access the metrics data for Azure, see Overview of metrics in Microsoft Azure.
Q: How are credits accumulated?
A: The VM accumulation and consumption rates are set such that a VM running at exactly its base performance
level will have neither a net accumulation or consumption of bursting credits. A VM will have a net increase in
credits whenever it is running below its base performance level and will have a net decrease in credits whenever
the VM is utilizing the CPU more than its base performance level.
Example: I deploy a VM using the B1ms size for my small time and attendance database application. This size
allows my application to use up to 20% of a vCPU as my baseline, which is 0.2 credits per minute I can use or
bank.
My application is busy at the beginning and end of my employees work day, between 7:00-9:00 AM and 4:00 -
6:00PM. During the other 20 hours of the day, my application is typically at idle, only using 10% of the vCPU. For
the non-peak hours, I earn 0.2 credits per minute but only consume 0.l credits per minute, so my VM will bank 0.1
x 60 = 6 credits per hour. For the 20 hours that I am off-peak, I will bank 120 credits.
During peak hours my application averages 60% vCPU utilization, I still earn 0.2 credits per minute but I consume
0.6 credits per minute, for a net cost of 0.4 credits a minute or 0.4 x 60 = 24 credits per hour. I have 4 hours per day
of peak usage, so it costs 4 x 24 = 96 credits for my peak usage.
If I take the 120 credits I earned off-peak and subtract the 96 credits I used for my peak times, I bank an additional
24 credits per day that I can use for other bursts of activity.
Q: Does the B -Series support Premium Storage data disks?
A: Yes, all B -Series sizes support Premium Storage data disks.
Q: Why is my remaining credit set to 0 after a redeploy or a stop/start?
A : When a VM is “REDPLOYED” and the VM moves to another node, the accumulated credit is lost. If the VM is
stopped/started, but remains on the same node, the VM retains the accumulated credit. Whenever the VM starts
fresh on a node, it gets an initial credit, for Standard_B8ms it is 240 mins.
Other sizes
General purpose
Compute optimized
Memory optimized
Storage optimized
GPU optimized
Next steps
SKUs.
Compute optimized virtual machine sizes
Compute optimized VM sizes have a high CPU -to-memory ratio and are good for medium traffic web servers,
network appliances, batch processes, and application servers. This article provides information about the
number of vCPUs, data disks, and NICs as well as storage throughput and network bandwidth for each size in
this grouping.
Fsv2-series is based on the Intel® Xeon® Platinum 8168 processor, featuring a base core frequency of 2.7 GHz
and a maximum single-core turbo frequency of 3.7 GHz. Intel® AVX-512 instructions, which are new on Intel
Scalable Processors, will provide up to a 2X performance boost to vector processing workloads on both single
and double precision floating point operations. In other words, they are really fast for any computational
workload.
At a lower per-hour list price, the Fsv2-series is the best value in price-performance in the Azure portfolio based
on the Azure Compute Unit (ACU ) per vCPU.
F -series is based on the 2.4 GHz Intel Xeon® E5-2673 v3 (Haswell) processor, which can achieve clock speeds
as high as 3.1 GHz with the Intel Turbo Boost Technology 2.0. This is the same CPU performance as the Dv2-
series of VMs.
F -series VMs are an excellent choice for workloads that demand faster CPUs but do not need as much memory
or temporary storage per vCPU. Workloads such as analytics, gaming servers, web servers, and batch
processing will benefit from the value of the F -series.
The Fs-series provides all the advantages of the F -series, in addition to Premium storage.
Fsv2-series 1
ACU: 195 - 210
MAX
CACHED
AND TEMP
STORAGE MAX
SIZE VCPU'S GIB (SSD) GIB DISKS IN GIB) MBPS H (MBPS)
Standard_F 2 4 16 4 4000 / 31 3200 / 47 Moderate

2s_v2 (32)
Standard_F 4 8 32 8 8000 / 63 6400 / 95 Moderate

4s_v2 (64)
Standard_F 8 16 64 16 16000 / 12800 / High

8s_v2 127 (128) 190
MAX
CACHED
AND TEMP
STORAGE MAX
SIZE VCPU'S GIB (SSD) GIB DISKS IN GIB) MBPS H (MBPS)
Standard_F 16 32 128 32 32000 / 25600 / High

16s_v2 255 (256) 380
Standard_F 32 64 256 32 64000 / 51200 / Extremely

32s_v2 512 (512) 750 High

64s_v2 1024 1100 High
(1024)

72s_v22, 3 1152 1100 High
(1520)
1 Fsv2 -series VM’s feature Intel® Hyper-Threading Technology

2 More than 64vCPU’s require one of these supported guest OSes: Windows Server 2016, Ubuntu 16.04 LTS,
SLES 12 SP2, and Red Hat Enterprise Linux, CentOS 7.3, or Oracle Linux 7.3 with LIS 4.2.1
3 Instance is isolated to hardware dedicated to a single customer.
Fs-series 1
ACU: 210 - 250
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_F 1 2 4 4 4,000 / 32 3,200 / 48 2 / 750

1s (12)
Standard_F 2 4 8 8 8,000 / 64 6,400 / 96 2 / 1500

2s (24)
Standard_F 4 8 16 16 16,000 / 12,800 / 4 / 3000

4s 128 (48) 192
Standard_F 8 16 32 32 32,000 / 25,600 / 8 / 6000

8s 256 (96) 384
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_F 16 32 64 64 64,000 / 51,200 / 8 / 12000

16s 512 (192) 768
MBps = 10^6 bytes per second, and GiB = 1024^3 bytes.

1 The maximum disk throughput (IOPS or MBps) possible with a Fs series VM may be limited by the number,
size, and striping of the attached disk(s). For details, see Premium Storage: High-performance storage for Azure
virtual machine workloads.
F-series
ACU: 210 - 250
MAX TEMP
STORAGE MAX NICS /
TEMP IOPS / READ DISKS / NETWORK
STORAGE MBPS / WRITE THROUGHPUT: BANDWIDTH
Standard_F1 1 2 16 3000 / 46 / 4 / 4x500 2 / 750

23
Standard_F2 2 4 32 6000 / 93 / 8 / 8x500 2 / 1500

46
Standard_F4 4 8 64 12000 / 187 / 16 / 16x500 4 / 3000

93
Standard_F8 8 16 128 24000 / 375 / 32 / 32x500 8 / 6000

187
Standard_F16 16 32 256 48000 / 750 / 64 / 64x500 8 / 12000

375

bytes/sec.
per vCPU.
the right VM type for the intended application. Actual network performance will depend on a variety of
factors including network congestion, application loads, and network settings. For information on optimizing
network throughput, see Optimizing network throughput for Windows and Linux. To achieve the expected
network performance on Linux or Windows, it may be necessary to select a specific version or optimize your
VM. For more information, see How to reliably test for virtual machine throughput.
Other sizes
General purpose
Memory optimized
Storage optimized
GPU optimized
Next steps
SKUs.
Memory optimized virtual machine sizes
Memory optimized VM sizes offer a high memory-to-CPU ratio that are great for relational database servers,
medium to large caches, and in-memory analytics. This article provides information about the number of vCPUs,
data disks and NICs as well as storage throughput and network bandwidth for each size in this grouping.
The M -Series offers the highest vCPU count (up to 128 vCPUs) and largest memory (up to 3.8 TiB ) of any
VM in the cloud. It’s ideal for extremely large databases or other applications that benefit from high vCPU
counts and large amounts of memory.
Dv2-series, G -series, and the DSv2/GS counterparts are ideal for applications that demand faster vCPUs,
better temporary storage performance, or have higher memory demands. They offer a powerful
combination for many enterprise-grade applications.
Dv2-series, a follow -on to the original D -series, features a more powerful CPU. The Dv2-series CPU is
about 35% faster than the D -series CPU. It is based on the latest generation 2.4 GHz Intel Xeon® E5-2673
v3 2.4 GHz (Haswell) or E5-2673 v4 2.3 GHz (Broadwell) processors, and with the Intel Turbo Boost
Technology 2.0, can go up to 3.1 GHz. The Dv2-series has the same memory and disk configurations as the
D -series.
The Ev3-series features the E5-2673 v4 2.3 GHz (Broadwell) processor in a hyper-threaded configuration,
providing a better value proposition for most general purpose workloads, and bringing the Ev3 into
alignment with the general purpose VMs of most other clouds. Memory has been expanded (from 7
GiB/vCPU to 8 GiB/vCPU ) while disk and network limits have been adjusted on a per core basis to align
with the move to hyperthreading. The Ev3 is the follow up to the high memory VM sizes of the D/Dv2
families.
Azure Compute offers virtual machine sizes that are Isolated to a specific hardware type and dedicated to a
single customer. These virtual machine sizes are best suited for workloads that require a high degree of
isolation from other customers for workloads involving elements like compliance and regulatory
requirements. Customers can also choose to further subdivide the resources of these Isolated virtual
machines by using Azure support for nested virtual machines . Please see the tables of virtual machine
families below for your isolated VM options.
Esv3-series
ACU: 160-190 1
ESv3-series instances are based on the 2.3 GHz Intel XEON ® E5-2673 v4 (Broadwell) processor and can achieve
3.5GHz with Intel Turbo Boost Technology 2.0 and use premium storage. Ev3-series instances are ideal for
memory-intensive enterprise applications.
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_E 2 16 32 4 4,000 / 32 3,200 / 48 2 / 1,000

2s_v3 (50)
Standard_E 4 32 64 8 8,000 / 64 6,400 / 96 2 / 2,000

4s_v3 2 (100)
Standard_E 8 64 128 16 16,000 / 12,800 / 4 / 4,000

8s_v3 2 128 (200) 192
Standard_E 16 128 256 32 32,000 / 25,600 / 8 / 8,000

16s_v3 2 256 (400) 384
Standard_E 20 160 320 32 40,000 / 32,000 / 8 / 10,000

20s_v3 2 320 (400) 480
Standard_E 32 256 512 32 64,000 / 51,200 / 8 / 16,000

32s_v3 2 512 (800) 768
Standard_E 64 432 864 32 128,000/10 80,000 / 8 / 30,000

64s_v3 2 24 (1600) 1200
Standard_E 64 432 864 32 128,000/10 80,000 / 8 / 30,000

64is_v3 3 24 (1600) 1200
1 Esv3 -series VM’s feature Intel® Hyper-Threading Technology.

2 Constrained core sizes available.
Ev3-series
ACU: 160 - 190 1
Ev3-series instances are based on the 2.3 GHz Intel XEON ® E5-2673 v4 (Broadwell) processor and can achieve
3.5GHz with Intel Turbo Boost Technology 2.0. Ev3-series instances are ideal for memory-intensive enterprise
applications.
Data disk storage is billed separately from virtual machines. To use premium storage disks, use the ESv3 sizes. The
pricing and billing meters for ESv3 sizes are the same as Ev3-series.
MAX TEMP
STORAGE
THROUGHPUT:
IOPS / READ MAX NICS /
TEMP STORAGE MAX DATA MBPS / WRITE NETWORK
Standard_E2_ 2 16 50 4 3000/46/23 2 / 1,000

v3
Standard_E4_ 4 32 100 8 6000/93/46 2 / 2,000

v3
Standard_E8_ 8 64 200 16 12000/187/9 4 / 4,000

v3 3
Standard_E16 16 128 400 32 24000/375/1 8 / 8,000

_v3 87
Standard_E20 20 160 500 32 30000/469/2 8 / 10,000

_v3 34
Standard_E32 32 256 800 32 48000/750/3 8 / 16,000

_v3 75
Standard_E64 64 432 1600 32 96000/1000/ 8 / 30,000

_v3 500
Standard_E64i 64 432 1600 32 96000/1000/ 8 / 30,000

_v3 2, 3 500
1 Ev3 -series VM’s feature Intel® Hyper-Threading Technology.

M-series
ACU: 160-180 1
Write Accelerator: Supported
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_ 8 218.75 256 8 10,000 / 5,000 / 125 4 / 2,000

M8ms 3 100 (793)
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_ 16 437.5 512 16 20,000 / 10,000 / 8 / 4,000

M16ms 3 200 (1,587) 250
Standard_ 32 192 1,024 32 40,000 / 20,000 / 8 / 8,000

M32ts 400 (3,174) 500
Standard_ 32 256 1,024 32 40,000 / 20,000 / 8 / 8,000

M32ls 400 (3,174) 500
Standard_ 32 875 1,024 32 40,000 / 20,000 / 8 / 8,000

M32ms 3 400 (3,174) 500
Standard_ 64 1,024 2,048 64 80,000 / 40,000 / 8 / 16,000

M64s 800 (6,348) 1,000
Standard_ 64 512 2,048 64 80,000 / 40,000 / 8 / 16,000

M64ls 800 (6,348) 1,000
Standard_ 64 1,792 2,048 64 80,000 / 40,000 / 8 / 16,000

M64ms 3 800 (6,348) 1,000
Standard_ 128 2,048 4,096 64 160,000 / 80,000 / 8 / 30,000

M128s 2, 3 1,600 2,000
(12,696)
Standard_ 128 3,892 4,096 64 160,000 / 80,000 / 8 / 30,000

M128ms 2, 1,600 2,000
3, 4 (12,696)
Standard_ 64 1,024 7,168 64 80,000 / 40,000 / 8 / 16,000

M64 800 (1,228) 1,000
Standard_ 64 1,792 7,168 64 80,000 / 40,000 / 8 / 16,000

M64m 800 (1,228) 1,000
Standard_ 128 2,048 14,336 64 250,000 / 80,000 / 8 / 32,000

M128 2 1,600 2,000
(2,456)
Standard_ 128 3,892 14,336 64 250,000 / 80,000 / 8 / 32,000

M128m 2 1,600 2,000
(2,456)
1 M -series VM’s feature Intel® Hyper-Threading Technology

2 More than 64vCPU’s require one of these supported guest OSes: Windows Server 2016, Ubuntu 16.04 LTS,
SLES 12 SP2, and Red Hat Enterprise Linux, CentOS 7.3 or Oracle Linux 7.3 with LIS 4.2.1.
4
GS-series
ACU: 180 - 240 1
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_G 2 28 56 8 10,000 / 5,000 / 125 2 / 2000

S1 100 (264)
Standard_G 4 56 112 16 20,000 / 10,000 / 2 / 4000

S2 200 (528) 250
Standard_G 8 112 224 32 40,000 / 20,000 / 4 / 8000

S3 400 (1,056) 500
Standard_G 16 224 448 64 80,000 / 40,000 / 8 / 16000

S4 3 800 (2,112) 1,000
Standard_G 32 448 896 64 160,000 / 80,000 / 8 / 20000

S5 2, 3 1,600 2,000
(4,224)
1 The maximum disk throughput (IOPS or MBps) possible with a GS series VM may be limited by the number,
size and striping of the attached disk(s). For details, see Premium Storage: High-performance storage for Azure
G-series
ACU: 180 - 240
MAX TEMP
STORAGE MAX NICS /
IOPS / READ DISKS / NETWORK
TEMP STORAGE MBPS / WRITE THROUGHPUT: BANDWIDTH
Standard_G1 2 28 384 6000 / 93 / 8 / 8 x 500 2 / 2000

46
Standard_G2 4 56 768 12000 / 187 / 16 / 16 x 500 2 / 4000

93
Standard_G3 8 112 1,536 24000 / 375 / 32 / 32 x 500 4 / 8000

187
Standard_G4 16 224 3,072 48000 / 750 / 64 / 64 x 500 8 / 16000

375
Standard_G5 32 448 6,144 96000 / 1500 64 / 64 x 500 8 / 20000

1 / 750
DSv2-series 11-15
ACU: 210 - 250 1
MAX
CACHED
AND TEMP
STORAGE MAX
Standard_D 2 14 28 8 8,000 / 64 6,400 / 96 2 / 1500

S11_v2 3 (72)
Standard_D 4 28 56 16 16,000 / 12,800 / 4 / 3000

S12_v2 3 128 (144) 192
Standard_D 8 56 112 32 32,000 / 25,600 / 8 / 6000

S13_v2 3 256 (288) 384
Standard_D 16 112 224 64 64,000 / 51,200 / 8 / 12000

S14_v2 3 512 (576) 768
Standard_D 20 140 280 64 80,000 / 64,000 / 8 / 25000 4

S15_v2 2 640 (720) 960
1 The maximum disk throughput (IOPS or MBps) possible with a DSv2 series VM may be limited by the number,
size and striping of the attached disk(s). For details, see Premium Storage: High-performance storage for Azure
2
4 25000 Mbps with Accelerated Networking.
Dv2-series 11-15
ACU: 210 - 250
MAX TEMP
STORAGE MAX NICS /
Standard_D11 2 14 100 6000 / 93 / 8 / 8x500 2 / 1500

_v2 46
Standard_D12 4 28 200 12000 / 187 / 16 / 16x500 4 / 3000

_v2 93
Standard_D13 8 56 400 24000 / 375 / 32 / 32x500 8 / 6000

_v2 187
Standard_D14 16 112 800 48000 / 750 / 64 / 64x500 8 / 12000

_v2 375
Standard_D15 20 140 1,000 60000 / 937 / 64 / 64x500 8 / 25000 2

_v2 1 468
2 25000 Mbps with Accelerated Networking.

bytes/sec.
If you want to get the best performance for your VMs, you should limit the number of data disks to 2 disks per
vCPU.
the right VM type for the intended application. Actual network performance will depend on a variety of factors
including network congestion, application loads, and network settings. For information on optimizing network
throughput, see Optimizing network throughput for Windows and Linux. To achieve the expected network
performance on Linux or Windows, it may be necessary to select a specific version or optimize your VM. For
more information, see How to reliably test for virtual machine throughput.
Other sizes
General purpose
Compute optimized
Storage optimized
GPU optimized
Next steps
SKUs.
Constrained vCPU capable VM sizes
Some database workloads like SQL Server or Oracle require high memory, storage, and I/O bandwidth, but not a
high core count. Many database workloads are not CPU -intensive. Azure offers certain VM sizes where you can
constrain the VM vCPU count to reduce the cost of software licensing, while maintaining the same memory,
storage, and I/O bandwidth.
The vCPU count can be constrained to one half or one quarter of the original VM size. These new VM sizes have a
suffix that specifies the number of active vCPUs to make them easier for you to identify.
For example, the current VM size Standard_GS5 comes with 32 vCPUs, 448 GB RAM, 64 disks (up to 256 TB ), and
80,000 IOPs or 2 GB/s of I/O bandwidth. The new VM sizes Standard_GS5-16 and Standard_GS5-8 comes with
16 and 8 active vCPUs respectively, while maintaining the rest of the specs of the Standard_GS5 for memory,
storage, and I/O bandwidth.
The licensing fees charged for SQL Server or Oracle are constrained to the new vCPU count, and other products
should be charged based on the new vCPU count. This results in a 50% to 75% increase in the ratio of the VM
specs to active (billable) vCPUs. These new VM sizes that are only available in Azure, allowing workloads to push
higher CPU utilization at a fraction of the (per-core) licensing cost. At this time, the compute cost, which includes
OS licensing, remains the same one as the original size. For more information, see Azure VM sizes for more cost-
effective database workloads.
NAME VCPU SPECS
Standard_M8-2ms 2 Same as M8ms
Standard_E4-2s_v3 2 Same as E4s_v3

NAME VCPU SPECS
Standard_E32-16_v3 16 Same as E32s_v3
Standard_GS4-8 8 Same as GS4
Standard_DS11-1_v2 1 Same as DS11_v2
Other sizes
Compute optimized
Memory optimized
Storage optimized
GPU
Next steps
SKUs.
Storage optimized virtual machine sizes
Storage optimized VM sizes offer high disk throughput and IO, and are ideal for Big Data, SQL, and NoSQL
databases. This article provides information about the number of vCPUs, data disks and NICs as well as storage
throughput and network bandwidth for each size in this grouping.
The Ls-series offers up to 32 vCPUs, using the Intel® Xeon® processor E5 v3 family. The Ls-series gets the same
CPU performance as the G/GS -Series and comes with 8 GiB of memory per vCPU. Ls-series VMs are ideal for
applications requiring low latency, high throughput, and large local disk storage.
Example use cases include NoSQL databases such as Cassandra, MongoDB, Cloudera, and Redis, data
warehousing, and large transactional databases.
NOTE
The Ls-series is optimized for use of the temporary disk attached to the VM machine rather than use of durable data disks.
The high throughput and IOPS of the temporary disk makes the Ls-series ideal for NoSQL stores such as Apache
Cassandra and MongoDB which replicate data across multiple VMs to achieve persistence in the event of the failure of a
single VM. The Ls-series does not support the creation of a local cache to increase the IOPS achievable by durable data
disks.
Ls-series
ACU: 180-240
MAX
MAX TEMP UNCACHED MAX NICS /
STORAGE DISK EXPECTED
TEMP THROUGHP THROUGHP NETWORK
MEMORY: STORAGE MAX DATA UT: IOPS / UT: IOPS / BANDWIDT
SIZE VCPU GIB (SSD) GIB DISKS MBPS MBPS H (MBPS)
Standard_L 4 32 678 16 20,000 / 5,000 / 125 2 / 4,000

4s 200
Standard_L 8 64 1,388 32 40,000 / 10,000 / 4 / 8,000

8s 400 250
Standard_L 16 128 2,807 64 80,000 / 20,000 / 8 / 16,000

16s 800 500
Standard_L 32 256 5,630 64 160,000 / 40,000 / 8 / 20,000

32s 1 1,600 1,000
The maximum disk throughput possible with Ls-series VMs may be limited by the number, size, and striping of
any attached disks. For details, see Premium Storage: High-performance storage for Azure virtual machine
workloads.
bytes/sec.
vCPU.
the right VM type for the intended application. Actual network performance will depend on a variety of factors
Other sizes
General purpose
Compute optimized
Memory optimized
GPU optimized
Next steps
SKUs.
GPU optimized virtual machine sizes
GPU optimized VM sizes are specialized virtual machines available with single or multiple NVIDIA GPUs.
These sizes are designed for compute-intensive, graphics-intensive, and visualization workloads. This article
provides information about the number and type of GPUs, vCPUs, data disks, and NICs. Storage throughput
and network bandwidth are also included for each size in this grouping.
NC, NCv2, NCv3, and ND sizes are optimized for compute-intensive and network-intensive applications
and algorithms. Some examples are CUDA- and OpenCL -based applications and simulations, AI, and Deep
Learning. The NCv3-series is focused on high-performance computing workloads featuring NVIDIA’s Tesla
V100 GPU. The ND -series is focused on training and inference scenarios for deep learning. It uses the
NVIDIA Tesla P40 GPU.
NV and NVv2 sizes are optimized and designed for remote visualization, streaming, gaming, encoding, and
VDI scenarios using frameworks such as OpenGL and DirectX. These VMs are backed by the NVIDIA Tesla
M60 GPU.
NC-series
NC -series VMs are powered by the NVIDIA Tesla K80 card. Users can crunch through data faster by leveraging
CUDA for energy exploration applications, crash simulations, ray traced rendering, deep learning and more. The
NC24r configuration provides a low latency, high-throughput network interface optimized for tightly coupled
parallel computing workloads.
TEMP
STORAGE MAX DATA
SIZE VCPU MEMORY: GIB (SSD) GIB GPU DISKS MAX NICS
Standard_NC 6 56 340 1 24 1
6
Standard_NC 12 112 680 2 48 2

12
Standard_NC 24 224 1440 4 64 4

24
Standard_NC 24 224 1440 4 64 4

24r*
1 GPU = one-half K80 card.

*RDMA capable
NCv2-series
NCv2-series VMs are powered by NVIDIA Tesla P100 GPUs. These GPUs can provide more than 2x the
computational performance of the NC -series. Customers can take advantage of these updated GPUs for
traditional HPC workloads such as reservoir modeling, DNA sequencing, protein analysis, Monte Carlo
simulations, and others. The NC24rs v2 configuration provides a low latency, high-throughput network
interface optimized for tightly coupled parallel computing workloads.
IMPORTANT
For this size family, the vCPU (core) quota in your subscription is initially set to 0 in each region. Request a vCPU quota
increase for this family in an available region.
TEMP
STORAGE MAX DATA
Standard_NC 6 112 736 1 12 4

6s_v2
Standard_NC 12 224 1474 2 24 8

12s_v2
Standard_NC 24 448 2948 4 32 8

24s_v2
Standard_NC 24 448 2948 4 32 8

24rs_v2*
1 GPU = one P100 card.

*RDMA capable
NCv3-series
NCv3-series VMs are powered by NVIDIA Tesla V100 GPUs. These GPUs can provide 1.5x the computational
performance of the NCv2-series. Customers can take advantage of these updated GPUs for traditional HPC
workloads such as reservoir modeling, DNA sequencing, protein analysis, Monte Carlo simulations, and others.
The NC24rs v3 configuration provides a low latency, high-throughput network interface optimized for tightly
coupled parallel computing workloads.
IMPORTANT
For this size family, the vCPU (core) quota in your subscription is initially set to 0 in each region. Request a vCPU quota
TEMP
STORAGE MAX DATA
Standard_NC 6 112 736 1 12 4

6s_v3
TEMP
STORAGE MAX DATA
Standard_NC 12 224 1474 2 24 8

12s_v3
Standard_NC 24 448 2948 4 32 8

24s_v3
Standard_NC 24 448 2948 4 32 8

24rs_v3*
1 GPU = one V100 card.

*RDMA capable
ND-series
The ND -series virtual machines are a new addition to the GPU family designed for AI and Deep Learning
workloads. They offer excellent performance for training and inference. ND instances are powered by NVIDIA
Tesla P40 GPUs. These instances provide excellent performance for single-precision floating point operations,
for AI workloads utilizing Microsoft Cognitive Toolkit, TensorFlow, Caffe, and other frameworks. The ND -series
also offers a much larger GPU memory size (24 GB ), enabling to fit much larger neural net models. Like the
NC -series, the ND -series offers a configuration with a secondary low -latency, high-throughput network
through RDMA, and InfiniBand connectivity so you can run large-scale training jobs spanning many GPUs.
IMPORTANT
For this size family, the vCPU (core) quota per region in your subscription is initially set to 0. Request a vCPU quota
TEMP
STORAGE MAX DATA
Standard_ND 6 112 736 1 12 4

6s
Standard_ND 12 224 1474 2 24 8

12s
Standard_ND 24 448 2948 4 32 8

24s
Standard_ND 24 448 2948 4 32 8

24rs*
1 GPU = one P40 card.

*RDMA capable
NV-series
The NV -series virtual machines are powered by NVIDIA Tesla M60 GPUs and NVIDIA GRID technology for
desktop accelerated applications and virtual desktops where customers are able to visualize their data or
simulations. Users are able to visualize their graphics intensive workflows on the NV instances to get superior
graphics capability and additionally run single precision workloads such as encoding and rendering.
Each GPU in NV instances comes with a GRID license. This license gives you the flexibility to use an NV
instance as a virtual workstation for a single user, or 25 concurrent users can connect to the VM for a virtual
application scenario.
TEMP MAX VIRTUAL VIRTUAL

MEMORY: STORAGE DATA WORKSTA APPLICATI
SIZE VCPU GIB (SSD) GIB GPU DISKS MAX NICS TIONS ONS
Standard 6 56 340 1 24 1 1 25
_NV6
Standard 12 112 680 2 48 2 2 50

_NV12
Standard 24 224 1440 4 64 4 4 100

_NV24
1 GPU = one-half M60 card.
NVv2-series (Preview)
The NVv2-series virtual machines are powered by NVIDIA Tesla M60 GPUs and NVIDIA GRID technology
with Intel Broadwell CPUs. These virtual machines are targeted for GPU accelerated graphics applications and
virtual desktops where customers want to visualize their data, simulate results to view, work on CAD, or render
and stream content. Additionally, these virtual machines can run single precision workloads such as encoding
and rendering. NVv2 virtual machines support Premium Storage and come with twice the system memory
(RAM ) when compared with its predecessor NV -series.
Each GPU in NVv2 instances comes with a GRID license. This license gives you the flexibility to use an NV
instance as a virtual workstation for a single user, or 25 concurrent users can connect to the VM for a virtual
application scenario.

Standard 6 112 320 1 12 4 1 25

_NV6s_v2
Standard 12 224 640 2 24 8 2 50

_NV12s_v
2
Standard 24 448 1280 4 32 8 4 100

_NV24s_v
2
1 GPU = one-half M60 card.

bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may appear
smaller. For example, 1023 GiB = 1098.4 GB
bytes/sec.
per vCPU.
NICs, for all destinations. Upper limits are not guaranteed, but are intended to provide guidance for
selecting the right VM type for the intended application. Actual network performance will depend on a
variety of factors including network congestion, application loads, and network settings. For information on
optimizing network throughput, see Optimizing network throughput for Windows and Linux. To achieve the
expected network performance on Linux or Windows, it may be necessary to select a specific version or
optimize your VM. For more information, see How to reliably test for virtual machine throughput.
Supported operating systems and drivers

To take advantage of the GPU capabilities of Azure N -series VMs running Windows, NVIDIA GPU drivers must
be installed. The NVIDIA GPU Driver Extension installs appropriate NVIDIA CUDA or GRID drivers on an N -
series VM. Install or manage the extension using the Azure portal or tools such as Azure PowerShell or Azure
Resource Manager templates. See the NVIDIA GPU Driver Extension documentation for supported operating
systems and deployment steps. For general information about VM extensions, see Azure virtual machine
extensions and features.
If you choose to install NVIDIA GPU drivers manually, see N -series GPU driver setup for Windows for
supported operating systems, drivers, and installation and verification steps.
For availability of N -series VMs, see Products available by region.
N -series VMs can only be deployed in the Resource Manager deployment model.
N -series VMs differ in the type of Azure Storage they support for their disks. NC and NV VMs only
support VM disks that are backed by Standard Disk Storage (HDD ). NCv2, NCv3, ND, and NVv2 VMs
only support VM disks that are backed by Premium Disk Storage (SSD ).
If you want to deploy more than a few N -series VMs, consider a pay-as-you-go subscription or other
purchase options. If you're using an Azure free account, you can use only a limited number of Azure
compute cores.
You might need to increase the cores quota (per region) in your Azure subscription, and increase the
separate quota for NC, NCv2, NCv3, ND, NV, or NVv2 cores. To request a quota increase, open an
online customer support request at no charge. Default limits may vary depending on your subscription
category.
Other sizes
General purpose
Compute optimized
Memory optimized
Storage optimized
Next steps
SKUs.
Install NVIDIA GPU drivers on N-series VMs running
Windows
To take advantage of the GPU capabilities of Azure N -series VMs running Windows, NVIDIA GPU drivers must be
installed. The NVIDIA GPU Driver Extension installs appropriate NVIDIA CUDA or GRID drivers on an N -series
VM. Install or manage the extension using the Azure portal or tools such as Azure PowerShell or Azure Resource
Manager templates. See the NVIDIA GPU Driver Extension documentation for supported operating systems and
deployment steps.
If you choose to install GPU drivers manually, this article provides supported operating systems, drivers, and
installation and verification steps. Manual driver setup information is also available for Linux VMs.
For basic specs, storage capacities, and disk details, see GPU Windows VM sizes.
Supported operating systems and drivers

NVIDIA Tesla (CUDA ) drivers
NVIDIA Tesla (CUDA) drivers for NC, NCv2, NCv3, and ND -series VMs (optional for NV -series) are supported
only on the operating systems listed in the following table. Driver download links are current at time of publication.
For the latest drivers, visit the NVIDIA website.
TIP
As an alternative to manual CUDA driver installation on a Windows Server VM, you can deploy an Azure Data Science Virtual
Machine image. The DSVM editions for Windows Server 2016 pre-install NVIDIA CUDA drivers, the CUDA Deep Neural
Network Library, and other tools.
OS DRIVER
Windows Server 2016 398.75 (.exe)
Windows Server 2012 R2 398.75 (.exe)
NVIDIA GRID drivers

Microsoft redistributes NVIDIA GRID driver installers for NV amd NVv2-series VMs used as virtual workstations
or for virtual applications. Install only these GRID drivers on Azure NV VMs, only on the operating systems listed
in the following table. These drivers include licensing for GRID Virtual GPU Software in Azure.
OS DRIVER
Windows Server 2016 GRID 6.2 (391.81) (.exe)
Windows 10
Windows Server 2012 R2 GRID 6.2 (391.81) (.exe)
Driver installation
1. Connect by Remote Desktop to each N -series VM.
2. Download, extract, and install the supported driver for your Windows operating system.
After GRID driver installation on a VM, a restart is required. After CUDA driver installation, a restart is not
required.
Verify driver installation

You can verify driver installation in Device Manager. The following example shows successful configuration of the
Tesla K80 card on an Azure NC VM.
To query the GPU device state, run the nvidia-smi command-line utility installed with the driver.
1. Open a command prompt and change to the C:\Program Files\NVIDIA Corporation\NVSMI directory.
2. Run nvidia-smi . If the driver is installed you will see output similar to the following. Note that GPU -Util
shows 0% unless you are currently running a GPU workload on the VM. Your driver version and GPU
details may be different from the ones shown.
RDMA network connectivity
RDMA network connectivity can be enabled on RDMA-capable N -series VMs such as NC24r deployed in the
same availability set or in a single placement group in a VM scale set. The HpcVmDrivers extension must be added
to install Windows network device drivers that enable RDMA connectivity. To add the VM extension to an RDMA-
enabled N -series VM, use Azure PowerShell cmdlets for Azure Resource Manager.
To install the latest version 1.1 HpcVMDrivers extension on an existing RDMA-capable VM named myVM in the
West US region:
Set-AzureRmVMExtension -ResourceGroupName "myResourceGroup" -Location "westus" -VMName "myVM" -ExtensionName

"HpcVmDrivers" -Publisher "Microsoft.HpcCompute" -Type "HpcVmDrivers" -TypeHandlerVersion "1.1"
For more information, see Virtual machine extensions and features for Windows.
The RDMA network supports Message Passing Interface (MPI) traffic for applications running with Microsoft MPI
or Intel MPI 5.x.
Next steps
Developers building GPU -accelerated applications for the NVIDIA Tesla GPUs can also download and install
the latest CUDA Toolkit. For more information, see the CUDA Installation Guide.
High performance compute VM sizes
Azure H-series virtual machines are the latest in high performance computing VMs aimed at handling
workloads like batch processing, analytics, molecular modeling, and fluid dynamics. These 8 and 16 vCPU VMs
are built on the Intel Haswell E5-2667 V3 processor technology featuring DDR4 memory and SSD -based
temporary storage.
In addition to the substantial CPU power, the H-series offers diverse options for low latency RDMA networking
using FDR InfiniBand and several memory configurations to support memory intensive computational
requirements.
H-series
ACU: 290-300
TEMP MAX DISK

STORAGE MAX DATA THROUGHPUT:
SIZE VCPU MEMORY: GIB (SSD) GIB DISKS IOPS MAX NICS
Standard_H8 8 56 1000 32 32 x 500 2
Standard_H1 16 112 2000 64 64 x 500 4

6
Standard_H8 8 112 1000 32 32 x 500 2

m
Standard_H1 16 224 2000 64 64 x 500 4

6m
Standard_H1 16 112 2000 64 64 x 500 4

6r 1
Standard_H1 16 224 2000 64 64 x 500 4

6mr 1
1 For MPI applications, dedicated RDMA backend network is enabled by FDR InfiniBand network, which
delivers ultra-low -latency and high bandwidth.

Storage capacity is shown in units of GiB or 1024^3 bytes. When comparing disks measured in GB
(1000^3 bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may
appear smaller. For example, 1023 GiB = 1098.4 GB
bytes/sec.
Data disks can operate in cached or uncached modes. For cached data disk operation, the host cache mode
is set to ReadOnly or ReadWrite. For uncached data disk operation, the host cache mode is set to None.
per vCPU.
NICs, for all destinations. Upper limits are not guaranteed, but are intended to provide guidance for
selecting the right VM type for the intended application. Actual network performance will depend on a
variety of factors including network congestion, application loads, and network settings. For information on
optimizing network throughput, see Optimizing network throughput for Windows and Linux. To achieve the
expected network performance on Linux or Windows, it may be necessary to select a specific version or
optimize your VM. For more information, see How to reliably test for virtual machine throughput.
Azure subscription – To deploy more than a few compute-intensive instances, consider a pay-as-you-
go subscription or other purchase options. If you're using an Azure free account, you can use only a
limited number of Azure compute cores.
Pricing and availability - These VM sizes are offered only in the Standard pricing tier. Check Products
available by region for availability in Azure regions.
Cores quota – You might need to increase the cores quota in your Azure subscription from the default
value. Your subscription might also limit the number of cores you can deploy in certain VM size families,
including the H-series. To request a quota increase, open an online customer support request at no
charge. (Default limits may vary depending on your subscription category.)
NOTE
Contact Azure Support if you have large-scale capacity needs. Azure quotas are credit limits, not capacity
guarantees. Regardless of your quota, you are only charged for cores that you use.
Virtual network – An Azure virtual network is not required to use the compute-intensive instances.
However, for many deployments you need at least a cloud-based Azure virtual network, or a site-to-site
connection if you need to access on-premises resources. When needed, create a new virtual network to
deploy the instances. Adding compute-intensive VMs to a virtual network in an affinity group is not
supported.
Resizing – Because of their specialized hardware, you can only resize compute-intensive instances within
the same size family (H-series or compute-intensive A-series). For example, you can only resize an H-series
VM from one H-series size to another. In addition, resizing from a non-compute-intensive size to a
compute-intensive size is not supported.
RDMA-capable instances
A subset of the compute-intensive instances (H16r, H16mr, A8, and A9) feature a network interface for remote
direct memory access (RDMA) connectivity. (Selected N -series sizes designated with 'r' such as NC24r are also
RDMA-capable.) This interface is in addition to the standard Azure network interface available to other VM
sizes.
This interface allows the RDMA-capable instances to communicate over an InfiniBand (IB ) network, operating
at FDR rates for H16r, H16mr, and RDMA-capable N -series virtual machines, and QDR rates for A8 and A9
virtual machines. These RDMA capabilities can boost the scalability and performance of certain Message
Passing Interface (MPI) applications.
NOTE
In Azure, IP over IB is not supported. Only RDMA over IB is supported.
Operating system - Windows Server 2016, Windows Server 2012 R2, Windows Server 2012
MPI - Microsoft MPI (MS -MPI) 2012 R2 or later, Intel MPI Library 5.x
Supported MPI implementations use the Microsoft Network Direct interface to communicate between
instances.
RDMA network address space - The RDMA network in Azure reserves the address space
172.16.0.0/16. To run MPI applications on instances deployed in an Azure virtual network, make sure
that the virtual network address space does not overlap the RDMA network.
HpcVmDrivers VM extension - On RDMA-capable VMs, add the HpcVmDrivers extension to install
Windows network device drivers for RDMA connectivity. (In certain deployments of A8 and A9
instances, the HpcVmDrivers extension is added automatically.) To add the VM extension to a VM, you
can use Azure PowerShell cmdlets.
The following command installs the latest version 1.1 HpcVMDrivers extension on an existing RDMA-capable
VM named myVM deployed in the resource group named myResourceGroup in the West US region:
Set-AzureRmVMExtension -ResourceGroupName "myResourceGroup" -Location "westus" -VMName "myVM" -

ExtensionName "HpcVmDrivers" -Publisher "Microsoft.HpcCompute" -Type "HpcVmDrivers" -TypeHandlerVersion
"1.1"
For more information, see Virtual machine extensions and features. You can also work with extensions for VMs
deployed in the classic deployment model.
Cluster configuration options
Azure provides several options to create clusters of Windows HPC VMs that can communicate using the
RDMA network, including:
Virtual machines - Deploy the RDMA-capable HPC VMs in the same availability set (when you use the
Azure Resource Manager deployment model). If you use the classic deployment model, deploy the VMs
in the same cloud service.
Virtual machine scale sets - In a VM scale set, ensure that you limit the deployment to a single
placement group. For example, in a Resource Manager template, set the singlePlacementGroup property
to true .
Azure CycleCloud - Create an HPC cluster in Azure CycleCloud to run MPI jobs on Windows nodes.
Azure Batch - Create an Azure Batch pool to run MPI workloads on Windows Server compute nodes.
For more information, see Use RDMA-capable or GPU -enabled instances in Batch pools. Also see the
Batch Shipyard project, for running container-based workloads on Batch.
Microsoft HPC Pack - HPC Pack includes a runtime environment for MS -MPI that uses the Azure
RDMA network when deployed on RDMA-capable Windows VMs. For example deployments, see Set
up a Windows RDMA cluster with HPC Pack to run MPI applications.
Other sizes
General purpose
Compute optimized
Memory optimized
Storage optimized
GPU optimized
Next steps
For checklists to use the compute-intensive instances with HPC Pack on Windows Server, see Set up a
Windows RDMA cluster with HPC Pack to run MPI applications.
To use compute-intensive instances when running MPI applications with Azure Batch, see Use multi-
instance tasks to run Message Passing Interface (MPI) applications in Azure Batch.
Learn more about how Azure compute units (ACU ) can help you compare compute performance across
Azure SKUs.
Previous generations of virtual machine sizes
This article provides information on previous generations of virtual machine sizes. These sizes can still be used,
but there are newer generations available.
DS-series
ACU: 160
MAX
CACHED
AND TEMP
STORAGE MAX
MEMORY: STORAGE MAX DATA (CACHE SIZE UT: IOPS / BANDWIDTH
SIZE VCPU GIB (SSD) GIB DISKS IN GIB) MBPS (MBPS)
Standard_D 1 3.5 7 4 4,000 / 32 3,200 / 32 2 / 500

S1 (43)
Standard_D 2 7 14 8 8,000 / 64 6,400 / 64 2 / 1000

S2 (86)
Standard_D 4 14 28 16 16,000 / 12,800 / 4 / 2000

S3 128 (172) 128
Standard_D 8 28 56 32 32,000 / 25,600 / 8 / 4000

S4 256 (344) 256
DS-series - memory optimized

ACU: 160 1
MAX
CACHED
AND TEMP
STORAGE MAX
MEMORY: STORAGE MAX DATA (CACHE SIZE UT: IOPS / BANDWIDTH
SIZE VCPU GIB (SSD) GIB DISKS IN GIB) MBPS (MBPS)
Standard_D 2 14 28 8 8,000 / 64 6,400 / 64 2 / 1000

S11 (72)
Standard_D 4 28 56 16 16,000 / 12,800 / 4 / 2000

S12 128 (144) 128
Standard_D 8 56 112 32 32,000 / 25,600 / 8 / 4000

S13 256 (288) 256
Standard_D 16 112 224 64 64,000 / 51,200 / 8 / 8000

S14 512 (576) 512
1 The maximum disk throughput (IOPS or MBps) possible with a DS series VM may be limited by the number, size
and striping of the attached disk(s). For details, see Premium Storage: High-performance storage for Azure virtual
machine workloads.
D-series
ACU: 160
MAX TEMP
STORAGE MAX NICS /
Standard_D1 1 3.5 50 3000 / 46 / 4 / 4x500 2 / 500

23
Standard_D2 2 7 100 6000 / 93 / 8 / 8x500 2 / 1000

46
Standard_D3 4 14 200 12000 / 187 / 16 / 16x500 4 / 2000

93
Standard_D4 8 28 400 24000 / 375 / 32 / 32x500 8 / 4000

187
D-series - memory optimized

ACU: 160
MAX TEMP
STORAGE MAX NICS /
Standard_D11 2 14 100 6000 / 93 / 8 / 8x500 2 / 1000

46
Standard_D12 4 28 200 12000 / 187 / 16 / 16x500 4 / 2000

93
Standard_D13 8 56 400 24000 / 375 / 32 / 32x500 8 / 4000

187
Standard_D14 16 112 800 48000 / 750 / 64 / 64x500 8 / 8000

375
A-series - compute-intensive instances

ACU: 225
The A8-A11 and H-series sizes are also known as compute-intensive instances. The hardware that runs these sizes
is designed and optimized for compute-intensive and network-intensive applications, including high-performance
computing (HPC ) cluster applications, modeling, and simulations. The A8-A11 series uses Intel Xeon E5-2670 @
2.6 GHZ and the H-series uses Intel Xeon E5-2667 v3 @ 3.2 GHz. This article provides information about the
number of vCPUs, data disks, and NICs as well as storage throughput and network bandwidth for each size in this
grouping.
MAX DATA
DISK
TEMP STORAGE MAX DATA THROUGHPUT:
SIZE VCPU MEMORY: GIB (HDD): GIB DISKS IOPS MAX NICS
Standard_A8 1 8 56 382 32 32x500 2
Standard_A9 1 16 112 382 64 64x500 4
Standard_A10 8 56 382 32 32x500 2
Standard_A11 16 112 382 64 64x500 4
1For MPI applications, dedicated RDMA backend network is enabled by FDR InfiniBand network, which delivers
ultra-low -latency and high bandwidth.
A-series
ACU: 50-100
MAX NICS /
MAX DATA EXPECTED
DISK NETWORK
TEMP STORAGE MAX DATA THROUGHPUT: BANDWIDTH
SIZE VCPU MEMORY: GIB (HDD): GIB DISKS IOPS (MBPS)
Standard_A0 1 1 0.768 20 1 1x500 2 / 100
Standard_A1 1 1.75 70 2 2x500 2 / 500
Standard_A2 2 3.5 135 4 4x500 2 / 500
Standard_A3 4 7 285 8 8x500 2 / 1000
Standard_A4 8 14 605 16 16x500 4 / 2000
Standard_A5 2 14 135 4 4x500 2 / 500
Standard_A6 4 28 285 8 8x500 2 / 1000
Standard_A7 8 56 605 16 16x500 4 / 2000
1 The A0 size is over-subscribed on the physical hardware. For this specific size only, other customer deployments
may impact the performance of your running workload. The relative performance is outlined below as the
expected baseline, subject to an approximate variability of 15 percent.
Standard A0 - A4 using CLI and PowerShell
In the classic deployment model, some VM size names are slightly different in CLI and PowerShell:
Standard_A0 is ExtraSmall
Standard_A1 is Small
Standard_A2 is Medium
Standard_A3 is Large
Standard_A4 is ExtraLarge
Basic A
The basic tier sizes are primarily for development workloads and other applications that don't require load
balancing, auto-scaling, or memory-intensive virtual machines.
MAX MAX. DATA

SIZE – TEMPORARY DISKS (1023 MAX. IOPS (300
SIZE\NAME VCPU MEMORY NICS (MAX) DISK SIZE GB EACH) PER DISK)
A0\Basic_A0 1 768 MB 2 20 GB 1 1x300

MAX MAX. DATA
SIZE – TEMPORARY DISKS (1023 MAX. IOPS (300
SIZE\NAME VCPU MEMORY NICS (MAX) DISK SIZE GB EACH) PER DISK)
A1\Basic_A1 1 1.75 GB 2 40 GB 2 2x300
A2\Basic_A2 2 3.5 GB 2 60 GB 4 4x300
A3\Basic_A3 4 7 GB 2 120 GB 8 8x300
A4\Basic_A4 8 14 GB 2 240 GB 16 16x300

bytes/sec.
vCPU.
Expected network bandwidth is the maximum aggregated bandwidth allocated per VM type across all NICs,
for all destinations. Upper limits are not guaranteed, but are intended to provide guidance for selecting the
right VM type for the intended application. Actual network performance will depend on a variety of factors
Other sizes
General purpose
Compute optimized
Memory optimized
Storage optimized
GPU
Next steps
SKUs.
Azure compute unit (ACU)
The concept of the Azure Compute Unit (ACU ) provides a way of comparing compute (CPU ) performance
across Azure SKUs. This will help you easily identify which SKU is most likely to satisfy your performance
needs. ACU is currently standardized on a Small (Standard_A1) VM being 100 and all other SKUs then
represent approximately how much faster that SKU can run a standard benchmark.
IMPORTANT
The ACU is only a guideline. The results for your workload may vary.
SKU FAMILY ACU \ VCPU VCPU: CORE
A0 50 1:1
A1-A4 100 1:1
A5-A7 100 1:1
A1_v2-A8_v2 100 1:1
A2m_v2-A8m_v2 100 1:1
A8-A11 225* 1:1
D1-D14 160 1:1
D1_v2-D15_v2 210 - 250* 1:1
DS1-DS14 160 1:1
DS1_v2-DS15_v2 210-250* 1:1
D_v3 160-190* 2:1**
Ds_v3 160-190* 2:1**
E_v3 160-190* 2:1**
Es_v3 160-190* 2:1**
F2s_v2-F72s_v2 195-210* 2:1**
F1-F16 210-250* 1:1
F1s-F16s 210-250* 1:1

SKU FAMILY ACU \ VCPU VCPU: CORE
G1-G5 180 - 240* 1:1
GS1-GS5 180 - 240* 1:1
H 290 - 300* 1:1
L4s-L32s 180 - 240* 1:1
M 160-180 2:1**
ACUs marked with a * use Intel® Turbo technology to increase CPU frequency and provide a performance
boost. The amount of the boost can vary based on the VM size, workload, and other workloads running on the
same host.
**Hyper-threaded and capable of running nested virtualization
Here are links to more information about the different sizes:
General-purpose
Memory optimized
Compute optimized
GPU optimized
Storage optimized
Compute benchmark scores for Windows VMs
The following SPECInt benchmark scores show compute performance for Azure's high-performance VM lineup
running Windows Server. Compute benchmark scores are also available for Linux VMs.
NOTE
The Linux numbers have been recently updated and contain a more comprehensive set of VMs.
A-series - compute-intensive
SIZE VCPUS NUMA NODES CPU RUNS AVG BASE RATE STDDEV
Standard_A8 8 1 Intel Xeon 10 236.1 1.1

CPU E5-2670
0 @ 2.6 GHz

CPU E5-2670
0 @ 2.6 GHz

CPU E5-2670
0 @ 2.6 GHz

CPU E5-2670
0 @ 2.6 GHz
Dv2-series
Standard_D1_ 1 1 Intel Xeon E5- 83 36.6 2.6

v2 2673 v3 @
2.4 GHz

v2 2673 v3 @
2.4 GHz

v2 2673 v3 @
2.4 GHz

v2 2673 v3 @
2.4 GHz

v2 2673 v3 @
2.4 GHz
Standard_D11 2 1 Intel Xeon E5- 19 70.1 3.7

_v2 2673 v3 @
2.4 GHz

_v2 2673 v3 @
2.4 GHz

_v2 2673 v3 @
2.4 GHz

_v2 2673 v3 @
2.4 GHz
G-series, GS-series
Standard_G1, 2 1 Intel Xeon E5- 31 71.8 6.5

Standard_GS1 2698B v3 @ 2
GHz

GHz

GHz

GHz

GHz
H-series
Standard_H8 8 1 Intel Xeon E5- 5 297.4 0.9

2667 v3 @
3.2 GHz

2667 v3 @
3.2 GHz

m 2667 v3 @
3.2 GHz

m 2667 v3 @
3.2 GHz

r 2667 v3 @
3.2 GHz

mr 2667 v3 @
3.2 GHz
About SPECint
Windows numbers were computed by running SPECint 2006 on Windows Server. SPECint was run using the
base rate option (SPECint_rate2006), with one copy per vCPU. SPECint consists of 12 separate tests, each run
three times, taking the median value from each test and weighting them to form a composite score. Those tests
were then run across multiple VMs to provide the average scores shown.
Next steps
For storage capacities, disk details, and additional considerations for choosing among VM sizes, see Sizes for
virtual machines.
Planned maintenance for virtual machines in Azure
Azure periodically performs updates to improve the reliability, performance, and security of the host infrastructure
for virtual machines. These updates range from patching software components in the hosting environment (like
operating system, hypervisor, and various agents deployed on the host), upgrading networking components, to
hardware decommissioning. The majority of these updates are performed without any impact to the hosted virtual
machines. However, there are cases where updates do have an impact:
If a reboot-less update is possible, Azure uses memory preserving maintenance to pause the VM while the
host is updated or the VM is moved to an already updated host altogether.
If maintenance requires a reboot, you get a notice of when the maintenance is planned. In these cases, you'll
also be given a time window where you can start the maintenance yourself, at a time that works for you.
This page describes how Microsoft Azure performs both types of maintenance. For more information about
unplanned events (outages), see Manage the availability of virtual machines for Windows or Linux.
Applications running in a virtual machine can gather information about upcoming updates by using the Azure
Metadata Service for Windows or Linux.
For "how -to" information on managing planned maintenance, see "Handling planned maintenance notifications"
for Linux or Windows.
Memory preserving maintenance

When updates don't require a full reboot, memory preserving maintenance mechanisms are used to limit the
impact to the virtual machine. The virtual machine is paused for up to 30 seconds, preserving the memory in RAM,
while the hosting environment applies the necessary updates and patches, or moves the VM to an already updated
host. The virtual machine is then resumed and the clock of the virtual machine is automatically synchronized.
These non-rebootful maintenance operations are applied fault domain by fault domain, and progress is stopped if
any warning health signals are received.
Some applications may be impacted by these types of updates. Applications that perform real-time event
processing, like media streaming or transcoding, or high throughput networking scenarios, may not be designed to
tolerate a 30 second pause. In case the VM is being moved to a different host, some sensitive workloads might
notice a slight performance degradation in the few minutes leading up to the Virtual Machine pause.
Maintenance requiring a reboot

When VMs need to be rebooted for planned maintenance, you are notified in advance. Planned maintenance has
two phases: the self-service window and a scheduled maintenance window.
The self-service window lets you initiate the maintenance on your VMs. During this time, you can query each VM
to see their status and check the result of your last maintenance request.
When you start self-service maintenance, your VM is moved to a node that has already been updated and then
powers it back on. Because the VM reboots, the temporary disk is lost and dynamic IP addresses associated with
virtual network interface are updated.
If you start self-service maintenance and there is an error during the process, the operation is stopped, the VM is
not updated and you get the option to retry the self-service maintenance.
When the self-service window has passed, the scheduled maintenance window begins. During this time
window, you can still query for the maintenance window, but no longer be able to start the maintenance yourself.
For information on managing maintenance requiring a reboot, see "Handling planned maintenance notifications"
for Linux or Windows.
Availability Considerations during Scheduled Maintenance
If you decide to wait until the scheduled maintenance window, there are a few things to consider for maintaining
the highest availability of your VMs.
Paired Regions
Each Azure region is paired with another region within the same geography, together they make a regional pair.
During scheduled maintenance, Azure will only update the VMs in a single region of a region pair. For example,
when updating the Virtual Machines in North Central US, Azure will not update any Virtual Machines in South
Central US at the same time. However, other regions such as North Europe can be under maintenance at the same
time as East US. Understanding how region pairs work can help you better distribute your VMs across regions. For
more information, see Azure region pairs.
Availability sets and scale sets
When deploying a workload on Azure VMs, you can create the VMs within an availability set to provide high
availability to your application. This ensures that during either an outage or rebootful maintenance events, at least
one virtual machine is available.
Within an availability set, individual VMs are spread across up to 20 update domains (UDs). During scheduled
maintenance, only a single update domain is impacted at any given time. Be aware that the order of update
domains being impacted does not necessarily happen sequentially.
Virtual machine scale sets are an Azure compute resource that enables you to deploy and manage a set of identical
VMs as a single resource. The scale set is automatically deployed across update domains, like VMs in an availability
set. Just like with availability sets, with scale sets only a single update domain is impacted at any given time during
scheduled maintenance.
For more information about configuring your virtual machines for high availability, see Manage the availability of
your virtual machines for Windows or Linux.
Next steps
For information on managing maintenance requiring a reboot, see Handling planned maintenance notifications.
About disks storage for Azure Windows VMs
Just like any other computer, virtual machines in Azure use disks as a place to store an operating system,
applications, and data. All Azure virtual machines have at least two disks – a Windows operating system disk and a
temporary disk. The operating system disk is created from an image, and both the operating system disk and the
image are virtual hard disks (VHDs) stored in an Azure storage account. Virtual machines also can have one or
more data disks, that are also stored as VHDs.
In this article, we will talk about the different uses for the disks, and then discuss the different types of disks you
can create and use. This article is also available for Linux virtual machines.
NOTE
Azure has two different deployment models for creating and working with resources: Resource Manager and classic. This
article covers using both models, but Microsoft recommends that most new deployments use the Resource Manager model.
Disks used by VMs

Let's take a look at how the disks are used by the VMs.
Operating system disk
Every virtual machine has one attached operating system disk. It's registered as a SATA drive and labeled as the C:
drive by default. This disk has a maximum capacity of 2048 gigabytes (GB ).
Temporary disk
Each VM contains a temporary disk. The temporary disk provides short-term storage for applications and
processes and is intended to only store data such as page or swap files. Data on the temporary disk may be lost
during a maintenance event or when you redeploy a VM. During a successful standard reboot of the VM, the data
on the temporary drive will persist.
The temporary disk is labeled as the D: drive by default and it used for storing pagefile.sys. To remap this disk to a
different drive letter, see Change the drive letter of the Windows temporary disk. The size of the temporary disk
varies, based on the size of the virtual machine. For more information, see Sizes for Windows virtual machines.
For more information on how Azure uses the temporary disk, see Understanding the temporary drive on
Microsoft Azure Virtual Machines
Data disk
A data disk is a VHD that's attached to a virtual machine to store application data, or other data you need to keep.
Data disks are registered as SCSI drives and are labeled with a letter that you choose. Each data disk has a
maximum capacity of 4,095 GB, managed disks have a maximum capacity of 32,767 TiB. The size of the virtual
machine determines how many data disks you can attach to it and the type of storage you can use to host the
disks.
NOTE
For more information about virtual machines capacities, see Sizes for Windows virtual machines.
Azure creates an operating system disk when you create a virtual machine from an image. If you use an image that
includes data disks, Azure also creates the data disks when it creates the virtual machine. Otherwise, you add data
disks after you create the virtual machine.
You can add data disks to a virtual machine at any time, by attaching the disk to the virtual machine. You can use
a VHD that you've uploaded or copied to your storage account, or use an empty VHD that Azure creates for you.
Attaching a data disk associates the VHD file with the VM by placing a 'lease' on the VHD so it can't be deleted
from storage while it's still attached.
About VHDs
The VHDs used in Azure are .vhd files stored as page blobs in a standard or premium storage account in Azure.
For details about page blobs, see Understanding block blobs and page blobs. For details about premium storage,
see High-performance premium storage and Azure VMs.
Azure supports the fixed disk VHD format. The fixed format lays the logical disk out linearly within the file, so that
disk offset X is stored at blob offset X. A small footer at the end of the blob describes the properties of the VHD.
Often, the fixed-format wastes space because most disks have large unused ranges in them. However, Azure stores
.vhd files in a sparse format, so you receive the benefits of both the fixed and dynamic disks at the same time. For
more information, see Getting started with virtual hard disks .
All VHD files in Azure that you want to use as a source to create disks or images are read-only, except the .vhd files
uploaded or copied to Azure storage by the user (which can be either read-write or read-only). When you create a
disk or image, Azure makes copies of the source .vhd files. These copies can be read-only or read-and-write,
depending on how you use the VHD.
When you create a virtual machine from an image, Azure creates a disk for the virtual machine that is a copy of the
source .vhd file. To protect against accidental deletion, Azure places a lease on any source .vhd file that’s used to
create an image, an operating system disk, or a data disk.
Before you can delete a source .vhd file, you’ll need to remove the lease by deleting the disk or image. To delete a
.vhd file that is being used by a virtual machine as an operating system disk, you can delete the virtual machine,
the operating system disk, and the source .vhd file all at once by deleting the virtual machine and deleting all
associated disks. However, deleting a .vhd file that’s a source for a data disk requires several steps in a set order.
First you detach the disk from the virtual machine, then delete the disk, and then delete the .vhd file.
WARNING
If you delete a source .vhd file from storage, or delete your storage account, Microsoft can't recover that data for you.
Types of disks
Azure Disks are designed for 99.999% availability. Azure Disks have consistently delivered enterprise-grade
durability, with an industry-leading ZERO% Annualized Failure Rate.
There are three performance tiers for storage that you can choose from when creating your disks -- Premium SSD
Disks, Standard SSD, and Standard HDD Storage. Also, there are two types of disks -- unmanaged and managed.
Standard HDD disks
Standard HDD disks are backed by HDDs, and deliver cost-effective storage. Standard HDD storage can be
replicated locally in one datacenter, or be geo-redundant with primary and secondary data centers. For more
information about storage replication, see Azure Storage replication.
For more information about using Standard HDD disks, see Standard Storage and Disks.
Standard SSD disks
Standard SSD disks are designed to address the same kind of workloads as Standard HDD disks, but offer more
consistent performance and reliability than HDD. Standard SSD disks combine elements of Premium SSD disks
and Standard HDD disks to form a cost-effective solution best suited for applications like web servers that do not
need high IOPS on disks. Where available, Standard SSD disks are the recommended deployment option for most
workloads. Standard SSD disks are available as Managed Disks in all regions but are currently only available with
the locally redundant storage (LRS ) resiliency type.
Premium SSD disks
Premium SSD disks are backed by SSDs, and delivers high-performance, low -latency disk support for VMs
running I/O -intensive workloads. Typically you can use Premium SSD disks with sizes that include an "s" in the
series name. For example, there is the Dv3-Series and the Dsv3-series, the Dsv3-series can be used with Premium
SSD disks. For more information, please see Premium Storage.
Unmanaged disks
Unmanaged disks are the traditional type of disks that have been used by VMs. With these disks, you create your
own storage account and specify that storage account when you create the disk. Make sure you don't put too many
disks in the same storage account, because you could exceed the scalability targets of the storage account (20,000
IOPS, for example), resulting in the VMs being throttled. With unmanaged disks, you have to figure out how to
maximize the use of one or more storage accounts to get the best performance out of your VMs.
Managed disks
Managed Disks handles the storage account creation/management in the background for you, and ensures that
you do not have to worry about the scalability limits of the storage account. You simply specify the disk size and
the performance tier (Standard/Premium), and Azure creates and manages the disk for you. As you add disks or
scale the VM up and down, you don't have to worry about the storage being used.
You can also manage your custom images in one storage account per Azure region, and use them to create
hundreds of VMs in the same subscription. For more information about Managed Disks, see the Managed Disks
Overview.
We recommend that you use Azure Managed Disks for new VMs, and that you convert your previous unmanaged
disks to managed disks, to take advantage of the many features available in Managed Disks.
Disk comparison
The following table provides a comparison of Standard HDD, Standard SSD, and Premium SSD for unmanaged
and managed disks to help you decide what to use. Sizes denoted with an asterisk are currently in preview.
AZURE PREMIUM DISK AZURE STANDARD SSD DISK AZURE STANDARD HDD DISK
Disk Type Solid State Drives (SSD) Solid State Drives (SSD) Hard Disk Drives (HDD)
Overview SSD-based high- More consistent HDD-based cost effective

performance, low-latency performance and reliability disk for infrequent access
disk support for VMs than HDD. Optimized for
running IO-intensive low-IOPS workloads
workloads or hosting
mission critical production
environment
Scenario Production and performance Web servers, lightly used Backup, Non-critical,
sensitive workloads enterprise applications and Infrequent access
Dev/Test
AZURE PREMIUM DISK AZURE STANDARD SSD DISK AZURE STANDARD HDD DISK
Disk Size P4: 32 GiB (Managed Disks Managed Disks only: Unmanaged Disks: 1 GiB – 4
only) E10: 128 GiB TiB (4095 GiB)
P6: 64 GiB (Managed Disks E15: 256 GiB
only) E20: 512 GiB Managed Disks:
P10: 128 GiB E30: 1024 GiB S4: 32 GiB
P15: 256 GiB (Managed E40: 2048 GiB S6: 64 GiB
Disks only) E50: 4095 GiB S10: 128 GiB
P20: 512 GiB E60: 8,192 GiB * (8 TiB) S15: 256 GiB
P50: 4,095 GiB S40: 2048 GiB
P60: 8,192 GiB * (8 TiB) S50: 4095 GiB
P70: 16,384 GiB * (16 TiB) S60: 8,192 GiB * (8 TiB)
P80: 32,767 GiB * (32 TiB) S70: 16,384 GiB * (16 TiB)
S80: 32,384 GiB * (32 TiB)
Max Throughput per Disk P4: 25 MiB/s E10-E50: Up to 60 MiB/s S4 - S50: Upt o 60 MiB/s
P6: 50 MiB/s E60: Up to 300 MiB/s * S60: Up to 300 MiB/s *
P10: 100 MiB/s E70-E80: 500 MiB/s * S70-S80: Up to 500 MiB/s *
P15: 200 MiB/s
P20: 150 MiB/s
P30: 200 MiB/s
P40-P50: 250 MiB/s
P60: 480 MiB/s
P70-P80: 750 MiB/s
Max IOPS per Disk P4: 120 IOPS E10-E50: Up to 500 IOPS S4-S50: Up to 500 IOPS
P6: 240 IOPS E60: Up to 1300 IOPS * S60: Up to 1300 IOPS *
P10: 500 IOPS E70-E80: Up to 2000 IOPS * S70-S80: Up to 2000 IOPS *
P15: 1100 IOPS
P20: 2300 IOPS
P30: 5000 IOPS
P40-P50: 7500 IOPS
P60: 12,500 IOPS *
P70: 15,000 IOPS *
P80: 20,000 IOPS *
One last recommendation: Use TRIM with unmanaged standard disks

If you use unmanaged standard disks (HDD ), you should enable TRIM. TRIM discards unused blocks on the disk
so you are only billed for storage that you are actually using. This can save on costs if you create large files and
then delete them.
You can run this command to check the TRIM setting. Open a command prompt on your Windows VM and type:
fsutil behavior query DisableDeleteNotify
If the command returns 0, TRIM is enabled correctly. If it returns 1, run the following command to enable TRIM:
fsutil behavior set DisableDeleteNotify 0

NOTE
Note: Trim support starts with Windows Server 2012 / Windows 8 and above, see New API allows apps to send "TRIM and
Unmap" hints to storage media.
Next steps
Attach a disk to add additional storage for your VM.
Create a snapshot.
Convert to managed disks.
Azure Managed Disks Overview
Azure Managed Disks simplifies disk management for Azure IaaS VMs by managing the storage accounts
associated with the VM disks. You only have to specify the type (Standard HDD, Standard SSD, or Premium SSD )
and the size of disk you need, and Azure creates and manages the disk for you.
Benefits of managed disks

Let's take a look at some of the benefits you gain by using managed disks, starting with this Channel 9 video,
Better Azure VM Resiliency with Managed Disks.
Simple and scalable VM deployment

Managed Disks handles storage for you behind the scenes. Previously, you had to create storage accounts to hold
the disks (VHD files) for your Azure VMs. When scaling up, you had to make sure you created additional storage
accounts so you didn't exceed the IOPS limit for storage with any of your disks. With Managed Disks handling
storage, you are no longer limited by the storage account limits (such as 20,000 IOPS / account). You also no
longer have to copy your custom images (VHD files) to multiple storage accounts. You can manage them in a
central location – one storage account per Azure region – and use them to create hundreds of VMs in a
subscription.
Managed Disks will allow you to create up to 50,000 VM disks of a type in a subscription per region, which will
enable you to create thousands of VMs in a single subscription. This feature also further increases the scalability
of Virtual Machine Scale Sets by allowing you to create up to a thousand VMs in a virtual machine scale set using
a Marketplace image.
Better reliability for Availability Sets
Managed Disks provides better reliability for Availability Sets by ensuring that the disks of VMs in an Availability
Set are sufficiently isolated from each other to avoid single points of failure. Disks are automatically placed in
different storage scale units (stamps). If a stamp fails due to hardware or software failure, only the VM instances
with disks on those stamps fail. For example, let's say you have an application running on five VMs, and the VMs
are in an Availability Set. The disks for those VMs won't all be stored in the same stamp, so if one stamp goes
down, the other instances of the application continue to run.
Highly durable and available
Azure Disks are designed for 99.999% availability. Rest easier knowing that you have three replicas of your data
that enables high durability. If one or even two replicas experience issues, the remaining replicas help ensure
persistence of your data and high tolerance against failures. This architecture has helped Azure consistently
deliver enterprise-grade durability for IaaS disks, with an industry-leading ZERO% Annualized Failure Rate.
Granular access control
You can use Azure Role-Based Access Control (RBAC ) to assign specific permissions for a managed disk to one
or more users. Managed Disks exposes a variety of operations, including read, write (create/update), delete, and
retrieving a shared access signature (SAS ) URI for the disk. You can grant access to only the operations a person
needs to perform their job. For example, if you don't want a person to copy a managed disk to a storage account,
you can choose not to grant access to the export action for that managed disk. Similarly, if you don't want a
person to use an SAS URI to copy a managed disk, you can choose not to grant that permission to the managed
disk.
Azure Backup service support
Use Azure Backup service with Managed Disks to create a backup job with time-based backups, easy VM
restoration, and backup retention policies. Managed Disks only support Locally Redundant Storage (LRS ) as the
replication option. Three copies of the data are kept within a single region. For regional disaster recovery, you
must back up your VM disks in a different region using Azure Backup service and a GRS storage account as
backup vault. Currently Azure Backup supports the disk sizes up to 4TB disks. You need to upgrade VM backup
stack to V2 for support of 4TB disks. For more information, see Using Azure Backup service for VMs with
Managed Disks.
Pricing and Billing

When using Managed Disks, the following billing considerations apply:
Storage Type
Disk Size
Number of transactions
Outbound data transfers
Managed Disk Snapshots (full disk copy)
Let's take a closer look at these options.
Storage Type: Managed Disks offers 3 performance tiers: Standard HDD, Standard SSD, and Premium. The
billing of a managed disk depends on which type of storage you have selected for the disk.
Disk Size: Billing for managed disks depends on the provisioned size of the disk. Azure maps the provisioned
size (rounded up) to the nearest Managed Disks option as specified in the tables below. Each managed disk maps
to one of the supported provisioned sizes and is billed accordingly. For example, if you create a standard
managed disk and specify a provisioned size of 200 GB, you are billed as per the pricing of the S15 Disk type.
Here are the disk sizes available for a premium managed disk:
PREMI
UM
HDD
MANA
GED
DISK
TYPE P4 P6 P10 P15 P20 P30 P40 P50 P60 P70 P80
Disk 32 64 128 256 512 1,024 2,048 4,095 8,192 16,38 32,76
Size GiB GiB GiB GiB GiB GiB (1 GiB (2 GiB (4 GiB (8 4 GiB 7 GiB
TiB) TiB) TiB) TiB) (16 (TiB)
TiB)
Here are the disk sizes available for a standard SSD managed disk:
STANDA
RD SSD
MANAGE
D
DISK
TYPE E10 E15 E20 E30 E40 E50 E60 E70 E80
Disk Size 128 GiB 256 GiB 512 GiB 1,024 2,048 4,095 8,192 16,384 32,767
GiB (1 GiB (2 GiB (4 GiB (8 GiB (16 GiB (TiB)
TiB) TiB) TiB) TiB) TiB)
Here are the disk sizes available for a standard HDD managed disk:
STAN
DARD
HDD
MANA
GED
DISK
TYPE S4 S6 S10 S15 S20 S30 S40 S50 S60 S70 S80
Disk 32 64 128 256 512 1,024 2,048 4,095 8,192 16,38 32,76
TiB) TiB) TiB) TiB) (16 (TiB)
TiB)
Number of transactions: You are billed for the number of transactions that you perform on a standard
managed disk.
Standard SSD Disks use IO Unit size of 256KB. If the data being transferred is less than 256 KB, it is considered 1
I/O unit. Larger I/O sizes are counted as multiple I/Os of size 256 KB. For example, a 1,100 KB I/O is counted as
five I/O units.
There is no cost for transactions for a premium managed disk.
Outbound data transfers: Outbound data transfers (data going out of Azure data centers) incur billing for
bandwidth usage.
For detailed information on pricing for Managed Disks, see Managed Disks Pricing.
Managed Disk Snapshots

A Managed Snapshot is a read-only full copy of a managed disk that is stored as a standard managed disk by
default. With snapshots, you can back up your managed disks at any point in time. These snapshots exist
independent of the source disk and can be used to create new Managed Disks. They are billed based on the used
size. For example, if you create a snapshot of a managed disk with provisioned capacity of 64 GiB and actual used
data size of 10 GiB, snapshot will be billed only for the used data size of 10 GiB.
Incremental snapshots are currently not supported for Managed Disks.
To learn more about how to create snapshots with Managed Disks, see the following resources:
Create copy of VHD stored as a Managed Disk using Snapshots in Windows
Create copy of VHD stored as a Managed Disk using Snapshots in Linux
Images
Managed Disks also support creating a managed custom image. You can create an image from your custom VHD
in a storage account or directly from a generalized (sys-prepped) VM. This process captures in a single image all
managed disks associated with a VM, including both the OS and data disks. This managed custom image enables
creating hundreds of VMs using your custom image without the need to copy or manage any storage accounts.
For information on creating images, see the following articles:
How to capture a managed image of a generalized VM in Azure
How to generalize and capture a Linux virtual machine using the Azure CLI
Images versus snapshots

You often see the word "image" used with VMs, and now you see "snapshots" as well. It's important to
understand the difference between these terms. With Managed Disks, you can take an image of a generalized VM
that has been deallocated. This image will include all of the disks attached to the VM. You can use this image to
create a new VM, and it will include all of the disks.
A snapshot is a copy of a disk at the point in time it is taken. It only applies to one disk. If you have a VM that only
has one disk (the OS ), you can take a snapshot or an image of it and create a VM from either the snapshot or the
image.
What if a VM has five disks and they are striped? You could take a snapshot of each of the disks, but there is no
awareness within the VM of the state of the disks – the snapshots only know about that one disk. In this case, the
snapshots would need to be coordinated with each other, and that is not currently supported.
Managed Disks and Encryption

There are two kinds of encryption to discuss in reference to managed disks. The first one is Storage Service
Encryption (SSE ), which is performed by the storage service. The second one is Azure Disk Encryption, which you
can enable on the OS and data disks for your VMs.
Storage Service Encryption (SSE)
Azure Storage Service Encryption provides encryption-at-rest and safeguard your data to meet your
organizational security and compliance commitments. SSE is enabled by default for all Managed Disks,
Snapshots, and Images in all the regions where managed disks are available. Starting June 10th, 2017, all new
managed disks/snapshots/images and new data written to existing managed disks are automatically encrypted-
at-rest with keys managed by Microsoft by default. Visit the Managed Disks FAQ page for more details.
Azure Disk Encryption (ADE)
Azure Disk Encryption allows you to encrypt the OS and Data disks used by an IaaS Virtual Machine. This
encryption includes managed disks. For Windows, the drives are encrypted using industry-standard BitLocker
encryption technology. For Linux, the disks are encrypted using the DM -Crypt technology. The encryption
process is integrated with Azure Key Vault to allow you to control and manage the disk encryption keys. For more
information, see Azure Disk Encryption for Windows and Linux IaaS VMs.
Next steps
For more information about Managed Disks, please refer to the following articles.
Get started with Managed Disks
Create a VM using Resource Manager and PowerShell
Create a Linux VM using the Azure CLI
Attach a managed data disk to a Windows VM using PowerShell
Add a managed disk to a Linux VM
Managed Disks PowerShell Sample Scripts
Use Managed Disks in Azure Resource Manager templates
Compare Managed Disks storage options
Premium SSD disks
Standard SSD and HDD disks
Operational guidance
Migrate from AWS and other platforms to Managed Disks in Azure
Convert Azure VMs to managed disks in Azure
High-performance Premium Storage and managed
disks for VMs
Azure Premium Storage delivers high-performance, low -latency disk support for virtual machines (VMs) with
input/output (I/O )-intensive workloads. VM disks that use Premium Storage store data on solid-state drives
(SSDs). To take advantage of the speed and performance of premium storage disks, you can migrate existing
VM disks to Premium Storage.
In Azure, you can attach several premium storage disks to a VM. Using multiple disks gives your applications
up to 256 TB of storage per VM, if you use the preview sizes your application can have up to around 2 PiB of
storage per VM. With Premium Storage, your applications can achieve 80,000 I/O operations per second
(IOPS ) per VM, and a disk throughput of up to 2,000 megabytes per second (MB/s) per VM. Read operations
give you very low latencies.
With Premium Storage, Azure offers the ability to truly lift-and-shift demanding enterprise applications like
Dynamics AX, Dynamics CRM, Exchange Server, SAP Business Suite, and SharePoint farms to the cloud. You
can run performance-intensive database workloads in applications like SQL Server, Oracle, MongoDB, MySQL,
and Redis, which require consistent high performance and low latency.
NOTE
For the best performance for your application, we recommend that you migrate any VM disk that requires high IOPS to
Premium Storage. If your disk does not require high IOPS, you can help limit costs by keeping it in standard Azure
Storage. In standard storage, VM disk data is stored on hard disk drives (HDDs) instead of on SSDs.
Azure offers two ways to create premium storage disks for VMs:
Unmanaged disks
The original method is to use unmanaged disks. In an unmanaged disk, you manage the storage
accounts that you use to store the virtual hard disk (VHD ) files that correspond to your VM disks. VHD
files are stored as page blobs in Azure storage accounts.
Managed disks
When you choose Azure Managed Disks, Azure manages the storage accounts that you use for your VM
disks. You specify the disk type (Premium or Standard) and the size of the disk that you need. Azure
creates and manages the disk for you. You don't have to worry about placing the disks in multiple
storage accounts to ensure that you stay within scalability limits for your storage accounts. Azure
handles that for you.
We recommend that you choose managed disks, to take advantage of their many features.
To get started with Premium Storage, create your free Azure account.
For information about migrating your existing VMs to Premium Storage, see Convert a Windows VM from
unmanaged disks to managed disks or Convert a Linux VM from unmanaged disks to managed disks.
NOTE
Premium Storage is available in most regions. For the list of available regions, see the row for Disk Storage in Azure
products available by region.
Features
Here are some of the features of Premium Storage:
Premium storage disks
Premium Storage supports VM disks that can be attached to specific size-series VMs. Premium Storage
supports a wide variety of Azure VMs. You have a choice of eight GA disk sizes: P4 (32 GiB ), P6 (64 GiB ),
P10 (128 GiB ), P15 (256 GiB ), P20 (512 GiB ), P30 (1,024 GiB ), P40 (2,048 GiB ), P50 (4,095 GiB ). As well
as three preview disk sizes: P60 8,192 GiB (8 TiB ), P70 16,348 GiB (16 TiB ), P80 32,767 GiB (32 TiB ). P4,
P6, P15, P60, P70, and P80 disk sizes are currently only supported for Managed Disks. Each disk size
has its own performance specifications. Depending on your application requirements, you can attach one
or more disks to your VM. We describe the specifications in more detail in Premium Storage scalability
and performance targets.
Premium page blobs
Premium Storage supports page blobs. Use page blobs to store persistent, unmanaged disks for VMs in
Premium Storage. Unlike standard Azure Storage, Premium Storage does not support block blobs,
append blobs, files, tables, or queues. Premium page blobs support six sizes from P10 to P50, and P60
(8191GiB ). P60 Premium page blob is not supported to be attached as VM disks.
Any object placed in a premium storage account will be a page blob. The page blob snaps to one of the
supported provisioned sizes. This is why a premium storage account is not intended to be used to store
tiny blobs.
Premium storage account
To start using Premium Storage, create a premium storage account for unmanaged disks. In the Azure
portal, to create a premium storage account, choose the Premium performance tier. Select the Locally-
redundant storage (LRS ) replication option. You also can create a premium storage account by setting
the performance tier to Premium_LRS. To change the performance tier, use one of the following
approaches:
PowerShell for Azure Storage
Azure CLI for Azure Storage
Azure Storage Resource Provider REST API (for Azure Resource Manager deployments) or one
of the Azure Storage resource provider client libraries
To learn about premium storage account limits, see Premium Storage scalability and performance
targets.
Premium locally redundant storage
A premium storage account supports only locally redundant storage as the replication option. Locally
redundant storage keeps three copies of the data within a single region. For regional disaster recovery,
you must back up your VM disks in a different region by using Azure Backup. You also must use a geo-
redundant storage (GRS ) account as the backup vault.
Azure uses your storage account as a container for your unmanaged disks. When you create an Azure
VM that supports Premium Storage with unmanaged disks, and you select a premium storage account,
your operating system and data disks are stored in that storage account.
Supported VMs
Premium Storage is supported on a wide variety of Azure VMs. You can use standard and premium storage
disks with these VM types. You cannot use premium storage disks with VM series that are not Premium
Storage-compatible.
For information about VM types and sizes in Azure for Windows, see Windows VM sizes. For information
about VM types and sizes in Azure for Linux, see Linux VM sizes.
These are some of the features supported on premium storage enabled VMs:
Availability Set
Using the example of a DS -series VMs, you can add a DS -series VM to a cloud service that has only DS -
series VMs. Do not add DS -series VMs to an existing cloud service that has any type other than DS -
series VMs. You can migrate your existing VHDs to a new cloud service that runs only DS -series VMs. If
you want to use the same virtual IP address for the new cloud service that hosts your DS -series VMs,
use reserved IP addresses.
Operating system disk
You can set up your Premium Storage VM to use either a premium or a standard operating system disk.
For the best experience, we recommend using a Premium Storage-based operating system disk.
Data disks
You can use premium and standard disks in the same Premium Storage VM. With Premium Storage,
you can provision a VM and attach several persistent data disks to the VM. If needed, to increase the
capacity and performance of the volume, you can stripe across your disks.
NOTE
If you stripe premium storage data disks by using Storage Spaces, set up Storage Spaces with 1 column for each
disk that you use. Otherwise, overall performance of the striped volume might be lower than expected because of
uneven distribution of traffic across the disks. By default, in Server Manager, you can set up columns for up to 8
disks. If you have more than 8 disks, use PowerShell to create the volume. Specify the number of columns
manually. Otherwise, the Server Manager UI continues to use 8 columns, even if you have more disks. For
example, if you have 32 disks in a single stripe set, specify 32 columns. To specify the number of columns the
virtual disk uses, in the New-VirtualDisk PowerShell cmdlet, use the NumberOfColumns parameter. For more
information, see Storage Spaces Overview and Storage Spaces FAQs.
Cache
Virtual Machines (VMs) that support Premium Storage have a unique caching capability for higher
levels of throughput and reduced latency. Their caching capability exceeds underlying premium storage
disk performance. Not all VMs support caching however, so please review the VM specifications for the
VM sizes you are interested for more information. VMs that do support caching will indicate this in their
spec with a "Max cached and temp storage throughput" measurement. They are also specified directly
under the VM title.
With caching, you can set the disk caching policy on premium storage disks to ReadOnly, ReadWrite,
or None. The default disk caching policy is ReadOnly for all premium data disks, and ReadWrite for
operating system disks. For optimal performance for your application please remember to use the
correct cache setting.
As an example, for read-heavy or read-only data disks, such as SQL Server data files, set the disk
caching policy to ReadOnly. For write-heavy or write-only data disks, such as SQL Server log files, set
the disk caching policy to None.
To learn more about optimizing your design with Premium Storage, see Design for performance with
Premium Storage.
Analytics
To analyze VM performance by using disks in Premium Storage, turn on VM diagnostics in the Azure
portal. For more information, see Azure VM monitoring with Azure Diagnostics Extension.
To see disk performance, use operating system-based tools like Windows Performance Monitor for
Windows VMs and the iostat command for Linux VMs.
VM scale limits and performance
Each Premium Storage-supported VM size has scale limits and performance specifications for IOPS,
bandwidth, and the number of disks that can be attached per VM. When you use premium storage disks
with VMs, make sure that there is sufficient IOPS and bandwidth on your VM to drive disk traffic.
For example, a STANDARD_DS1 VM has a dedicated bandwidth of 32 MB/s for premium storage disk
traffic. A P10 premium storage disk can provide a bandwidth of 100 MB/s. If a P10 premium storage
disk is attached to this VM, it can only go up to 32 MB/s. It cannot use the maximum 100 MB/s that the
P10 disk can provide.
Currently, the largest VM in the DS -series is the Standard_DS15_v2. The Standard_DS15_v2 can
provide up to 960 MB/s across all disks. The largest VM in the GS -series is the Standard_GS5. The
Standard_GS5 can provide up to 2,000 MB/s across all disks.
These limits are for disk traffic only. These limits don't include cache hits and network traffic. A separate
bandwidth is available for VM network traffic. Bandwidth for network traffic is different from the
dedicated bandwidth used by premium storage disks.
For the most up-to-date information about maximum IOPS and throughput (bandwidth) for Premium
Storage-supported VMs, see Windows VM sizes or Linux VM sizes.
For more information about premium storage disks and their IOPS and throughput limits, see the table
in the next section.
Scalability and performance targets

In this section, we describe the scalability and performance targets to consider when you use Premium Storage.
Premium storage accounts have the following scalability targets:
TOTAL BANDWIDTH FOR A LOCALLY REDUNDANT STORAGE

TOTAL ACCOUNT CAPACITY ACCOUNT
Disk capacity: 35 TB Up to 50 gigabits per second for inbound1 + outbound2

Snapshot capacity: 10 TB
1 All data (requests) that are sent to a storage account
2 All data (responses) that are received from a storage account

For more information, see Azure Storage scalability and performance targets.
If you are using premium storage accounts for unmanaged disks and your application exceeds the scalability
targets of a single storage account, you might want to migrate to managed disks. If you don't want to migrate
to managed disks, build your application to use multiple storage accounts. Then, partition your data across
those storage accounts. For example, if you want to attach 51-TB disks across multiple VMs, spread them
across two storage accounts. 35 TB is the limit for a single premium storage account. Make sure that a single
premium storage account never has more than 35 TB of provisioned disks.
Premium Storage disk limits
When you provision a premium storage disk, the size of the disk determines the maximum IOPS and
throughput (bandwidth). Azure offers eight GA types of premium storage disks: P4 (Managed Disks only), P6
(Managed Disks only), P10, P15 (Managed Disks only), P20, P30, P40, and P50. As well as three preview disk
sizes: P60, P70, and P80. Each premium storage disk type has specific limits for IOPS and throughput. Limits
for the disk types are described in the following table:
PREMI
UM
DISKS
Disk 32 64 128 256 512 1024 2048 4095 8192 16,38 32,76
size GiB GiB GiB GiB GiB GiB (1 GiB (2 GiB (4 GiB (8 4 GiB 7 GiB
TiB) TiB) TiB) TiB) (16 (32
TiB) TiB)
IOPS 120 240 500 1100 2300 5000 7500 7500 12,50 15,00 20,00
per 0 0 0
disk
Throu 25 50 100 125 150 200 250 250 480 750 750
ghput MB MB MB MB MB MB MB MB MB MB MB
per per per per per per per per per per per per
disk secon secon secon secon secon secon secon secon secon secon secon
d d d d d d d d d d d
NOTE
Make sure sufficient bandwidth is available on your VM to drive disk traffic, as described in Premium Storage-supported
VMs. Otherwise, your disk throughput and IOPS is constrained to lower values. Maximum throughput and IOPS are
based on the VM limits, not on the disk limits described in the preceding table.
Azure has designed Premium Storage platform to be massively parallel. Designing your application to be multi-threaded
will help to achieve the high performance target offered on the larger disk sizes.
Here are some important things to know about Premium Storage scalability and performance targets:
Provisioned capacity and performance
When you provision a premium storage disk, unlike standard storage, you are guaranteed the capacity,
IOPS, and throughput of that disk. For example, if you create a P50 disk, Azure provisions 4,095-GB
storage capacity, 7,500 IOPS, and 250-MB/s throughput for that disk. Your application can use all or
part of the capacity and performance.
Disk size
Azure maps the disk size (rounded up) to the nearest premium storage disk option, as specified in the
table in the preceding section. For example, a disk size of 100 GB is classified as a P10 option. It can
perform up to 500 IOPS, with up to 100-MB/s throughput. Similarly, a disk of size 400 GB is classified
as a P20. It can perform up to 2,300 IOPS, with 150-MB/s throughput.
NOTE
You can easily increase the size of existing disks. For example, you might want to increase the size of a 30-GB disk
to 128 GB, or even to 1 TB. Or, you might want to convert your P20 disk to a P30 disk because you need more
capacity or more IOPS and throughput.
I/O size
The size of an I/O is 256 KB. If the data being transferred is less than 256 KB, it is considered 1 I/O unit.
Larger I/O sizes are counted as multiple I/Os of size 256 KB. For example, 1,100 KB I/O is counted as 5
I/O units.
Throughput
The throughput limit includes writes to the disk, and it includes disk read operations that aren't served
from the cache. For example, a P10 disk has 100-MB/s throughput per disk. Some examples of valid
throughput for a P10 disk are shown in the following table:
MAX THROUGHPUT PER P10 DISK NON-CACHE READS FROM DISK NON-CACHE WRITES TO DISK
100 MB/s 100 MB/s 0
100 MB/s 0 100 MB/s
100 MB/s 60 MB/s 40 MB/s
Cache hits
Cache hits are not limited by the allocated IOPS or throughput of the disk. For example, when you use a
data disk with a ReadOnly cache setting on a VM that is supported by Premium Storage, reads that are
served from the cache are not subject to the IOPS and throughput caps of the disk. If the workload of a
disk is predominantly reads, you might get very high throughput. The cache is subject to separate IOPS
and throughput limits at the VM level, based on the VM size. DS -series VMs have roughly 4,000 IOPS
and 33-MB/s throughput per core for cache and local SSD I/Os. GS -series VMs have a limit of 5,000
IOPS and 50-MB/s throughput per core for cache and local SSD I/Os.
Throttling
Throttling might occur, if your application IOPS or throughput exceeds the allocated limits for a premium
storage disk. Throttling also might occur if your total disk traffic across all disks on the VM exceeds the disk
bandwidth limit available for the VM. To avoid throttling, we recommend that you limit the number of pending
I/O requests for the disk. Use a limit based on scalability and performance targets for the disk you have
provisioned, and on the disk bandwidth available to the VM.
Your application can achieve the lowest latency when it is designed to avoid throttling. However, if the number
of pending I/O requests for the disk is too small, your application cannot take advantage of the maximum IOPS
and throughput levels that are available to the disk.
The following examples demonstrate how to calculate throttling levels. All calculations are based on an I/O unit
size of 256 KB.
Example 1
Your application has processed 495 I/O units of 16-KB size in one second on a P10 disk. The I/O units are
counted as 495 IOPS. If you try a 2-MB I/O in the same second, the total of I/O units is equal to 495 + 8 IOPS.
This is because 2 MB I/O = 2,048 KB / 256 KB = 8 I/O units, when the I/O unit size is 256 KB. Because the sum
of 495 + 8 exceeds the 500 IOPS limit for the disk, throttling occurs.
Example 2
Your application has processed 400 I/O units of 256-KB size on a P10 disk. The total bandwidth consumed is
(400 × 256) / 1,024 KB = 100 MB/s. A P10 disk has a throughput limit of 100 MB/s. If your application tries to
perform more I/O operations in that second, it is throttled because it exceeds the allocated limit.
Example 3
You have a DS4 VM with two P30 disks attached. Each P30 disk is capable of 200-MB/s throughput. However,
a DS4 VM has a total disk bandwidth capacity of 256 MB/s. You cannot drive both attached disks to the
maximum throughput on this DS4 VM at the same time. To resolve this, you can sustain traffic of 200 MB/s on
one disk and 56 MB/s on the other disk. If the sum of your disk traffic goes over 256 MB/s, disk traffic is
throttled.
NOTE
If your disk traffic mostly consists of small I/O sizes, your application likely will hit the IOPS limit before the throughput
limit. However, if the disk traffic mostly consists of large I/O sizes, your application likely will hit the throughput limit first,
instead of the IOPS limit. You can maximize your application's IOPS and throughput capacity by using optimal I/O sizes.
Also, you can limit the number of pending I/O requests for a disk.
To learn more about designing for high performance by using Premium Storage, see Design for performance
with Premium Storage.
Snapshots and Copy Blob

To the Storage service, the VHD file is a page blob. You can take snapshots of page blobs, and copy them to
another location, such as to a different storage account.
Unmanaged disks
Create incremental snapshots for unmanaged premium disks the same way you use snapshots with standard
storage. Premium Storage supports only locally redundant storage as the replication option. We recommend
that you create snapshots, and then copy the snapshots to a geo-redundant standard storage account. For more
information, see Azure Storage redundancy options.
If a disk is attached to a VM, some API operations on the disk are not permitted. For example, you cannot
perform a Copy Blob operation on that blob if the disk is attached to a VM. Instead, first create a snapshot of
that blob by using the Snapshot Blob REST API. Then, perform the Copy Blob of the snapshot to copy the
attached disk. Alternatively, you can detach the disk, and then perform any necessary operations.
The following limits apply to premium storage blob snapshots:
PREMIUM STORAGE LIMIT VALUE
Maximum number of snapshots per blob 100
Storage account capacity for snapshots 10 TB

(Includes data in snapshots only. Does not include data in
base blob.)
Minimum time between consecutive snapshots 10 minutes
To maintain geo-redundant copies of your snapshots, you can copy snapshots from a premium storage account
to a geo-redundant standard storage account by using AzCopy or Copy Blob. For more information, see
Transfer data with the AzCopy command-line utility and Copy Blob.
For detailed information about performing REST operations against page blobs in a premium storage account,
see Blob service operations with Azure Premium Storage.
Managed disks
A snapshot for a managed disk is a read-only copy of the managed disk. The snapshot is stored as a standard
managed disk. Currently, incremental snapshots are not supported for managed disks. To learn how to take a
snapshot for a managed disk, see Create a copy of a VHD stored as an Azure managed disk by using managed
snapshots in Windows or Create a copy of a VHD stored as an Azure managed disk by using managed
snapshots in Linux.
If a managed disk is attached to a VM, some API operations on the disk are not permitted. For example, you
cannot generate a shared access signature (SAS ) to perform a copy operation while the disk is attached to a
VM. Instead, first create a snapshot of the disk, and then perform the copy of the snapshot. Alternately, you can
detach the disk and then generate an SAS to perform the copy operation.
Premium Storage for Linux VMs

You can use the following information to help you set up your Linux VMs in Premium Storage:
To achieve scalability targets in Premium Storage, for all premium storage disks with cache set to ReadOnly or
None, you must disable "barriers" when you mount the file system. You don't need barriers in this scenario
because the writes to premium storage disks are durable for these cache settings. When the write request
successfully finishes, data has been written to the persistent store. To disable "barriers," use one of the following
methods. Choose the one for your file system:
For reiserFS, to disable barriers, use the barrier=none mount option. (To enable barriers, use
barrier=flush .)
For ext3/ext4, to disable barriers, use the barrier=0 mount option. (To enable barriers, use barrier=1 .)
For XFS, to disable barriers, use the nobarrier mount option. (To enable barriers, use barrier .)
For premium storage disks with cache set to ReadWrite, enable barriers for write durability.
For volume labels to persist after you restart the VM, you must update /etc/fstab with the universally unique
identifier (UUID ) references to the disks. For more information, see Add a managed disk to a Linux VM.
The following Linux distributions have been validated for Azure Premium Storage. For better performance and
stability with Premium Storage, we recommend that you upgrade your VMs to one of these versions, at a
minimum (or to a later version). Some of the versions require the latest Linux Integration Services (LIS ), v4.0,
for Azure. To download and install a distribution, follow the link listed in the following table. We add images to
the list as we complete validation. Note that our validations show that performance varies for each image.
Performance depends on workload characteristics and your image settings. Different images are tuned for
different kinds of workloads.
DISTRIBUTION VERSION SUPPORTED KERNEL DETAILS
Ubuntu 12.04 3.2.0-75.110+ Ubuntu-12_04_5-LTS-

amd64-server-20150119-
en-us-30GB
Ubuntu 14.04 3.13.0-44.73+ Ubuntu-14_04_1-LTS-

amd64-server-20150123-
en-us-30GB
Debian 7.x, 8.x 3.16.7-ckt4-1+

DISTRIBUTION VERSION SUPPORTED KERNEL DETAILS
SUSE SLES 12 3.12.36-38.1+ suse-sles-12-priority-

v20150213
suse-sles-12-v20150213
SUSE SLES 11 SP4 3.0.101-0.63.1+
CoreOS 584.0.0+ 3.18.4+ CoreOS 584.0.0
CentOS 6.5, 6.6, 6.7, 7.0 LIS4 required

See note in the next section
CentOS 7.1+ 3.10.0-229.1.2.el7+ LIS4 recommended

See note in the next section
Red Hat Enterprise Linux 6.8+, 7.2+

(RHEL)
Oracle 6.0+, 7.2+ UEK4 or RHCK
Oracle 7.0-7.1 UEK4 or RHCK w/LIS 4.1+
Oracle 6.4-6.7 UEK4 or RHCK w/LIS 4.1+
LIS drivers for OpenLogic CentOS

If you are running OpenLogic CentOS VMs, run the following command to install the latest drivers:
sudo rpm -e hypervkvpd ## (Might return an error if not installed. That's OK.)
sudo yum install microsoft-hyper-v
To activate the new drivers, restart the VM.
Pricing and billing

When you use Premium Storage, the following billing considerations apply:
Premium storage disk and blob size
Billing for a premium storage disk or blob depends on the provisioned size of the disk or blob. Azure
maps the provisioned size (rounded up) to the nearest premium storage disk option. For details, see the
table in Premium Storage scalability and performance targets. Each disk maps to a supported
provisioned disk size, and is billed accordingly. Billing for any provisioned disk is prorated hourly by
using the monthly price for the Premium Storage offer. For example, if you provisioned a P10 disk and
deleted it after 20 hours, you are billed for the P10 offering prorated to 20 hours. This is regardless of
the amount of actual data written to the disk or the IOPS and throughput used.
Premium unmanaged disks snapshots
Snapshots on premium unmanaged disks are billed for the additional capacity used by the snapshots.
For more information about snapshots, see Create a snapshot of a blob.
Premium managed disks snapshots
A snapshot of a managed disk is a read-only copy of the disk. The disk is stored as a standard managed
disk. A snapshot costs the same as a standard managed disk. For example, if you take a snapshot of a
128-GB premium managed disk, the cost of the snapshot is equivalent to a 128-GB standard managed
disk.
Outbound data transfers (data going out of Azure datacenters) incur billing for bandwidth usage.
For detailed information about pricing for Premium Storage, Premium Storage-supported VMs, and managed
disks, see these articles:
Azure Storage pricing
Virtual Machines pricing
Managed disks pricing
Azure Backup support

For regional disaster recovery, you must back up your VM disks in a different region by using Azure Backup
and a GRS storage account as a backup vault.
To create a backup job with time-based backups, easy VM restoration, and backup retention policies, use Azure
Backup. You can use Backup both with unmanaged and managed disks. For more information, see Azure
Backup for VMs with unmanaged disks and Azure Backup for VMs with managed disks.
Next steps
For more information about Premium Storage, see the following articles.
Design and implement with Premium Storage
Design for performance with Premium Storage
Blob storage operations with Premium Storage
Operational guidance
Migrate to Azure Premium Storage
Blog posts
Azure Premium Storage generally available
Announcing the GS -series: Adding Premium Storage support to the largest VMs in the public cloud
Standard SSD Managed Disks for Azure Virtual
machine workloads
Azure Standard Solid State Drives (SSD ) Managed Disks are a cost-effective storage option optimized for
workloads that need consistent performance at lower IOPS levels. Standard SSD offers a good entry level
experience for those who wish to move to the cloud, especially if you experience issues with the variance of
workloads running on your HDD solutions on premises. Standard SSDs deliver better availability, consistency,
reliability and latency compared to HDD Disks, and are suitable for Web servers, low IOPS application servers,
lightly used enterprise applications, and Dev/Test workloads.
Standard SSD Features

Managed Disks: Standard SSDs are only available as Managed Disks. Unmanaged Disks and Page Blobs are not
supported on Standard SSD. While creating the Managed Disk, you specify the disk type as Standard SSD and
indicate the size of disk you need, and Azure creates and manages the disk for you. Standard SSDs support all
service management operations offered by Managed Disks. For example, you can create, copy or snapshot
Standard SSD Managed Disks in the same way you do with Managed Disks.
Virtual Machines: Standard SSDs can be used with all Azure VMs, including the VM types that do not support
Premium Disks. For example, if you're using an A-series VM, or N -series VM, or DS -series, or any other Azure VM
series, you can use Standard SSDs with that VM. With the introduction of Standard SSD, we are enabling a broad
range of workloads that previously used HDD -based disks to transition to SSD -based disks, and experience the
consistent performance, higher availability, better latency, and an overall better experience that come with SSDs.
Highly durable and available: Standard SSDs are built on the same Azure Disks platform, which has
consistently delivered high availability and durability for disks. Azure Disks are designed for 99.999 percent
availability. Like all Managed Disks, Standard SSDs will also offer Local Redundant Storage (LRS ). With LRS, the
platform maintains multiple replicas of data for every disk and has consistently delivered enterprise-grade
durability for IaaS disks, with an industry-leading ZERO percent Annualized Failure Rate.
Snapshots: Like all Managed Disks, Standard SSDs also support creation of Snapshots. Snapshot type can be
either Standard (HDD ) or Premium (SSD ). For cost saving, we recommend Snapshot type of Standard (HDD ) for
all Azure disk types. This is because when you create a managed disk from a snapshot, you're always able to
choose a higher tier such as Standard SSD or Premium SSD.

The following table contains disk sizes, which are currently offered for Standard SSD.
STANDARD SSD DISK TYPE DISK SIZE IOPS PER DISK THROUGHPUT PER DISK
E10 128 GiB Up to 500 Up to 60 MiB per second
E30 1,024 GiB Up to 500 Up to 60 MiB per second

STANDARD SSD DISK TYPE DISK SIZE IOPS PER DISK THROUGHPUT PER DISK
E60 8,192 GiB Up to 1,300 Up to 300 MiB per second
Standard SSDs are designed to provide single-digit millisecond latencies for most IO operations, and to deliver the
IOPS and throughput up to the limits described in the above table. Actual IOPS and Throughput may vary
sometimes depending on the traffic patterns. Standard SSDs will provide more consistent performance than the
HDD disks with the lower latency.
Premium SSDs on the other hand, perform better than Standard SSDs, with low latencies, high IOPS/throughput,
and even better consistency with provisioned disk performance. It is the recommended disk type for critical
production workloads. If your workload requires high-performance, low -latency disk support, you should consider
using Premium Storage.
Like the Premium SSDs, Standard SSDs also use IO Unit size of 256 KiB. If the data being transferred is less than
256 KiB, it is considered 1 I/O unit. Larger I/O sizes are counted as multiple I/Os of size 256 KiB. For example, a
1,100 KiB I/O is counted as five I/O units.
Pricing and billing

When using Standard SSDs, the following billing considerations apply:
Managed Disk Size
Snapshots
Transactions
Managed Disk Size: Managed disks are billed on the provisioned size. Azure maps the provisioned size (rounded
up) to the nearest disk size offer. For details of the disk sizes offered, see the table in Scalability and Performance
Targets section above. Each disk maps to a supported provisioned disk size and billed accordingly. For example, if
you provisioned a 200 GiB Standard SSD, it will map to the disk size offer of E15 (256 GiB ). Billing for any
provisioned disk is prorated hourly by using the monthly price for the Premium Storage offer. For example, if you
provisioned an E10 disk and deleted it after 20 hours, you're billed for the E10 offering prorated to 20 hours. This is
regardless of the amount of actual data written to the disk.
Snapshots: Snapshots of Managed Disks are billed for the capacity used by the snapshots, at the target and at the
source, if any. For more information on snapshots, see Managed Disk Snapshots.
bandwidth usage.
Transactions: Similar to Standard HDD, transactions on Standard SSDs incur billing. Transactions include both
read and write operations on the disk. I/O unit size used for accounting the transactions on Standard SSD is 256
KiB. Larger I/O sizes are counted as multiple I/Os of size 256 KiB.
For more information on pricing for Virtual Machines and Managed Disks, see:
Virtual Machines Pricing
Managed Disks Pricing
Next steps
To learn more about creating standard SSDs, refer to the sample that shows how to create a VM with multiple
standard SSDs.
Create a VM with multiple standard SSDs
To learn how to convert your existing HDD disks to standard SSD disks, refer to our article on the process.
Convert Azure managed disks storage from standard to premium, and vice versa
Ultra SSD (preview) Managed Disks for Azure Virtual
Machine workloads
Azure Ultra SSD (preview ) delivers high throughput, high IOPS, and consistent low latency disk storage for Azure
IaaS VMs. This new offering provides top of the line performance at the same availability levels as our existing
disks offerings. Additional benefits of Ultra SSD include the ability to dynamically change the performance of the
disk along with your workloads without the need to restart your virtual machines. Ultra SSD is suited for data-
intensive workloads such as SAP HANA, top tier databases, and transaction-heavy workloads.
Ultra SSD Features

Managed Disks: Ultra SSDs are only available as Managed Disks. Ultra SSDs cannot be deployed as an
Unmanaged Disk or Page Blob. While creating a Managed Disk, you specify the disk sku type as UltraSSD_LRS
and indicate the size of disk, the IOPS, and throughput you need, and Azure creates and manages the disk for you.
Virtual Machines: Ultra SSDs are designed to work with all Premium SSD enabled Azure Virtual Machine SKUs,
however at preview time the VM sizes will be limited to ES/DS v3 VM instances.
Dynamic Performance Configuration: Ultra SSDs allow you to dynamically change the performance (IOPS and
throughput) of the disk along with your workload needs without having to restart your virtual machines.

When you provision an Ultra SSD, you will have the option to independently configure the capacity and the
performance of the disk. Ultra SSDs come in several fixed sizes from 4 GiB up to 64 TiB and feature a flexible
performance configuration model that allows you to independently configure IOPS and throughput. Ultra SSDs
can only be leveraged as data disks. We recommend using Premium SSDs as OS disks.
Some key capabilities of Ultra SSD are:
Disk Capacity: Ultra SSD offers you a range of different disk sizes from 4 GiB up to 64 TiB.
Disk IOPS: Ultra SSDs support IOPS limits of 300 IOPS/GiB, up to a maximum of 160 K IOPS per disk. To
achieve the IOPS that you provisioned, ensure that the selected Disk IOPS are less than the VM IOPS. The
minimum disk IOPS is 100 IOPS.
Disk Throughput: With Ultra SSDs, the throughput limit of a single disk is 256 KiB/s for each provisioned IOPS,
up to a maximum of 2000 MBps per disk (where MBps = 10^6 Bytes per second). The minimum disk
throughput is 1 MiB.
The following table summarizes the different configurations supported for different disk sizes:
Ultra SSD Managed Disk Offerings
DISK SIZE (GIB) IOPS CAPS THROUGHPUT CAP (MBPS)
4 1,200 300
8 2,400 600
16 4,800 1,200
DISK SIZE (GIB) IOPS CAPS THROUGHPUT CAP (MBPS)
32 9,600 2,000
64 19,200 2,000
128 38,400 2,000
256 76,800 2,000
512 80,000 2,000
1,024-65,536 (sizes in this range 160,000 2,000

increasing in increments of 1 TiB)
Pricing and billing

When using Ultra SSDs, the following billing considerations apply:
Managed Disk Size
Managed Disk Provisioned IOPS
Managed Disk Provisioned Throughput
Ultra SSD VM reservation fee
Managed Disk Size
Managed Disks are billed on the provisioned sizes. Azure maps the provisioned size (rounded up) to the nearest
disk size offer. For details of the disk sizes offered, see the table in Scalability and Performance Targets section
above. Each disk maps to a supported provisioned disk size and billed accordingly on an hourly basis. For example,
if you provisioned a 200 GiB Ultra SSD Disk and deleted it after 20 hours, it will map to the disk size offer of 256
GiB and you'll be billed for the 256 GiB for 20 hours. This is regardless of the amount of actual data written to the
disk.
Managed Disk Provisioned IOPS
IOPS are the number of requests that your application is sending to the disks in one second. An input/output
operation could be sequential or random, reads, or writes. Like disk size, the provisioned IOPS are billed on an
hourly basis. For details of the disk IOPS offered, see the table in Scalability and Performance Targets section
above.
Managed Disk Provisioned Throughput
Throughput is the amount of data that your application is sending to the disks in a specified interval, measured in
bytes/second. If your application is performing large input/output operations, it requires high throughput.
There is a relation between Throughput and IOPS as shown in the following formula: IOPS x IO size = Throughput
Therefore, it is important to determine the optimal Throughput and IOPS values that your application requires. As
you try to optimize one, the other also gets affected. We recommend starting with a throughput corresponding to
16 KiB IO size, and adjusting if more throughput is needed.
For details on the supported disk throughput on Ultra SSD, see the table in Scalability and Performance Targets
section above. Like disk size and IOPS, the provisioned throughput is billed on an hourly basis per MBps
provisioned.
Ultra SSD VM reservation fee
We are introducing a capability on the VMthat indicates your VM is Ultra SSD compatible. An Ultra SSD
Compatible VM allocates dedicated bandwidth capacity between the compute VM instance and the block storage
scale unit to optimize the performance and reduce latency. Adding this capability on the VM results in a reservation
charge that is only imposed if you enabled Ultra SSD capability on the VM without attaching an Ultra SSD disk to
it. When an Ultra SSD disk is attached to the Ultra SSD compatible VM, this charge would not be applied. This
charge is per vCPU provisioned on the VM.
Refer to the Azure Disks pricing page for the new Ultra SSD Disks price details available in limited preview.
Ultra SSD Preview Scope and Limitations
At preview, Ultra SSD Disks:
Will be initially supported in East US 2 in a single Availability Zone
Can only be used with Availability Zones (Availability Sets and Single VM deployments outside of Zones will
not have the ability to attach an Ultra SSD Disk)
Are only supported on ES/DS v3 VMs
Are only available as data disks and only support 4k physical sector size
Can only be created as Empty disks
Currently can only be deployed using Resource Manager templates, CLI, and Python SDK.
Will not (yet) support disk snapshots, VM Images, Availability Sets, Virtual Machine Scale Sets and Azure Disk
Encryption.
Will not (yet) support integration with Azure Backup or Azure Site Recovery.
As withmost previews, this feature should not be used for production workloads until General Availability (GA).
Next steps
If you're interested in trying ultra SSDs, see the how to guide on the subject.
How to enable and use Ultra SSDs
Azure Premium Storage: Design for High
Performance
This article provides guidelines for building high performance applications using Azure Premium Storage. You can
use the instructions provided in this document combined with performance best practices applicable to
technologies used by your application. To illustrate the guidelines, we have used SQL Server running on Premium
Storage as an example throughout this document.
While we address performance scenarios for the Storage layer in this article, you will need to optimize the
application layer. For example, if you are hosting a SharePoint Farm on Azure Premium Storage, you can use the
SQL Server examples from this article to optimize the database server. Additionally, optimize the SharePoint
Farm's Web server and Application server to get the most performance.
This article will help answer following common questions about optimizing application performance on Azure
Premium Storage,
How to measure your application performance?
Why are you not seeing expected high performance?
Which factors influence your application performance on Premium Storage?
How do these factors influence performance of your application on Premium Storage?
How can you optimize for IOPS, Bandwidth and Latency?
We have provided these guidelines specifically for Premium Storage because workloads running on Premium
Storage are highly performance sensitive. We have provided examples where appropriate. You can also apply
some of these guidelines to applications running on IaaS VMs with Standard Storage disks.
Before you begin, if you are new to Premium Storage, first read the Premium Storage: High-Performance Storage
for Azure Virtual Machine Workloads and Azure Storage Scalability and Performance Targets articles.
Application Performance Indicators

We assess whether an application is performing well or not using performance indicators like, how fast an
application is processing a user request, how much data an application is processing per request, how many
requests is an application processing in a specific period of time, how long a user has to wait to get a response
after submitting their request. The technical terms for these performance indicators are, IOPS, Throughput or
Bandwidth, and Latency.
In this section, we will discuss the common performance indicators in the context of Premium Storage. In the
following section, Gathering Application Requirements, you will learn how to measure these performance
indicators for your application. Later in Optimizing Application Performance, you will learn about the factors
affecting these performance indicators and recommendations to optimize them.
IOPS
IOPS is number of requests that your application is sending to the storage disks in one second. An input/output
operation could be read or write, sequential or random. OLTP applications like an online retail website need to
process many concurrent user requests immediately. The user requests are insert and update intensive database
transactions, which the application must process quickly. Therefore, OLTP applications require very high IOPS.
Such applications handle millions of small and random IO requests. If you have such an application, you must
design the application infrastructure to optimize for IOPS. In the later section, Optimizing Application
Performance, we discuss in detail all the factors that you must consider to get high IOPS.
When you attach a premium storage disk to your high scale VM, Azure provisions for you a guaranteed number
of IOPS as per the disk specification. For example, a P50 disk provisions 7500 IOPS. Each high scale VM size also
has a specific IOPS limit that it can sustain. For example, a Standard GS5 VM has 80,000 IOPS limit.
Throughput
Throughput or Bandwidth is the amount of data that your application is sending to the storage disks in a specified
interval. If your application is performing input/output operations with large IO unit sizes, it requires high
Throughput. Data warehouse applications tend to issue scan intensive operations that access large portions of
data at a time and commonly perform bulk operations. In other words, such applications require higher
Throughput. If you have such an application, you must design its infrastructure to optimize for Throughput. In the
next section, we discuss in detail the factors you must tune to achieve this.
When you attach a premium storage disk to a high scale VM, Azure provisions Throughput as per that disk
specification. For example, a P50 disk provisions 250 MB per second disk Throughput. Each high scale VM size
also has as specific Throughput limit that it can sustain. For example, Standard GS5 VM has a maximum
throughput of 2,000 MB per second.
There is a relation between Throughput and IOPS as shown in the formula below.
Therefore, it is important to determine the optimal Throughput and IOPS values that your application requires. As
you try to optimize one, the other also gets affected. In a later section, Optimizing Application Performance, we
will discuss in more details about optimizing IOPS and Throughput.
Latency
Latency is the time it takes an application to receive a single request, send it to the storage disks and send the
response to the client. This is a critical measure of an application's performance in addition to IOPS and
Throughput. The Latency of a premium storage disk is the time it takes to retrieve the information for a request
and communicate it back to your application. Premium Storage provides consistent low latencies. If you enable
ReadOnly host caching on premium storage disks, you can get much lower read latency. We will discuss Disk
Caching in more detail in later section on Optimizing Application Performance.
When you are optimizing your application to get higher IOPS and Throughput, it will affect the Latency of your
application. After tuning the application performance, always evaluate the Latency of the application to avoid
unexpected high latency behavior.
Gather Application Performance Requirements

The first step in designing high performance applications running on Azure Premium Storage is, to understand
the performance requirements of your application. After you gather performance requirements, you can optimize
your application to achieve the most optimal performance.
In the previous section, we explained the common performance indicators, IOPS, Throughput and Latency. You
must identify which of these performance indicators are critical to your application to deliver the desired user
experience. For example, high IOPS matters most to OLTP applications processing millions of transactions in a
second. Whereas, high Throughput is critical for Data Warehouse applications processing large amounts of data in
a second. Extremely low Latency is crucial for real-time applications like live video streaming websites.
Next, measure the maximum performance requirements of your application throughout its lifetime. Use the
sample checklist below as a start. Record the maximum performance requirements during normal, peak and off-
hours workload periods. By identifying requirements for all workloads levels, you will be able to determine the
overall performance requirement of your application. For example, the normal workload of an e-commerce
website will be the transactions it serves during most days in a year. The peak workload of the website will be the
transactions it serves during holiday season or special sale events. The peak workload is typically experienced for a
limited period, but can require your application to scale two or more times its normal operation. Find out the 50
percentile, 90 percentile and 99 percentile requirements. This helps filter out any outliers in the performance
requirements and you can focus your efforts on optimizing for the right values.
Application Performance Requirements Checklist
PERFORMANCE
REQUIREMENTS 50 PERCENTILE 90 PERCENTILE 99 PERCENTILE
Max. Transactions per

second
% Read operations
% Write operations
% Random operations
% Sequential operations
IO request size
Average Throughput
Max. Throughput
Min. Latency
Average Latency
Max. CPU
Average CPU
Max. Memory
Average Memory
Queue Depth
NOTE
You should consider scaling these numbers based on expected future growth of your application. It is a good idea to plan for
growth ahead of time, because it could be harder to change the infrastructure for improving performance later.
If you have an existing application and want to move to Premium Storage, first build the checklist above for the
existing application. Then, build a prototype of your application on Premium Storage and design the application
based on guidelines described in Optimizing Application Performance in a later section of this document. The next
section describes the tools you can use to gather the performance measurements.
Create a checklist similar to your existing application for the prototype. Using Benchmarking tools you can
simulate the workloads and measure performance on the prototype application. See the section on Benchmarking
to learn more. By doing so you can determine whether Premium Storage can match or surpass your application
performance requirements. Then you can implement the same guidelines for your production application.
Counters to measure application performance requirements
The best way to measure performance requirements of your application, is to use performance-monitoring tools
provided by the operating system of the server. You can use PerfMon for Windows and iostat for Linux. These
tools capture counters corresponding to each measure explained in the above section. You must capture the values
of these counters when your application is running its normal, peak and off-hours workloads.
The PerfMon counters are available for processor, memory and, each logical disk and physical disk of your server.
When you use premium storage disks with a VM, the physical disk counters are for each premium storage disk,
and logical disk counters are for each volume created on the premium storage disks. You must capture the values
for the disks that host your application workload. If there is a one to one mapping between logical and physical
disks, you can refer to physical disk counters; otherwise refer to the logical disk counters. On Linux, the iostat
command generates a CPU and disk utilization report. The disk utilization report provides statistics per physical
device or partition. If you have a database server with its data and log on separate disks, collect this data for both
disks. Below table describes counters for disks, processor and memory:
COUNTER DESCRIPTION PERFMON IOSTAT
IOPS or Transactions per Number of I/O requests Disk Reads/sec tps

second issued to the storage disk Disk Writes/sec r/s
per second. w/s
Disk Reads and Writes % of Reads and Write % Disk Read Time r/s
operations performed on % Disk Write Time w/s
the disk.
Throughput Amount of data read from Disk Read Bytes/sec kB_read/s

or written to the disk per Disk Write Bytes/sec kB_wrtn/s
second.
Latency Total time to complete a disk Average Disk sec/Read await

IO request. Average disk sec/Write svctm
IO size The size of I/O requests Average Disk Bytes/Read avgrq-sz

issues to the storage disks. Average Disk Bytes/Write
Queue Depth Number of outstanding I/O Current Disk Queue Length avgqu-sz
requests waiting to be read
from or written to the
storage disk.
Max. Memory Amount of memory % Committed Bytes in Use Use vmstat

required to run application
smoothly
Max. CPU Amount CPU required to % Processor time %util

run application smoothly
Learn more about iostat and PerfMon.

Optimizing Application Performance
The main factors that influence performance of an application running on Premium Storage are Nature of IO
Requests, VM size, Disk size, Number of disks, Disk Caching, Multithreading and Queue Depth. You can control
some of these factors with knobs provided by the system. Most applications may not give you an option to alter
the IO size and Queue Depth directly. For example, if you are using SQL Server, you cannot choose the IO size
and queue depth. SQL Server chooses the optimal IO size and queue depth values to get the most performance. It
is important to understand the effects of both types of factors on your application performance, so that you can
provision appropriate resources to meet performance needs.
Throughout this section, refer to the application requirements checklist that you created, to identify how much you
need to optimize your application performance. Based on that, you will be able to determine which factors from
this section you will need to tune. To witness the effects of each factor on your application performance, run
benchmarking tools on your application setup. Refer to the Benchmarking section at the end of this article for
steps to run common benchmarking tools on Windows and Linux VMs.
Optimizing IOPS, Throughput and Latency at a glance
The table below summarizes performance factors and the steps necessary to optimize IOPS, throughput and
latency. The sections following this summary will describe each factor is much more depth.
For more information on VM sizes and on the IOPS, throughput, and latency available for each type of VM, see
Linux VM sizes or Windows VM sizes.
IOPS THROUGHPUT LATENCY
Example Scenario Enterprise OLTP application Enterprise Data warehousing Near real-time applications
requiring very high application processing large requiring instant responses
transactions per second rate. amounts of data. to user requests, like online
gaming.
Performance factors
IO size Smaller IO size yields higher Larger IO size to yields

IOPS. higher Throughput.
VM size Use a VM size that offers Use a VM size with Use a VM size that offers
IOPS greater than your throughput limit greater scale limits greater than
application requirement. than your application your application
requirement. requirement.
Disk size Use a disk size that offers Use a disk size with Use a disk size that offers
IOPS greater than your Throughput limit greater scale limits greater than
application requirement. than your application your application
requirement. requirement.
VM and Disk Scale Limits IOPS limit of the VM size Throughput limit of the VM Scale limits of the VM size
chosen should be greater size chosen should be chosen must be greater
than total IOPS driven by greater than total than total scale limits of
premium storage disks Throughput driven by attached premium storage
attached to it. premium storage disks disks.
attached to it.
Disk Caching Enable ReadOnly Cache on Enable ReadOnly Cache on

premium storage disks with premium storage disks with
Read heavy operations to Ready heavy operations to
get higher Read IOPS. get very low Read latencies.
IOPS THROUGHPUT LATENCY
Disk Striping Use multiple disks and stripe

them together to get a
combined higher IOPS and
Throughput limit. Note that
the combined limit per VM
should be higher than the
combined limits of attached
premium disks.
Stripe Size Smaller stripe size for Larger stripe size for
random small IO pattern sequential large IO pattern
seen in OLTP applications. seen in Data Warehouse
E.g., use stripe size of 64KB applications. E.g., use 256KB
for SQL Server OLTP stripe size for SQL Server
application. Data warehouse application.
Multithreading Use multithreading to push

higher number of requests
to Premium Storage that will
lead to higher IOPS and
Throughput. For example,
on SQL Server set a high
MAXDOP value to allocate
more CPUs to SQL Server.
Queue Depth Larger Queue Depth yields Larger Queue Depth yields Smaller Queue Depth yields
higher IOPS. higher Throughput. lower latencies.
Nature of IO Requests
An IO request is a unit of input/output operation that your application will be performing. Identifying the nature of
IO requests, random or sequential, read or write, small or large, will help you determine the performance
requirements of your application. It is very important to understand the nature of IO requests, to make the right
decisions when designing your application infrastructure.
IO size is one of the more important factors. The IO size is the size of the input/output operation request
generated by your application. The IO size has a significant impact on performance especially on the IOPS and
Bandwidth that the application is able to achieve. The following formula shows the relationship between IOPS, IO
size and Bandwidth/Throughput.
Some applications allow you to alter their IO size, while some applications do not. For example, SQL Server
determines the optimal IO size itself, and does not provide users with any knobs to change it. On the other hand,
Oracle provides a parameter called DB_BLOCK_SIZE using which you can configure the I/O request size of the
database.
If you are using an application, which does not allow you to change the IO size, use the guidelines in this article to
optimize the performance KPI that is most relevant to your application. For example,
An OLTP application generates millions of small and random IO requests. To handle these type of IO requests,
you must design your application infrastructure to get higher IOPS.
A data warehousing application generates large and sequential IO requests. To handle these type of IO
requests, you must design your application infrastructure to get higher Bandwidth or Throughput.
If you are using an application, which allows you to change the IO size, use this rule of thumb for the IO size in
addition to other performance guidelines,
Smaller IO size to get higher IOPS. For example, 8 KB for an OLTP application.
Larger IO size to get higher Bandwidth/Throughput. For example, 1024 KB for a data warehouse application.
Here is an example on how you can calculate the IOPS and Throughput/Bandwidth for your application. Consider
an application using a P30 disk. The maximum IOPS and Throughput/Bandwidth a P30 disk can achieve is 5000
IOPS and 200 MB per second respectively. Now, if your application requires the maximum IOPS from the P30
disk and you use a smaller IO size like 8 KB, the resulting Bandwidth you will be able to get is 40 MB per second.
However, if your application requires the maximum Throughput/Bandwidth from P30 disk, and you use a larger
IO size like 1024 KB, the resulting IOPS will be less, 200 IOPS. Therefore, tune the IO size such that it meets both
your application's IOPS and Throughput/Bandwidth requirement. Table below summarizes the different IO sizes
and their corresponding IOPS and Throughput for a P30 disk.
APPLICATION REQUIREMENT I/O SIZE IOPS THROUGHPUT/BANDWIDTH
Max IOPS 8 KB 5,000 40 MB per second
Max Throughput 1024 KB 200 200 MB per second
Max Throughput + high 64 KB 3,200 200 MB per second

IOPS
Max IOPS + high 32 KB 5,000 160 MB per second

Throughput
To get IOPS and Bandwidth higher than the maximum value of a single premium storage disk, use multiple
premium disks striped together. For example, stripe two P30 disks to get a combined IOPS of 10,000 IOPS or a
combined Throughput of 400 MB per second. As explained in the next section, you must use a VM size that
supports the combined disk IOPS and Throughput.
NOTE
As you increase either IOPS or Throughput the other also increases, make sure you do not hit throughput or IOPS limits of
the disk or VM when increasing either one.
To witness the effects of IO size on application performance, you can run benchmarking tools on your VM and
disks. Create multiple test runs and use different IO size for each run to see the impact. Refer to the Benchmarking
section at the end of this article for more details.
High Scale VM Sizes

When you start designing an application, one of the first things to do is, choose a VM to host your application.
Premium Storage comes with High Scale VM sizes that can run applications requiring higher compute power and
a high local disk I/O performance. These VMs provide faster processors, a higher memory-to-core ratio, and a
Solid-State Drive (SSD ) for the local disk. Examples of High Scale VMs supporting Premium Storage are the DS,
DSv2 and GS series VMs.
High Scale VMs are available in different sizes with a different number of CPU cores, memory, OS and temporary
disk size. Each VM size also has maximum number of data disks that you can attach to the VM. Therefore, the
chosen VM size will affect how much processing, memory, and storage capacity is available for your application. It
also affects the Compute and Storage cost. For example, below are the specifications of the largest VM size in a
DS series, DSv2 series and a GS series:
BANDWIDT
VM DISK MAX. DATA H CACHE IO
VM SIZE CPU CORES MEMORY SIZES DISKS CACHE SIZE IOPS LIMITS
Standard_D 16 112 GB OS = 1023 32 576 GB 50,000 4,000 IOPS

S14 GB IOPS and 33 MB
Local SSD = 512 MB per per second
224 GB second
Standard_G 32 448 GB OS = 1023 64 4224 GB 80,000 5,000 IOPS

S5 GB IOPS and 50 MB
Local SSD = 2,000 MB per second
896 GB per second
To view a complete list of all available Azure VM sizes, refer to Windows VM sizes or Linux VM sizes. Choose a
VM size that can meet and scale to your desired application performance requirements. In addition to this, take
into account following important considerations when choosing VM sizes.
Scale Limits
The maximum IOPS limits per VM and per disk are different and independent of each other. Make sure that the
application is driving IOPS within the limits of the VM as well as the premium disks attached to it. Otherwise,
application performance will experience throttling.
As an example, suppose an application requirement is a maximum of 4,000 IOPS. To achieve this, you provision a
P30 disk on a DS1 VM. The P30 disk can deliver up to 5,000 IOPS. However, the DS1 VM is limited to 3,200
IOPS. Consequently, the application performance will be constrained by the VM limit at 3,200 IOPS and there will
be degraded performance. To prevent this situation, choose a VM and disk size that will both meet application
requirements.
Cost of Operation
In many cases, it is possible that your overall cost of operation using Premium Storage is lower than using
Standard Storage.
For example, consider an application requiring 16,000 IOPS. To achieve this performance, you will need a
Standard_D14 Azure IaaS VM, which can give a maximum IOPS of 16,000 using 32 standard storage 1TB disks.
Each 1TB standard storage disk can achieve a maximum of 500 IOPS. The estimated cost of this VM per month
will be $1,570. The monthly cost of 32 standard storage disks will be $1,638. The estimated total monthly cost will
be $3,208.
However, if you hosted the same application on Premium Storage, you will need a smaller VM size and fewer
premium storage disks, thus reducing the overall cost. A Standard_DS13 VM can meet the 16,000 IOPS
requirement using four P30 disks. The DS13 VM has a maximum IOPS of 25,600 and each P30 disk has a
maximum IOPS of 5,000. Overall, this configuration can achieve 5,000 x 4 = 20,000 IOPS. The estimated cost of
this VM per month will be $1,003. The monthly cost of four P30 premium storage disks will be $544.34. The
estimated total monthly cost will be $1,544.
Table below summarizes the cost breakdown of this scenario for Standard and Premium Storage.
STANDARD PREMIUM
Cost of VM per month $1,570.58 (Standard_D14) $1,003.66 (Standard_DS13)
Cost of Disks per month $1,638.40 (32 x 1 TB disks) $544.34 (4 x P30 disks)
Overall Cost per month $3,208.98 $1,544.34
Linux Distros
With Azure Premium Storage, you get the same level of Performance for VMs running Windows and Linux. We
support many flavors of Linux distros, and you can see the complete list here. It is important to note that different
distros are better suited for different types of workloads. You will see different levels of performance depending on
the distro your workload is running on. Test the Linux distros with your application and choose the one that works
best.
When running Linux with Premium Storage, check the latest updates about required drivers to ensure high
performance.
Premium Storage Disk Sizes

Azure Premium Storage offers eight GA disk sizes and three disk sizes which are in preview, currently. Each disk
size has a different scale limit for IOPS, Bandwidth and Storage. Choose the right Premium Storage Disk size
depending on the application requirements and the high scale VM size. The table below shows the eleven disks
sizes and their capabilities. P4, P6, P15, P60, P70, and P80 sizes are currently only supported for Managed Disks.
PREMI
UM
DISKS
Disk 32 64 128 256 512 1,024 2,048 4,095 8,192 16,38 32,76
size GiB GiB GiB GiB GB GiB (1 GiB (2 GiB (4 GiB (8 4 GiB 7 GiB
TiB) GiB)
IOPS 120 240 500 1100 2300 5000 7500 7500 12,50 15,00 20,00
per 0 0 0
disk
Throu 25 50 100 125 150 200 250 250 480 750 750
ghput MiB MiB MiB MiB MiB MiB MiB MiB MiB MiB MiB
per per per per per per per per per per per per
disk secon secon secon secon secon secon secon secon secon secon secon
d d d d d d d d d d d
How many disks you choose depends on the disk size chosen. You could use a single P50 disk or multiple P10
disks to meet your application requirement. Take into account considerations listed below when making the
choice.
Scale Limits (IOPS and Throughput)
The IOPS and Throughput limits of each Premium disk size is different and independent from the VM scale limits.
Make sure that the total IOPS and Throughput from the disks are within scale limits of the chosen VM size.
For example, if an application requirement is a maximum of 250 MB/sec Throughput and you are using a DS4
VM with a single P30 disk. The DS4 VM can give up to 256 MB/sec Throughput. However, a single P30 disk has
Throughput limit of 200 MB/sec. Consequently, the application will be constrained at 200 MB/sec due to the disk
limit. To overcome this limit, provision more than one data disks to the VM or resize your disks to P40 or P50.
NOTE
Reads served by the cache are not included in the disk IOPS and Throughput, hence not subject to disk limits. Cache has its
separate IOPS and Throughput limit per VM.
For example, initially your reads and writes are 60MB/sec and 40MB/sec respectively. Over time, the cache warms up and
serves more and more of the reads from the cache. Then, you can get higher write Throughput from the disk.
Number of Disks
Determine the number of disks you will need by assessing application requirements. Each VM size also has a limit
on the number of disks that you can attach to the VM. Typically, this is twice the number of cores. Ensure that the
VM size you choose can support the number of disks needed.
Remember, the Premium Storage disks have higher performance capabilities compared to Standard Storage disks.
Therefore, if you are migrating your application from Azure IaaS VM using Standard Storage to Premium Storage,
you will likely need fewer premium disks to achieve the same or higher performance for your application.
Disk Caching
High Scale VMs that leverage Azure Premium Storage have a multi-tier caching technology called BlobCache.
BlobCache uses a combination of the Virtual Machine RAM and local SSD for caching. This cache is available for
the Premium Storage persistent disks and the VM local disks. By default, this cache setting is set to Read/Write for
OS disks and ReadOnly for data disks hosted on Premium Storage. With disk caching enabled on the Premium
Storage disks, the high scale VMs can achieve extremely high levels of performance that exceed the underlying
disk performance.
WARNING
Disk Caching is only supported for disk sizes up to 4 TiB. Changing the cache setting of an Azure disk detaches and re-
attaches the target disk. If it is the operating system disk, the VM is restarted. Stop all applications/services that might be
affected by this disruption before changing the disk cache setting.
To learn more about how BlobCache works, refer to the Inside Azure Premium Storage blog post.
It is important to enable cache on the right set of disks. Whether you should enable disk caching on a premium
disk or not will depend on the workload pattern that disk will be handling. Table below shows the default cache
settings for OS and Data disks.
DISK TYPE DEFAULT CACHE SETTING
OS disk ReadWrite
Data disk ReadOnly
Following are the recommended disk cache settings for data disks,
DISK CACHING SETTING RECOMMENDATION ON WHEN TO USE THIS SETTING
None Configure host-cache as None for write-only and write-heavy

disks.
ReadOnly Configure host-cache as ReadOnly for read-only and read-

write disks.
ReadWrite Configure host-cache as ReadWrite only if your application

properly handles writing cached data to persistent disks when
needed.
ReadOnly
By configuring ReadOnly caching on Premium Storage data disks, you can achieve low Read latency and get very
high Read IOPS and Throughput for your application. This is due two reasons,
1. Reads performed from cache, which is on the VM memory and local SSD, are much faster than reads from the
data disk, which is on the Azure blob storage.
2. Premium Storage does not count the Reads served from cache, towards the disk IOPS and Throughput.
Therefore, your application is able to achieve higher total IOPS and Throughput.
ReadWrite
By default, the OS disks have ReadWrite caching enabled. We have recently added support for ReadWrite caching
on data disks as well. If you are using ReadWrite caching, you must have a proper way to write the data from
cache to persistent disks. For example, SQL Server handles writing cached data to the persistent storage disks on
its own. Using ReadWrite cache with an application that does not handle persisting the required data can lead to
data loss, if the VM crashes.
As an example, you can apply these guidelines to SQL Server running on Premium Storage by doing the
following,
1. Configure "ReadOnly" cache on premium storage disks hosting data files.
a. The fast reads from cache lower the SQL Server query time since data pages are retrieved much faster from
the cache compared to directly from the data disks.
b. Serving reads from cache, means there is additional Throughput available from premium data disks. SQL
Server can use this additional Throughput towards retrieving more data pages and other operations like
backup/restore, batch loads, and index rebuilds.
2. Configure "None" cache on premium storage disks hosting the log files.
a. Log files have primarily write-heavy operations. Therefore, they do not benefit from the ReadOnly cache.
Disk Striping
When a high scale VM is attached with several premium storage persistent disks, the disks can be striped together
to aggregate their IOPs, bandwidth, and storage capacity.
On Windows, you can use Storage Spaces to stripe disks together. You must configure one column for each disk in
a pool. Otherwise, the overall performance of striped volume can be lower than expected, due to uneven
distribution of traffic across the disks.
Important: Using Server Manager UI, you can set the total number of columns up to 8 for a striped volume. When
attaching more than 8 disks, use PowerShell to create the volume. Using PowerShell, you can set the number of
columns equal to the number of disks. For example, if there are 16 disks in a single stripe set; specify 16 columns
in the NumberOfColumns parameter of the New -VirtualDisk PowerShell cmdlet.
On Linux, use the MDADM utility to stripe disks together. For detailed steps on striping disks on Linux refer to
Configure Software RAID on Linux.
Stripe Size
An important configuration in disk striping is the stripe size. The stripe size or block size is the smallest chunk of
data that application can address on a striped volume. The stripe size you configure depends on the type of
application and its request pattern. If you choose the wrong stripe size, it could lead to IO misalignment, which
leads to degraded performance of your application.
For example, if an IO request generated by your application is bigger than the disk stripe size, the storage system
writes it across stripe unit boundaries on more than one disk. When it is time to access that data, it will have to
seek across more than one stripe units to complete the request. The cumulative effect of such behavior can lead to
substantial performance degradation. On the other hand, if the IO request size is smaller than stripe size, and if it
is random in nature, the IO requests may add up on the same disk causing a bottleneck and ultimately degrading
the IO performance.
Depending on the type of workload your application is running, choose an appropriate stripe size. For random
small IO requests, use a smaller stripe size. Whereas, for large sequential IO requests use a larger stripe size. Find
out the stripe size recommendations for the application you will be running on Premium Storage. For SQL Server,
configure stripe size of 64KB for OLTP workloads and 256KB for data warehousing workloads. See Performance
best practices for SQL Server on Azure VMs to learn more.
NOTE
You can stripe together a maximum of 32 premium storage disks on a DS series VM and 64 premium storage disks on a GS
series VM.
Multi-threading
Azure has designed Premium Storage platform to be massively parallel. Therefore, a multi-threaded application
achieves much higher performance than a single-threaded application. A multi-threaded application splits up its
tasks across multiple threads and increases efficiency of its execution by utilizing the VM and disk resources to the
maximum.
For example, if your application is running on a single core VM using two threads, the CPU can switch between
the two threads to achieve efficiency. While one thread is waiting on a disk IO to complete, the CPU can switch to
the other thread. In this way, two threads can accomplish more than a single thread would. If the VM has more
than one core, it further decreases running time since each core can execute tasks in parallel.
You may not be able to change the way an off-the-shelf application implements single threading or multi-
threading. For example, SQL Server is capable of handling multi-CPU and multi-core. However, SQL Server
decides under what conditions it will leverage one or more threads to process a query. It can run queries and build
indexes using multi-threading. For a query that involves joining large tables and sorting data before returning to
the user, SQL Server will likely use multiple threads. However, a user cannot control whether SQL Server executes
a query using a single thread or multiple threads.
There are configuration settings that you can alter to influence this multi-threading or parallel processing of an
application. For example, in case of SQL Server it is the maximum Degree of Parallelism configuration. This
setting called MAXDOP, allows you to configure the maximum number of processors SQL Server can use when
parallel processing. You can configure MAXDOP for individual queries or index operations. This is beneficial when
you want to balance resources of your system for a performance critical application.
For example, say your application using SQL Server is executing a large query and an index operation at the same
time. Let us assume that you wanted the index operation to be more performant compared to the large query. In
such a case, you can set MAXDOP value of the index operation to be higher than the MAXDOP value for the
query. This way, SQL Server has more number of processors that it can leverage for the index operation
compared to the number of processors it can dedicate to the large query. Remember, you do not control the
number of threads SQL Server will use for each operation. You can control the maximum number of processors
being dedicated for multi-threading.
Learn more about Degrees of Parallelism in SQL Server. Find out such settings that influence multi-threading in
your application and their configurations to optimize performance.
Queue Depth
The Queue Depth or Queue Length or Queue Size is the number of pending IO requests in the system. The value
of Queue Depth determines how many IO operations your application can line up, which the storage disks will be
processing. It affects all the three application performance indicators that we discussed in this article viz., IOPS,
Throughput and Latency.
Queue Depth and multi-threading are closely related. The Queue Depth value indicates how much multi-
threading can be achieved by the application. If the Queue Depth is large, application can execute more operations
concurrently, in other words, more multi-threading. If the Queue Depth is small, even though application is multi-
threaded, it will not have enough requests lined up for concurrent execution.
Typically, off the shelf applications do not allow you to change the queue depth, because if set incorrectly it will do
more harm than good. Applications will set the right value of queue depth to get the optimal performance.
However, it is important to understand this concept so that you can troubleshoot performance issues with your
application. You can also observe the effects of queue depth by running benchmarking tools on your system.
Some applications provide settings to influence the Queue Depth. For example, the MAXDOP (maximum degree
of parallelism) setting in SQL Server explained in previous section. MAXDOP is a way to influence Queue Depth
and multi-threading, although it does not directly change the Queue Depth value of SQL Server.
High Queue Depth
A high queue depth lines up more operations on the disk. The disk knows the next request in its queue ahead of
time. Consequently, the disk can schedule operations ahead of time and process them in an optimal sequence.
Since the application is sending more requests to the disk, the disk can process more parallel IOs. Ultimately, the
application will be able to achieve higher IOPS. Since application is processing more requests, the total
Throughput of the application also increases.
Typically, an application can achieve maximum Throughput with 8-16+ outstanding IOs per attached disk. If a
Queue Depth is one, application is not pushing enough IOs to the system, and it will process less amount of in a
given period. In other words, less Throughput.
For example, in SQL Server, setting the MAXDOP value for a query to "4" informs SQL Server that it can use up
to four cores to execute the query. SQL Server will determine what is best queue depth value and the number of
cores for the query execution.
Optimal Queue Depth
Very high queue depth value also has its drawbacks. If queue depth value is too high, the application will try to
drive very high IOPS. Unless application has persistent disks with sufficient provisioned IOPS, this can negatively
affect application latencies. Following formula shows the relationship between IOPS, Latency and Queue Depth.
You should not configure Queue Depth to any high value, but to an optimal value, which can deliver enough IOPS
for the application without affecting latencies. For example, if the application latency needs to be 1 millisecond, the
Queue Depth required to achieve 5,000 IOPS is, QD = 5000 x 0.001 = 5.
Queue Depth for Striped Volume
For a striped volume, maintain a high enough queue depth such that, every disk has a peak queue depth
individually. For example, consider an application that pushes a queue depth of 2 and there are 4 disks in the
stripe. The two IO requests will go to two disks and remaining two disks will be idle. Therefore, configure the
queue depth such that all the disks can be busy. Formula below shows how to determine the queue depth of
striped volumes.
Throttling
Azure Premium Storage provisions specified number of IOPS and Throughput depending on the VM sizes and
disk sizes you choose. Anytime your application tries to drive IOPS or Throughput above these limits of what the
VM or disk can handle, Premium Storage will throttle it. This manifests in the form of degraded performance in
your application. This can mean higher latency, lower Throughput or lower IOPS. If Premium Storage does not
throttle, your application could completely fail by exceeding what its resources are capable of achieving. So, to
avoid performance issues due to throttling, always provision sufficient resources for your application. Take into
consideration what we discussed in the VM sizes and Disk sizes sections above. Benchmarking is the best way to
figure out what resources you will need to host your application.
Benchmarking
Benchmarking is the process of simulating different workloads on your application and measuring the application
performance for each workload. Using the steps described in an earlier section, you have gathered the application
performance requirements. By running benchmarking tools on the VMs hosting the application, you can
determine the performance levels that your application can achieve with Premium Storage. In this section, we
provide you examples of benchmarking a Standard DS14 VM provisioned with Azure Premium Storage disks.
We have used common benchmarking tools Iometer and FIO, for Windows and Linux respectively. These tools
spawn multiple threads simulating a production like workload, and measure the system performance. Using the
tools you can also configure parameters like block size and queue depth, which you normally cannot change for an
application. This gives you more flexibility to drive the maximum performance on a high scale VM provisioned
with premium disks for different types of application workloads. To learn more about each benchmarking tool visit
Iometer and FIO.
To follow the examples below, create a Standard DS14 VM and attach 11 Premium Storage disks to the VM. Of
the 11 disks, configure 10 disks with host caching as "None" and stripe them into a volume called NoCacheWrites.
Configure host caching as "ReadOnly" on the remaining disk and create a volume called CacheReads with this
disk. Using this setup, you will be able to see the maximum Read and Write performance from a Standard DS14
VM. For detailed steps about creating a DS14 VM with premium disks, go to Create and use a Premium Storage
account for a virtual machine data disk.
Warming up the Cache
The disk with ReadOnly host caching will be able to give higher IOPS than the disk limit. To get this maximum
read performance from the host cache, first you must warm up the cache of this disk. This ensures that the Read
IOs which benchmarking tool will drive on CacheReads volume actually hits the cache and not the disk directly.
The cache hits result in additional IOPS from the single cache enabled disk.
Important:
You must warm up the cache before running benchmarking, every time VM is rebooted.
Iometer
Download the Iometer tool on the VM.
Test file
Iometer uses a test file that is stored on the volume on which you will run the benchmarking test. It drives Reads
and Writes on this test file to measure the disk IOPS and Throughput. Iometer creates this test file if you have not
provided one. Create a 200GB test file called iobw.tst on the CacheReads and NoCacheWrites volumes.
Access Specifications
The specifications, request IO size, % read/write, % random/sequential are configured using the "Access
Specifications" tab in Iometer. Create an access specification for each of the scenarios described below. Create the
access specifications and "Save" with an appropriate name like – RandomWrites_8K, RandomReads_8K. Select the
corresponding specification when running the test scenario.
An example of access specifications for maximum Write IOPS scenario is shown below,
Maximum IOPS Test Specifications
To demonstrate maximum IOPs, use smaller request size. Use 8K request size and create specifications for
Random Writes and Reads.
ACCESS SPECIFICATION REQUEST SIZE RANDOM % READ %
RandomWrites_8K 8K 100 0
RandomReads_8K 8K 100 100
Maximum Throughput Test Specifications

To demonstrate maximum Throughput, use larger request size. Use 64K request size and create specifications for
Random Writes and Reads.
ACCESS SPECIFICATION REQUEST SIZE RANDOM % READ %
RandomWrites_64K 64K 100 0
RandomReads_64K 64K 100 100
Running the Iometer Test

Perform the steps below to warm up cache
1. Create two access specifications with values shown below,
NAME REQUEST SIZE RANDOM % READ %
RandomWrites_1MB 1MB 100 0
RandomReads_1MB 1MB 100 100

2. Run the Iometer test for initializing cache disk with following parameters. Use three worker threads for the
target volume and a queue depth of 128. Set the "Run time" duration of the test to 2hrs on the "Test Setup"
tab.
SCENARIO TARGET VOLUME NAME DURATION
Initialize Cache Disk CacheReads RandomWrites_1MB 2hrs
3. Run the Iometer test for warming up cache disk with following parameters. Use three worker threads for
the target volume and a queue depth of 128. Set the "Run time" duration of the test to 2hrs on the "Test
Setup" tab.
SCENARIO TARGET VOLUME NAME DURATION
Warm up Cache Disk CacheReads RandomReads_1MB 2hrs
After cache disk is warmed up, proceed with the test scenarios listed below. To run the Iometer test, use at least
three worker threads for each target volume. For each worker thread, select the target volume, set queue depth
and select one of the saved test specifications, as shown in the table below, to run the corresponding test scenario.
The table also shows expected results for IOPS and Throughput when running these tests. For all scenarios, a
small IO size of 8KB and a high queue depth of 128 is used.
TEST SCENARIO TARGET VOLUME NAME RESULT
Max. Read IOPS CacheReads RandomWrites_8K 50,000 IOPS
Max. Write IOPS NoCacheWrites RandomReads_8K 64,000 IOPS
Max. Combined IOPS CacheReads RandomWrites_8K 100,000 IOPS
NoCacheWrites RandomReads_8K
Max. Read MB/sec CacheReads RandomWrites_64K 524 MB/sec
Max. Write MB/sec NoCacheWrites RandomReads_64K 524 MB/sec
Combined MB/sec CacheReads RandomWrites_64K 1000 MB/sec
NoCacheWrites RandomReads_64K
Below are screenshots of the Iometer test results for combined IOPS and Throughput scenarios.
Combined Reads and Writes Maximum IOPS
Combined Reads and Writes Maximum Throughput
FIO
FIO is a popular tool to benchmark storage on the Linux VMs. It has the flexibility to select different IO sizes,
sequential or random reads and writes. It spawns worker threads or processes to perform the specified I/O
operations. You can specify the type of I/O operations each worker thread must perform using job files. We
created one job file per scenario illustrated in the examples below. You can change the specifications in these job
files to benchmark different workloads running on Premium Storage. In the examples, we are using a Standard DS
14 VM running Ubuntu. Use the same setup described in the beginning of the Benchmarking section and warm
up the cache before running the benchmarking tests.
Before you begin, download FIO and install it on your virtual machine.
Run the following command for Ubuntu,
apt-get install fio
We will use four worker threads for driving Write operations and four worker threads for driving Read operations
on the disks. The Write workers will be driving traffic on the "nocache" volume, which has 10 disks with cache set
to "None". The Read workers will be driving traffic on the "readcache" volume, which has 1 disk with cache set to
"ReadOnly".
Maximum Write IOPS
Create the job file with following specifications to get maximum Write IOPS. Name it "fiowrite.ini".
[global]
size=30g
direct=1
iodepth=256
ioengine=libaio
bs=8k
[writer1]
rw=randwrite
directory=/mnt/nocache
[writer2]
rw=randwrite
[writer3]
rw=randwrite
[writer4]
rw=randwrite
Note the follow key things that are in line with the design guidelines discussed in previous sections. These
specifications are essential to drive maximum IOPS,
A high queue depth of 256.
A small block size of 8KB.
Multiple threads performing random writes.
Run the following command to kick off the FIO test for 30 seconds,
sudo fio --runtime 30 fiowrite.ini
While the test runs, you will be able to see the number of write IOPS the VM and Premium disks are delivering.
As shown in the sample below, the DS14 VM is delivering its maximum write IOPS limit of 50,000 IOPS.
Maximum Read IOPS

Create the job file with following specifications to get maximum Read IOPS. Name it "fioread.ini".
[global]
size=30g
direct=1
iodepth=256
ioengine=libaio
bs=8k
[reader1]
rw=randread
directory=/mnt/readcache
[reader2]
rw=randread
[reader3]
rw=randread
[reader4]
rw=randread
Multiple threads performing random writes.
sudo fio --runtime 30 fioread.ini
While the test runs, you will be able to see the number of read IOPS the VM and Premium disks are delivering. As
shown in the sample below, the DS14 VM is delivering more than 64,000 Read IOPS. This is a combination of the
disk and the cache performance.
Maximum Read and Write IOPS

Create the job file with following specifications to get maximum combined Read and Write IOPS. Name it
"fioreadwrite.ini".
[global]
size=30g
direct=1
iodepth=128
ioengine=libaio
bs=4k
[reader1]
rw=randread
[reader2]
rw=randread
[reader3]
rw=randread
[reader4]
rw=randread
[writer1]
rw=randwrite
rate_iops=12500
[writer2]
rw=randwrite
rate_iops=12500
[writer3]
rw=randwrite
rate_iops=12500
[writer4]
rw=randwrite
rate_iops=12500
Multiple threads performing random reads and writes.
sudo fio --runtime 30 fioreadwrite.ini
While the test runs, you will be able to see the number of combined read and write IOPS the VM and Premium
disks are delivering. As shown in the sample below, the DS14 VM is delivering more than 100,000 combined Read
and Write IOPS. This is a combination of the disk and the cache performance.
Maximum Combined Throughput

To get the maximum combined Read and Write Throughput, use a larger block size and large queue depth with
multiple threads performing reads and writes. You can use a block size of 64KB and queue depth of 128.
Next Steps
Learn more about Azure Premium Storage:
Premium Storage: High-Performance Storage for Azure Virtual Machine Workloads
For SQL Server users, read articles on Performance Best Practices for SQL Server:
Performance Best Practices for SQL Server in Azure Virtual Machines
Azure Premium Storage provides highest performance for SQL Server in Azure VM
Cost-effective Standard Storage and unmanaged and
managed Azure VM disks
Azure Standard Storage delivers reliable, low -cost disk support for VMs running latency-insensitive workloads. It
also supports blobs, tables, queues, and files. With Standard Storage, the data is stored on hard disk drives
(HDDs). When working with VMs, you can use standard SSD and HDD disks for Dev/Test scenarios and less
critical workloads, and premium SSD disks for mission-critical production applications. Standard Storage is
available in all Azure regions.
This article focuses on the use of standard SSD and HDD disks. For more information about the use of storage
with blobs, tables, queues, and files, see Introduction to Storage.
Disk types
There are two ways to create standard disks for Azure VMs:
Unmanaged disks: This type of disk is the original method where you manage the storage accounts used to
store the VHD files that correspond to the VM disks. VHD files are stored as page blobs in storage accounts.
Unmanaged disks can be attached to any Azure VM size, including the VMs that primarily use Premium Storage,
such as the DSv2 and GS series. Azure VMs support attaching several standard disks, allowing up to 256 TiB of
storage per VM. If you use the preview disk sizes, you can have up to around 2 PiB of storage per VM.
Azure Managed Disks: This feature manages the storage accounts used for the VM disks for you. You specify
the type (Premium SSD, Standard SSD, or Standard HDD ) and size of disk you need, and Azure creates and
manages the disk for you. You don't have to worry about placing the disks across multiple storage accounts in
order to ensure you stay within the scalability limits for the storage accounts -- Azure handles that for you.
Even though both types of disks are available, we recommend using Managed Disks to take advantage of their
many features.
To get started with Azure Standard Storage, visit Get started for free.
For information on how to create a VM with Managed Disks, see one of the following articles.
Standard Storage Features

Let's take a look at some of the features of Standard Storage. For more details, please see Introduction to Azure
Storage.
Standard Storage: Azure Standard Storage supports Azure Disks, Azure Blobs, Azure Files, Azure Tables, and
Azure Queues. To use Standard Storage services, start with Create an Azure Storage account.
Standard SSD disks: Standard SSD disks provide more reliable performance than Standard HDD disks, and are
currently available. For more information about region availability of Standard SSD disks, see region availability
of Standard SSD disks.
Standard HDD disks: Standard HDD disks can be attached to all Azure VMs including size-series VMs used with
Premium Storage such as the DSv2 and GS series. A Standard HDD disk can only be attached to one VM.
However, you can attach one or more of these disks to a VM, up to the maximum disk count defined for that VM
size. In the following section on Standard Storage Scalability and Performance Targets, we describe the
specifications in more detail.
Standard page blob: Standard page blobs are used to hold persistent disks for VMs and can also be accessed
directly through REST like other types of Azure Blobs. Page blobs are a collection of 512-byte pages optimized for
random read and write operations.
Storage Replication: In most regions, data in a standard storage account can be replicated locally or geo-
replicated across multiple data centers. The four types of replication available are Locally-Redundant Storage
(LRS ), Zone-Redundant Storage (ZRS ), Geo-Redundant Storage (GRS ), and Read Access Geo-Redundant Storage
(RA-GRS ). Managed Disks in Standard Storage currently support Locally-Redundant Storage (LRS ) only. For
more information, please see Storage Replication.
Scalability and Performance Targets

In this section, we describe the Scalability and Performance targets you need to consider when using standard
storage.
Account limits – does not apply to managed disks
RESOURCE DEFAULT LIMIT
TB per storage account 500 TB
Max ingress1 per storage account (US Regions) 10 Gbps if GRS/ZRS enabled, 20 Gbps for LRS
Max egress1 per storage account (US Regions) 20 Gbps if RA-GRS/GRS/ZRS enabled, 30 Gbps for LRS
Max ingress1 per storage account (European and Asian 5 Gbps if GRS/ZRS enabled, 10 Gbps for LRS
Regions)
Max egress1 per storage account (European and Asian 10 Gbps if RA-GRS/GRS/ZRS enabled, 15 Gbps for LRS
Regions)
Total Request Rate (assuming 1 KB object size) per storage Up to 20,000 IOPS, entities per second, or messages per
account second
1 Ingress refers to all data (requests) being sent to a storage account. Egress refers to all data (responses) being
received from a storage account.

For more information, see Azure Storage Scalability and Performance Targets.
If the needs of your application exceed the scalability targets of a single storage account, build your application to
use multiple storage accounts and partition your data across those storage accounts. Alternately, you can Azure
Managed Disks, and Azure manages the partitioning and placement of your data for you.
Standard Disks Limits
Unlike Premium Disks, the input/output operations per second (IOPS ) and throughput (bandwidth) of Standard
Disks are not provisioned. The performance of standard disks varies with the VM size to which the disk is
attached, not to the size of the disk. You could expect to achieve up to the performance limit listed in the table
below.
Standard disks limits (managed and unmanaged)
VM TIER BASIC TIER VM STANDARD TIER VM
Max Disk size 32,767 GiB 32,767 GiB
Max 8 KB IOPS per disk Up to 2,000 Up to 2,000
Max Bandwidth per disk Up to 500 MB/s Up to 500 MB/s
If your workload requires high-performance, low -latency disk support, you should consider using Premium
Storage. To know more benefits of Premium Storage, visit High-Performance Premium Storage and Azure VM
Disks.
Snapshots and copy blob

To the Storage service, the VHD file is a page blob. You can take snapshots of page blobs, and copy them to
another location, such as a different storage account.
Unmanaged disks
You can create incremental snapshots for unmanaged standard disks in the same way you use snapshots with
Standard Storage. We recommend that you create snapshots and then copy those snapshots to a geo-redundant
standard storage account if your source disk is in a locally-redundant storage account. For more information, see
Azure Storage Redundancy Options.
If a disk is attached to a VM, certain API operations are not permitted on the disks. For example, you cannot
perform a Copy Blob operation on that blob as long as the disk is attached to a VM. Instead, first create a
snapshot of that blob by using the Snapshot Blob REST API method, and then perform the Copy Blob of the
snapshot to copy the attached disk. Alternatively, you can detach the disk and then perform any necessary
operations.
To maintain geo-redundant copies of your snapshots, you can copy snapshots from a locally-redundant storage
account to a geo-redundant standard storage account by using AzCopy or Copy Blob. For more information, see
Transfer data with the AzCopy Command-Line Utility and Copy Blob.
For detailed information on performing REST operations against page blobs in standard storage accounts, see
Azure Storage Services REST API.
Managed disks
A snapshot for a managed disk is a read-only copy of the managed disk which is stored as a standard managed
disk. Incremental Snapshots are not currently supported for Managed Disks but will be supported in the future.
If a managed disk is attached to a VM, certain API operations are not permitted on the disks. For example, you
cannot generate a shared access signature (SAS ) to perform a copy operation while the disk is attached to a VM.
Instead, first create a snapshot of the disk, and then perform the copy of the snapshot. Alternately, you can detach
the disk and then generate a shared access signature (SAS ) to perform the copy operation.
Pricing and Billing

When using Standard Storage, the following billing considerations apply:
Standard storage unmanaged disks/data size
Standard managed disks
Standard storage snapshots
Transactions
Unmanaged storage data and disk size: For unmanaged disks and other data (blobs, tables, queues, and files),
you are charged only for the amount of space you are using. For example, if you have a VM whose page blob is
provisioned as 127 GB, but the VM is really only using 10 GB of space, you are billed for 10 GB of space. We
support Standard storage up to 8191 GB, and Standard unmanaged disks up to 4095 GB.
Managed disks: Billing for standard managed disks depends on the provisioned size of the disk. Azure maps the
provisioned size (rounded up) to the nearest Managed Disks option as specified in the tables below. Each
managed disk maps to one of the supported provisioned sizes and is billed accordingly. For example, if you create
a standard managed disk and specify a provisioned size of 200 GiB, you are billed as per the pricing of the S15
Disk type.
STAN
DARD
HDD
MANA
GED
DISK
Disk 32 64 128 256 512 1,024 2,048 4,095 8,192 16,38 32,76
TiB) TiB)
Snapshots: Snapshots of standard disks are billed for the additional capacity used by the snapshots. For
information on snapshots, see Creating a Snapshot of a Blob.
bandwidth usage.
Transaction: Azure charges $0.0036 per 100,000 transactions for standard storage. Transactions include both
read and write operations to storage.
For detailed information on pricing for Standard Storage, Virtual Machines, and Managed Disks, see:
Azure Storage Pricing
Virtual Machines Pricing
Managed Disks Pricing
Azure Backup service support

Virtual machines with unmanaged disks can be backed up using Azure Backup. More details.
You can also use the Azure Backup service with Managed Disks to create a backup job with time-based backups,
easy VM restoration and backup retention policies. You can read more about this at Using Azure Backup service
for VMs with Managed Disks.
Next steps
Introduction to Azure Storage
Create a storage account
Managed Disks Overview
Scalability and performance targets for VM disks on
Windows
An Azure virtual machine supports attaching a number of data disks. This article describes scalability and
performance targets for a VM's data disks. Use these targets to help decide the number and type of disk that you
need to meet your performance and capacity requirements.
IMPORTANT
For optimal performance, limit the number of highly utilized disks attached to the virtual machine to avoid possible throttling.
If all attached disks are not highly utilized at the same time, then the virtual machine can support a larger number of disks.
For Azure Managed Disks:
RESOURCE DEFAULT LIMIT MAXIMUM LIMIT
Standard Managed Disks 10,000 50,000
Standard SSD Managed Disks 10,000 50,000
Premium Managed Disks 10,000 50,000
Standard_LRS Snapshots 10,000 50,000
Standard_ZRS Snapshots 10,000 50,000
Premium_LRS Snapshots 10,000 50,000
Managed Image 10,000 50,000
For standard storage accounts: A standard storage account has a maximum total request rate of 20,000
IOPS. The total IOPS across all of your virtual machine disks in a standard storage account should not
exceed this limit.
You can roughly calculate the number of highly utilized disks supported by a single standard storage
account based on the request rate limit. For example, for a Basic Tier VM, the maximum number of highly
utilized disks is about 66 (20,000/300 IOPS per disk), and for a Standard Tier VM, it is about 40 (20,000/500
IOPS per disk).
For premium storage accounts: A premium storage account has a maximum total throughput rate of 50
Gbps. The total throughput across all of your VM disks should not exceed this limit.
See Windows VM sizes for additional details.
Managed virtual machine disks

Standard managed virtual machine HDDs
STAND
ARD
DISK
Disk 32 64 128 256 512 1,024 2,048 4,095 8,192 16,38 32,76
size in 4 7
GiB
IOPS Up to Up to Up to Up to Up to Up to Up to Up to Up to Up to Up to
per 500 500 500 500 500 500 500 500 1,300 2,000 2,000
disk
Throu Up to Up to Up to Up to Up to Up to Up to Up to Up to Up to Up to
ghput 60 60 60 60 60 60 60 60 300 500 500
per MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s
disk ec ec ec ec ec ec ec ec ec ec ec
Standard managed virtual machine SSDs
STANDA
RD SSD
DISK
TYPE E10 E15 E20 E30 E40 E50 E60 E70 E80
Disk size 128 256 512 1,024 2,048 4,095 8,192 16,384 32,767
in GiB
IOPS per Up to Up to Up to Up to Up to Up to Up to Up to Up to
disk 500 500 500 500 500 500 500 500 1,300
Through Up to Up to Up to Up to Up to Up to Up to Up to 60 Up to
put per 60 60 60 60 60 60 60 MB/sec 300
disk MB/sec MB/sec MB/sec MB/sec MB/sec MB/sec MB/sec MiB/sec
Premium managed virtual machine disks: per disk limits
PREMI
UM
DISK
TYPE P4 P6 P10 P15 P20 S30 P40 P50 P60 P70 P80
Disk 32 64 128 256 512 1,024 2,048 4,095 8,192 16,38 32,76
size in 4 7
GiB
IOPS Up to Up to Up to Up to Up to Up to Up to Up to Up to Up to Up to
per 120 240 500 1,100 2,300 5,000 7,500 7,500 12,50 15,00 20,00
disk 0 0 0
Throu Up to Up to Up to Up to Up to Up to Up to Up to Up to Up to Up to
ghput 25 50 100 125 150 200 250 250 480 750 750
per MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s MiB/s
disk ec ec ec ec ec ec ec ec ec ec ec
Premium managed virtual machine disks: per VM limits

Max IOPS Per VM 80,000 IOPS with GS5 VM
Max throughput per VM 2,000 MB/s with GS5 VM
Unmanaged virtual machine disks

Standard unmanaged virtual machine disks: per disk limits
VM TIER BASIC TIER VM STANDARD TIER VM
Disk size 4095 GB 4095 GB
Max 8 KB IOPS per persistent disk 300 500
Max number of disks performing max 66 40

IOPS
Premium unmanaged virtual machine disks: per account limits
Total disk capacity per account 35 TB
Total snapshot capacity per account 10 TB
Max bandwidth per account (ingress + egress1 ) <=50 Gbps
1Ingress refers to all data (requests) being sent to a storage account. Egress refers to all data (responses) being
received from a storage account.

Premium unmanaged virtual machine disks: per disk limits
PREMIUM
STORAGE DISK
TYPE P10 P20 P30 P40 P50
Disk size 128 GiB 512 GiB 1024 GiB (1 TB) 2048 GiB (2 TB) 4095 GiB (4 TB)
Max IOPS per 500 2300 5000 7500 7500

disk
Max throughput 100 MB/s 150 MB/s 200 MB/s 250 MB/s 250 MB/s
per disk
Max number of 280 70 35 17 8

disks per storage
account
Premium unmanaged virtual machine disks: per VM limits

Max IOPS Per VM 80,000 IOPS with GS5 VM
Max throughput per VM 2,000 MB/s with GS5 VM
See also
Azure subscription and service limits, quotas, and constraints
Backup and disaster recovery for Azure IaaS disks
This article explains how to plan for backup and disaster recovery (DR ) of IaaS virtual machines (VMs) and disks in
Azure. This document covers both managed and unmanaged disks.
First, we cover the built-in fault tolerance capabilities in the Azure platform that helps guard against local failures.
We then discuss the disaster scenarios not fully covered by the built-in capabilities. We also show several examples
of workload scenarios where different backup and DR considerations can apply. We then review possible solutions
for the DR of IaaS disks.
Introduction
The Azure platform uses various methods for redundancy and fault tolerance to help protect customers from
localized hardware failures. Local failures can include problems with an Azure Storage server machine that stores
part of the data for a virtual disk or failures of an SSD or HDD on that server. Such isolated hardware component
failures can happen during normal operations.
The Azure platform is designed to be resilient to these failures. Major disasters can result in failures or the
inaccessibility of many storage servers or even a whole datacenter. Although your VMs and disks are normally
protected from localized failures, additional steps are necessary to protect your workload from region-wide
catastrophic failures, such as a major disaster, that can affect your VM and disks.
In addition to the possibility of platform failures, problems with a customer application or data can occur. For
example, a new version of your application might inadvertently make a change to the data that causes it to break.
In that case, you might want to revert the application and the data to a prior version that contains the last known
good state. This requires maintaining regular backups.
For regional disaster recovery, you must back up your IaaS VM disks to a different region.
Before we look at backup and DR options, let’s recap a few methods available for handling localized failures.
Azure IaaS resiliency
Resiliency refers to the tolerance for normal failures that occur in hardware components. Resiliency is the ability to
recover from failures and continue to function. It's not about avoiding failures, but responding to failures in a way
that avoids downtime or data loss. The goal of resiliency is to return the application to a fully functioning state
following a failure. Azure virtual machines and disks are designed to be resilient to common hardware faults. Let's
look at how the Azure IaaS platform provides this resiliency.
A virtual machine consists mainly of two parts: a compute server and the persistent disks. Both affect the fault
tolerance of a virtual machine.
If the Azure compute host server that houses your VM experiences a hardware failure, which is rare, Azure is
designed to automatically restore the VM on another server. If this scenario, your computer reboots, and the VM
comes back up after some time. Azure automatically detects such hardware failures and executes recoveries to help
ensure the customer VM is available as soon as possible.
Regarding IaaS disks, the durability of data is critical for a persistent storage platform. Azure customers have
important business applications running on IaaS, and they depend on the persistence of the data. Azure designs
protection for these IaaS disks, with three redundant copies of the data that is stored locally. These copies provide
for high durability against local failures. If one of the hardware components that holds your disk fails, your VM is
not affected, because there are two additional copies to support disk requests. It works fine, even if two different
hardware components that support a disk fail at the same time (which is rare).
To ensure that you always maintain three replicas, Azure Storage automatically spawns a new copy of the data in
the background if one of the three copies becomes unavailable. Therefore, it should not be necessary to use RAID
with Azure disks for fault tolerance. A simple RAID 0 configuration should be sufficient for striping the disks, if
necessary, to create larger volumes.
Because of this architecture, Azure has consistently delivered enterprise-grade durability for IaaS disks, with an
industry-leading zero percent annualized failure rate.
Localized hardware faults on the compute host or in the Storage platform can sometimes result in of the
temporary unavailability of the VM that is covered by the Azure SLA for VM availability. Azure also provides an
industry-leading SLA for single VM instances that use Azure Premium SSD disks.
To safeguard application workloads from downtime due to the temporary unavailability of a disk or VM, customers
can use availability sets. Two or more virtual machines in an availability set provide redundancy for the application.
Azure then creates these VMs and disks in separate fault domains with different power, network, and server
components.
Because of these separate fault domains, localized hardware failures typically do not affect multiple VMs in the set
at the same time. Having separate fault domains provides high availability for your application. It's considered a
good practice to use availability sets when high availability is required. The next section covers the disaster
recovery aspect.
Backup and disaster recovery
Disaster recovery is the ability to recover from rare, but major, incidents. These incidents include non-transient,
wide-scale failures, such as service disruption that affects an entire region. Disaster recovery includes data backup
and archiving, and might include manual intervention, such as restoring a database from a backup.
The Azure platform’s built-in protection against localized failures might not fully protect the VMs/disks if a major
disaster causes large-scale outages. These large-scale outages include catastrophic events, such as if a datacenter is
hit by a hurricane, earthquake, fire, or if there is a large-scale hardware unit failure. In addition, you might
encounter failures due to application or data issues.
To help protect your IaaS workloads from outages, you should plan for redundancy and have backups to enable
recovery. For disaster recovery, you should back up in a different geographic location away from the primary site.
This approach helps ensure your backup is not affected by the same event that originally affected the VM or disks.
For more information, see Disaster recovery for Azure applications.
Your DR considerations might include the following aspects:
High availability: The ability of the application to continue running in a healthy state, without significant
downtime. By healthy state, this state means that the application is responsive, and users can connect to the
application and interact with it. Certain mission-critical applications and databases might be required to
always be available, even when there are failures in the platform. For these workloads, you might need to
plan redundancy for the application, as well as the data.
Data durability: In some cases, the main consideration is ensuring that the data is preserved if a disaster
happens. Therefore, you might need a backup of your data in a different site. For such workloads, you might
not need full redundancy for the application, but only a regular backup of the disks.
Backup and DR scenarios

Let’s look at a few typical examples of application workload scenarios and the considerations for planning for
disaster recovery.
Scenario 1: Major database solutions
Consider a production database server, like SQL Server or Oracle, that can support high availability. Critical
production applications and users depend on this database. The disaster recovery plan for this system might need
to support the following requirements:
The data must be protected and recoverable.
The server must be available for use.
The disaster recovery plan might require maintaining a replica of the database in a different region as a backup.
Depending on the requirements for server availability and data recovery, the solution might range from an active-
active or active-passive replica site to periodic offline backups of the data. Relational databases, such as SQL Server
and Oracle, provide various options for replication. For SQL Server, use SQL Server AlwaysOn Availability Groups
for high availability.
NoSQL databases, like MongoDB, also support replicas for redundancy. The replicas for high availability are used.
Scenario 2: A cluster of redundant VMs
Consider a workload handled by a cluster of VMs that provide redundancy and load balancing. One example is a
Cassandra cluster deployed in a region. This type of architecture already provides a high level of redundancy within
that region. However, to protect the workload from a regional-level failure, you should consider spreading the
cluster across two regions or making periodic backups to another region.
Scenario 3: IaaS application workload
Let's look at the IaaS application workload. For example, this application might be a typical production workload
running on an Azure VM. It might be a web server or file server holding the content and other resources of a site. It
might also be a custom-built business application running on a VM that stored its data, resources, and application
state on the VM disks. In this case, it's important to make sure you take backups on a regular basis. Backup
frequency should be based on the nature of the VM workload. For example, if the application runs every day and
modifies data, then the backup should be taken every hour.
Another example is a reporting server that pulls data from other sources and generates aggregated reports. The
loss of this VM or disks might lead to the loss of the reports. However, it might be possible to rerun the reporting
process and regenerate the output. In that case, you don’t really have a loss of data, even if the reporting server is
hit with a disaster. As a result, you might have a higher level of tolerance for losing part of the data on the reporting
server. In that case, less frequent backups are an option to reduce costs.
Scenario 4: IaaS application data issues
IaaS application data issues are another possibility. Consider an application that computes, maintains, and serves
critical commercial data, such as pricing information. A new version of your application had a software bug that
incorrectly computed the pricing and corrupted the existing commerce data served by the platform. Here, the best
course of action is to revert to the earlier version of the application and the data. To enable this, take periodic
backups of your system.
Disaster recovery solution: Azure Backup

Azure Backup is used for backups and DR, and it works with managed disks as well as unmanaged disks. You can
create a backup job with time-based backups, easy VM restoration, and backup retention policies.
If you use Premium SSD disks, managed disks, or other disk types with the locally redundant storage option, it's
especially important to make periodic DR backups. Azure Backup stores the data in your recovery services vault
for long-term retention. Choose the geo-redundant storage option for the backup recovery services vault. That
option ensures that backups are replicated to a different Azure region for safeguarding from regional disasters.
For unmanaged disks, you can use the locally redundant storage type for IaaS disks, but ensure that Azure Backup
is enabled with the geo-redundant storage option for the recovery services vault.
NOTE
If you use the geo-redundant storage or read-access geo-redundant storage option for your unmanaged disks, you still need
consistent snapshots for backup and DR. Use either Azure Backup or consistent snapshots.
The following table is a summary of the solutions available for DR.
SCENARIO AUTOMATIC REPLICATION DR SOLUTION
Premium SSD disks Local (locally redundant storage) Azure Backup
Managed disks Local (locally redundant storage) Azure Backup
Unmanaged locally redundant storage Local (locally redundant storage) Azure Backup
disks
Unmanaged geo-redundant storage Cross region (geo-redundant storage) Azure Backup

disks Consistent snapshots
Unmanaged read-access geo- Cross region (read-access geo- Azure Backup

redundant storage disks redundant storage) Consistent snapshots
High availability is best met by using managed disks in an availability set along with Azure Backup. If you use
unmanaged disks, you can still use Azure Backup for DR. If you are unable to use Azure Backup, then taking
consistent snapshots, as described in a later section, is an alternative solution for backup and DR.
Your choices for high availability, backup, and DR at application or infrastructure levels can be represented as
follows:
LEVEL HIGH AVAILABILITY BACKUP OR DR
Application SQL Server AlwaysOn Azure Backup
Infrastructure Availability set Geo-redundant storage with consistent

snapshots
Using Azure Backup

Azure Backup can back up your VMs running Windows or Linux to the Azure recovery services vault. Backing up
and restoring business-critical data is complicated by the fact that business-critical data must be backed up while
the applications that produce the data are running.
To address this issue, Azure Backup provides application-consistent backups for Microsoft workloads. It uses the
volume shadow service to ensure that data is written correctly to storage. For Linux VMs, only file-consistent
backups are possible, because Linux does not have functionality equivalent to the volume shadow service.
When Azure Backup initiates a backup job at the scheduled time, it triggers the backup extension installed in the
VM to take a point-in-time snapshot. A snapshot is taken in coordination with the volume shadow service to get a
consistent snapshot of the disks in the virtual machine without having to shut it down. The backup extension in the
VM flushes all writes before taking a consistent snapshot of all of the disks. After taking the snapshot, the data is
transferred by Azure Backup to the backup vault. To make the backup process more efficient, the service identifies
and transfers only the blocks of data that have changed after the last backup.
To restore, you can view the available backups through Azure Backup and then initiate a restore. You can create and
restore Azure backups through the Azure portal, by using PowerShell, or by using the Azure CLI.
Steps to enable a backup
Use the following steps to enable backups of your VMs by using the Azure portal. There is some variation
depending on your exact scenario. Refer to the Azure Backup documentation for full details. Azure Backup also
supports VMs with managed disks.
1. Create a recovery services vault for a VM:
a. In the Azure portal, browse All resources and find Recovery Services vaults.
b. On the Recovery Services vaults menu, click Add and follow the steps to create a new vault in the same
region as the VM. For example, if your VM is in the West US region, pick West US for the vault.
2. Verify the storage replication for the newly created vault. Access the vault under Recovery Services vaults
and go to Settings > Backup Configuration. Ensure the geo-redundant storage option is selected by
default. This option ensures that your vault is automatically replicated to a secondary datacenter. For
example, your vault in West US is automatically replicated to East US.
3. Configure the backup policy and select the VM from the same UI.
4. Make sure the Backup Agent is installed on the VM. If your VM is created by using an Azure gallery image,
then the Backup Agent is already installed. Otherwise (that is, if you use a custom image), use the
instructions to install the VM agent on a virtual machine.
5. Make sure that the VM allows network connectivity for the backup service to function. Follow the
instructions for network connectivity.
6. After the previous steps are completed, the backup runs at regular intervals as specified in the backup policy.
If necessary, you can trigger the first backup manually from the vault dashboard on the Azure portal.
For automating Azure Backup by using scripts, refer to PowerShell cmdlets for VM backup.
Steps for recovery
If you need to repair or rebuild a VM, you can restore the VM from any of the backup recovery points in the vault.
There are a couple of different options for performing the recovery:
You can create a new VM as a point-in-time representation of your backed-up VM.
You can restore the disks, and then use the template for the VM to customize and rebuild the restored VM.
For more information, see the instructions to use the Azure portal to restore virtual machines . This document also
explains the specific steps for restoring backed-up VMs to a paired datacenter by using your geo-redundant
backup vault if there is a disaster at the primary datacenter. In that case, Azure Backup uses the Compute service
from the secondary region to create the restored virtual machine.
You can also use PowerShell for restoring a VM or for creating a new VM from restored disks.
Alternative solution: Consistent snapshots

If you are unable to use Azure Backup, you can implement your own backup mechanism by using snapshots.
Creating consistent snapshots for all the disks used by a VM and then replicating those snapshots to another
region is complicated. For this reason, Azure considers using the Backup service as a better option than building a
custom solution.
If you use read-access geo-redundant storage/geo-redundant storage for disks, snapshots are automatically
replicated to a secondary datacenter. If you use locally redundant storage for disks, you need to replicate the data
yourself. For more information, see Back up Azure-unmanaged VM disks with incremental snapshots.
A snapshot is a representation of an object at a specific point in time. A snapshot incurs billing for the incremental
size of the data it holds. For more information, see Create a blob snapshot.
Create snapshots while the VM is running
Although you can take a snapshot at any time, if the VM is running, there is still data being streamed to the disks.
The snapshots might contain partial operations that were in flight. Also, if there are several disks involved, the
snapshots of different disks might have occurred at different times. These scenarios may cause to the snapshots to
be uncoordinated. This lack of co-ordination is especially problematic for striped volumes whose files might be
corrupted if changes were being made during backup.
To avoid this situation, the backup process must implement the following steps:
1. Freeze all the disks.
2. Flush all the pending writes.
3. Create a blob snapshot for all the disks.
Some Windows applications, like SQL Server, provide a coordinated backup mechanism via a volume shadow
service to create application-consistent backups. On Linux, you can use a tool like fsfreeze for coordinating the
disks. This tool provides file-consistent backups, but not application-consistent snapshots. This process is complex,
so you should consider using Azure Backup or a third-party backup solution that already implements this
procedure.
The previous process results in a collection of coordinated snapshots for all of the VM disks, representing a specific
point-in-time view of the VM. This is a backup restore point for the VM. You can repeat the process at scheduled
intervals to create periodic backups. See Copy the backups to another region for steps to copy the snapshots to
another region for DR.
Create snapshots while the VM is offline
Another option to create consistent backups is to shut down the VM and take blob snapshots of each disk. Taking
blob snapshots is easier than coordinating snapshots of a running VM, but it requires a few minutes of downtime.
1. Shut down the VM.
2. Create a snapshot of each virtual hard drive blob, which only takes a few seconds.
To create a snapshot, you can use PowerShell, the Azure Storage REST API, Azure CLI, or one of the Azure
Storage client libraries, such as the Storage client library for .NET.
3. Start the VM, which ends the downtime. Typically, the entire process finishes within a few minutes.
This process yields a collection of consistent snapshots for all the disks, providing a backup restore point for the
VM.
Copy the snapshots to another region
Creation of the snapshots alone might not be sufficient for DR. You must also replicate the snapshot backups to
another region.
If you use geo-redundant storage or read-access geo-redundant storage for your disks, then the snapshots are
replicated to the secondary region automatically. There can be a few minutes of lag before the replication. If the
primary datacenter goes down before the snapshots finish replicating, you cannot access the snapshots from the
secondary datacenter. The likelihood of this is small.
NOTE
Only having the disks in a geo-redundant storage or read-access geo-redundant storage account does not protect the VM
from disasters. You must also create coordinated snapshots or use Azure Backup. This is required to recover a VM to a
consistent state.
If you use locally redundant storage, you must copy the snapshots to a different storage account immediately after
creating the snapshot. The copy target might be a locally redundant storage account in a different region, resulting
in the copy being in a remote region. You can also copy the snapshot to a read-access geo-redundant storage
account in the same region. In this case, the snapshot is lazily replicated to the remote secondary region. Your
backup is protected from disasters at the primary site after the copying and replication is complete.
To copy your incremental snapshots for DR efficiently, review the instructions in Back up Azure unmanaged VM
disks with incremental snapshots.
Recovery from snapshots

To retrieve a snapshot, copy it to make a new blob. If you are copying the snapshot from the primary account, you
can copy the snapshot over to the base blob of the snapshot. This process reverts the disk to the snapshot. This
process is known as promoting the snapshot. If you are copying the snapshot backup from a secondary account, in
the case of a read-access geo-redundant storage account, you must copy it to a primary account. You can copy a
snapshot by using PowerShell or by using the AzCopy utility. For more information, see Transfer data with the
AzCopy command-line utility.
For VMs with multiple disks, you must copy all the snapshots that are part of the same coordinated restore point.
After you copy the snapshots to writable VHD blobs, you can use the blobs to recreate your VM by using the
template for the VM.
Other options
SQL Server
SQL Server running in a VM has its own built-in capabilities to back up your SQL Server database to Azure Blob
storage or a file share. If the storage account is geo-redundant storage or read-access geo-redundant storage, you
can access those backups in the storage account’s secondary datacenter in the event of a disaster, with the same
restrictions as previously discussed. For more information, see Back up and restore for SQL Server in Azure virtual
machines. In addition to back up and restore, SQL Server AlwaysOn availability groups can maintain secondary
replicas of databases. This ability greatly reduces the disaster recovery time.
Other considerations
This article has discussed how to back up or take snapshots of your VMs and their disks to support disaster
recovery and how to use those backups or snapshots to recover your data. With the Azure Resource Manager
model, many people use templates to create their VMs and other infrastructures in Azure. You can use a template
to create a VM that has the same configuration every time. If you use custom images for creating your VMs, you
must also make sure that your images are protected by using a read-access geo-redundant storage account to
store them.
Consequently, your backup process can be a combination of two things:
Back up the data (disks).
Back up the configuration (templates and custom images).
Depending on the backup option you choose, you might have to handle the backup of both the data and the
configuration, or the backup service might handle all of that for you.
Appendix: Understanding the impact of data redundancy

For storage accounts in Azure, there are three types of data redundancy that you should consider regarding
disaster recovery: locally redundant, geo-redundant, or geo-redundant with read access.
Locally redundant storage retains three copies of the data in the same datacenter. When the VM writes the data, all
three copies are updated before success is returned to the caller, so you know they are identical. Your disk is
protected from local failures, because it's unlikely that all three copies are affected at the same time. In the case of
locally redundant storage, there is no geo-redundancy, so the disk is not protected from catastrophic failures that
can affect an entire datacenter or storage unit.
With geo-redundant storage and read-access geo-redundant storage, three copies of your data are retained in the
primary region that is selected by you. Three more copies of your data are retained in a corresponding secondary
region that is set by Azure. For example, if you store data in West US, the data is replicated to East US. Copy
retention is done asynchronously, and there is a small delay between updates to the primary and secondary sites.
Replicas of the disks on the secondary site are consistent on a per-disk basis (with the delay), but replicas of
multiple active disks might not be in sync with each other. To have consistent replicas across multiple disks,
consistent snapshots are needed.
The main difference between geo-redundant storage and read-access geo-redundant storage is that with read-
access geo-redundant storage, you can read the secondary copy at any time. If there is a problem that renders the
data in the primary region inaccessible, the Azure team makes every effort to restore access. While the primary is
down, if you have read-access geo-redundant storage enabled, you can access the data in the secondary datacenter.
Therefore, if you plan to read from the replica while the primary is inaccessible, then read-access geo-redundant
storage should be considered.
If it turns out to be a significant outage, the Azure team might trigger a geo-failover and change the primary DNS
entries to point to secondary storage. At this point, if you have either geo-redundant storage or read-access geo-
redundant storage enabled, you can access the data in the region that used to be the secondary. In other words, if
your storage account is geo-redundant storage and there is a problem, you can access the secondary storage only
if there is a geo-failover.
For more information, see What to do if an Azure Storage outage occurs.
NOTE
Microsoft controls whether a failover occurs. Failover is not controlled per storage account, so it's not decided by individual
customers. To implement disaster recovery for specific storage accounts or virtual machine disks, you must use the
techniques described previously in this article.
Virtual networks and virtual machines in Azure
When you create an Azure virtual machine (VM ), you must create a virtual network (VNet) or use an existing VNet.
You also need to decide how your VMs are intended to be accessed on the VNet. It is important to plan before
creating resources and make sure that you understand the limits of networking resources.
In the following figure, VMs are represented as web servers and database servers. Each set of VMs are assigned to
separate subnets in the VNet.
You can create a VNet before you create a VM or you can as you create a VM. You create these resources to
support communication with a VM:
Network interfaces
IP addresses
Virtual network and subnets
In addition to those basic resources, you should also consider these optional resources:
Network security groups
Load balancers
Network interfaces
A network interface (NIC ) is the interconnection between a VM and a virtual network (VNet). A VM must have at
least one NIC, but can have more than one, depending on the size of the VM you create. Learn about how many
NICs each VM size supports for Windows or Linux.
You can create a VM with multiple NICs, and add or remove NICs through the lifecycle of a VM. Multiple NICs
allow a VM to connect to different subnets and send or receive traffic over the most appropriate interface.
If the VM is added to an availability set, all VMs within the availability set must have one or multiple NICs. VMs
with more than one NIC aren’t required to have the same number of NICs, but they must all have at least two.
Each NIC attached to a VM must exist in the same location and subscription as the VM. Each NIC must be
connected to a VNet that exists in the same Azure location and subscription as the NIC. You can change the subnet
a VM is connected to after it's created, but you cannot change the VNet. Each NIC attached to a VM is assigned a
MAC address that doesn’t change until the VM is deleted.
This table lists the methods that you can use to create a network interface.
METHOD DESCRIPTION
Azure portal When you create a VM in the Azure portal, a network

interface is automatically created for you (you cannot use a
NIC you create separately). The portal creates a VM with only
one NIC. If you want to create a VM with more than one NIC,
you must create it with a different method.
Azure PowerShell Use New-AzureRmNetworkInterface with the -

PublicIpAddressId parameter to provide the identifier of the
public IP address that you previously created.
Azure CLI To provide the identifier of the public IP address that you
previously created, use az network nic create with the --
public-ip-address parameter.
Template Use Network Interface in a Virtual Network with Public IP

Address as a guide for deploying a network interface using a
template.
IP addresses
You can assign these types of IP addresses to a NIC in Azure:
Public IP addresses - Used to communicate inbound and outbound (without network address translation
(NAT)) with the Internet and other Azure resources not connected to a VNet. Assigning a public IP address to a
NIC is optional. Public IP addresses have a nominal charge, and there's a maximum number that can be used
per subscription.
Private IP addresses - Used for communication within a VNet, your on-premises network, and the Internet
(with NAT). You must assign at least one private IP address to a VM. To learn more about NAT in Azure, read
Understanding outbound connections in Azure.
You can assign public IP addresses to VMs or internet-facing load balancers. You can assign private IP addresses to
VMs and internal load balancers. You assign IP addresses to a VM using a network interface.
There are two methods in which an IP address is allocated to a resource - dynamic or static. The default allocation
method is dynamic, where an IP address is not allocated when it's created. Instead, the IP address is allocated when
you create a VM or start a stopped VM. The IP address is released when you stop or delete the VM.
To ensure the IP address for the VM remains the same, you can set the allocation method explicitly to static. In this
case, an IP address is assigned immediately. It is released only when you delete the VM or change its allocation
method to dynamic.
This table lists the methods that you can use to create an IP address.
METHOD DESCRIPTION
Azure portal By default, public IP addresses are dynamic and the address
associated to them may change when the VM is stopped or
deleted. To guarantee that the VM always uses the same
public IP address, create a static public IP address. By default,
the portal assigns a dynamic private IP address to a NIC when
creating a VM. You can change this IP address to static after
the VM is created.
Azure PowerShell You use New-AzureRmPublicIpAddress with the -

AllocationMethod parameter as Dynamic or Static.
Azure CLI You use az network public-ip create with the --allocation-
method parameter as Dynamic or Static.
Template Use Network Interface in a Virtual Network with Public IP

Address as a guide for deploying a public IP address using a
template.
After you create a public IP address, you can associate it with a VM by assigning it to a NIC.

A subnet is a range of IP addresses in the VNet. You can divide a VNet into multiple subnets for organization and
security. Each NIC in a VM is connected to one subnet in one VNet. NICs connected to subnets (same or different)
within a VNet can communicate with each other without any extra configuration.
When you set up a VNet, you specify the topology, including the available address spaces and subnets. If the VNet
is to be connected to other VNets or on-premises networks, you must select address ranges that don't overlap. The
IP addresses are private and can't be accessed from the Internet, which was true only for the non-routable IP
addresses such as 10.0.0.0/8, 172.16.0.0/12, or 192.168.0.0/16. Now, Azure treats any address range as part of the
private VNet IP address space that is only reachable within the VNet, within interconnected VNets, and from your
on-premises location.
If you work within an organization in which someone else is responsible for the internal networks, you should talk
to that person before selecting your address space. Make sure there is no overlap and let them know the space you
want to use so they don’t try to use the same range of IP addresses.
By default, there is no security boundary between subnets, so VMs in each of these subnets can talk to one another.
However, you can set up Network Security Groups (NSGs), which allow you to control the traffic flow to and from
subnets and to and from VMs.
This table lists the methods that you can use to create a VNet and subnets.
METHOD DESCRIPTION
Azure portal If you let Azure create a VNet when you create a VM, the
name is a combination of the resource group name that
contains the VNet and -vnet. The address space is
10.0.0.0/24, the required subnet name is default, and the
subnet address range is 10.0.0.0/24.
Azure PowerShell You use New-AzureRmVirtualNetworkSubnetConfig and New-

AzureRmVirtualNetwork to create a subnet and a VNet. You
can also use Add-AzureRmVirtualNetworkSubnetConfig to add
a subnet to an existing VNet.
Azure CLI The subnet and the VNet are created at the same time.
Provide a --subnet-name parameter to az network vnet
create with the subnet name.
Template The easiest way to create a VNet and subnets is to download

an existing template, such as Virtual Network with two
subnets, and modify it for your needs.
Network security groups

A network security group (NSG ) contains a list of Access Control List (ACL ) rules that allow or deny network traffic
to subnets, NICs, or both. NSGs can be associated with either subnets or individual NICs connected to a subnet.
When an NSG is associated with a subnet, the ACL rules apply to all the VMs in that subnet. In addition, traffic to
an individual NIC can be restricted by associating an NSG directly to a NIC.
NSGs contain two sets of rules: inbound and outbound. The priority for a rule must be unique within each set. Each
rule has properties of protocol, source and destination port ranges, address prefixes, direction of traffic, priority,
and access type.
All NSGs contain a set of default rules. The default rules cannot be deleted, but because they are assigned the
lowest priority, they can be overridden by the rules that you create.
When you associate an NSG to a NIC, the network access rules in the NSG are applied only to that NIC. If an NSG
is applied to a single NIC on a multi-NIC VM, it does not affect traffic to the other NICs. You can associate different
NSGs to a NIC (or VM, depending on the deployment model) and the subnet that a NIC or VM is bound to.
Priority is given based on the direction of traffic.
Be sure to plan your NSGs when you plan your VMs and VNet.
This table lists the methods that you can use to create a network security group.
METHOD DESCRIPTION
Azure portal When you create a VM in the Azure portal, an NSG is

automatically created and associated to the NIC the portal
creates. The name of the NSG is a combination of the name of
the VM and -nsg. This NSG contains one inbound rule with a
priority of 1000, service set to RDP, the protocol set to TCP,
port set to 3389, and action set to Allow. If you want to allow
any other inbound traffic to the VM, you must add additional
rules to the NSG.
METHOD DESCRIPTION
Azure PowerShell Use New-AzureRmNetworkSecurityRuleConfig and provide the

required rule information. Use New-
AzureRmNetworkSecurityGroup to create the NSG. Use Set-
AzureRmVirtualNetworkSubnetConfig to configure the NSG
for the subnet. Use Set-AzureRmVirtualNetwork to add the
NSG to the VNet.
Azure CLI Use az network nsg create to initially create the NSG. Use az
network nsg rule create to add rules to the NSG. Use az
network vnet subnet update to add the NSG to the subnet.
Template Use Create a Network Security Group as a guide for deploying

a network security group using a template.
Load balancers
Azure Load Balancer delivers high availability and network performance to your applications. A load balancer can
be configured to balance incoming Internet traffic to VMs or balance traffic between VMs in a VNet. A load
balancer can also balance traffic between on-premises computers and VMs in a cross-premises network, or
forward external traffic to a specific VM.
The load balancer maps incoming and outgoing traffic between the public IP address and port on the load balancer
and the private IP address and port of the VM.
When you create a load balancer, you must also consider these configuration elements:
Front-end IP configuration – A load balancer can include one or more front-end IP addresses, otherwise
known as virtual IPs (VIPs). These IP addresses serve as ingress for the traffic.
Back-end address pool – IP addresses that are associated with the NIC to which load is distributed.
NAT rules - Defines how inbound traffic flows through the front-end IP and distributed to the back-end IP.
Load balancer rules - Maps a given front-end IP and port combination to a set of back-end IP addresses and
port combination. A single load balancer can have multiple load balancing rules. Each rule is a combination of a
front-end IP and port and back-end IP and port associated with VMs.
Probes - Monitors the health of VMs. When a probe fails to respond, the load balancer stops sending new
connections to the unhealthy VM. The existing connections are not affected, and new connections are sent to
healthy VMs.
This table lists the methods that you can use to create an internet-facing load balancer.
METHOD DESCRIPTION
Azure portal You can't currently create an internet-facing load balancer

using the Azure portal.
METHOD DESCRIPTION
Azure PowerShell To provide the identifier of the public IP address that you
previously created, use New-
AzureRmLoadBalancerFrontendIpConfig with the -
PublicIpAddress parameter. Use New-
AzureRmLoadBalancerBackendAddressPoolConfig to create
the configuration of the back-end address pool. Use New-
AzureRmLoadBalancerInboundNatRuleConfig to create
inbound NAT rules associated with the front-end IP
configuration that you created. Use New-
AzureRmLoadBalancerProbeConfig to create the probes that
you need. Use New-AzureRmLoadBalancerRuleConfig to
create the load balancer configuration. Use New-
AzureRmLoadBalancer to create the load balancer.
Azure CLI Use az network lb create to create the initial load balancer
configuration. Use az network lb frontend-ip create to add the
public IP address that you previously created. Use az network
lb address-pool create to add the configuration of the back-
end address pool. Use az network lb inbound-nat-rule create
to add NAT rules. Use az network lb rule create to add the
load balancer rules. Use az network lb probe create to add the
probes.
Template Use 2 VMs in a Load Balancer and configure NAT rules on the
LB as a guide for deploying a load balancer using a template.
This table lists the methods that you can use to create an internal load balancer.
METHOD DESCRIPTION
Azure portal You can't currently create an internal load balancer using the
Azure portal.
Azure PowerShell To provide a private IP address in the network subnet, use

New-AzureRmLoadBalancerFrontendIpConfig with the -
PrivateIpAddress parameter. Use New-
AzureRmLoadBalancerBackendAddressPoolConfig to create
the configuration of the back-end address pool. Use New-
AzureRmLoadBalancerInboundNatRuleConfig to create
inbound NAT rules associated with the front-end IP
configuration that you created. Use New-
AzureRmLoadBalancerProbeConfig to create the probes that
you need. Use New-AzureRmLoadBalancerRuleConfig to
create the load balancer configuration. Use New-
AzureRmLoadBalancer to create the load balancer.
Azure CLI Use the az network lb create command to create the initial
load balancer configuration. To define the private IP address,
use az network lb frontend-ip create with the --private-ip-
address parameter. Use az network lb address-pool create to
add the configuration of the back-end address pool. Use az
network lb inbound-nat-rule create to add NAT rules. Use az
network lb rule create to add the load balancer rules. Use az
network lb probe create to add the probes.
Template Use 2 VMs in a Load Balancer and configure NAT rules on the
LB as a guide for deploying a load balancer using a template.
VMs
VMs can be created in the same VNet and they can connect to each other using private IP addresses. They can
connect even if they are in different subnets without the need to configure a gateway or use public IP addresses. To
put VMs into a VNet, you create the VNet and then as you create each VM, you assign it to the VNet and subnet.
VMs acquire their network settings during deployment or startup.
VMs are assigned an IP address when they are deployed. If you deploy multiple VMs into a VNet or subnet, they
are assigned IP addresses as they boot up. A dynamic IP address (DIP ) is the internal IP address associated with a
VM. You can allocate a static DIP to a VM. If you allocate a static DIP, you should consider using a specific subnet
to avoid accidentally reusing a static DIP for another VM.
If you create a VM and later want to migrate it into a VNet, it is not a simple configuration change. You must
redeploy the VM into the VNet. The easiest way to redeploy is to delete the VM, but not any disks attached to it,
and then re-create the VM using the original disks in the VNet.
This table lists the methods that you can use to create a VM in a VNet.
METHOD DESCRIPTION
Azure portal Uses the default network settings that were previously
mentioned to create a VM with a single NIC. To create a VM
with multiple NICs, you must use a different method.
Azure PowerShell Includes the use of Add-AzureRmVMNetworkInterface to add

the NIC that you previously created to the VM configuration.
Azure CLI Create and connect a VM to a Vnet, subnet, and NIC that
build as individual steps.
Template Use Very simple deployment of a Windows VM as a guide for

deploying a VM using a template.
Next steps
For VM -specific steps on how to manage Azure virtual networks for VMs, see the Windows or Linux tutorials.
There are also tutorials on how to load balance VMs and create highly available applications for Windows or Linux.
Learn how to configure user-defined routes and IP forwarding.
Learn how to configure VNet to VNet connections.
Learn how to Troubleshoot routes.
Automatically scale virtual machines in Azure
You can easily automatically scale your virtual machines (VMs) when you use virtual machine scale sets and the
autoscaling feature of Azure Monitor. Your VMs need to be members of a scale set to be automatically scaled. This
article provides information that enables you to better understand how to scale your VMs both vertically and
horizontally using automatic and manual methods.
Horizontal or vertical scaling

The autoscale feature of Azure Monitor only scales horizontally, which is an increase ("out") or decrease ("in") of the
number of VMs. Horizontal scaling is more flexible in a cloud situation as it allows you to run potentially thousands
of VMs to handle load. You scale horizontally by either automatically or manually changing the capacity (or
instance count) of the scale set.
Vertical scaling keeps the same number of VMs, but makes the VMs more ("up") or less ("down") powerful. Power
is measured in attributes such as memory, CPU speed, or disk space. Vertical scaling is dependent on the
availability of larger hardware, which quickly hits an upper limit and can vary by region. Vertical scaling also usually
requires a VM to stop and restart. You scale vertically by setting a new size in the configuration of the VMs in the
scale set.
Using runbooks in Azure Automation, you can easily scale VMs in a scale set up or down.

Scale sets make it easy for you to deploy and manage identical VMs as a set. You can create Linux or Windows
scale sets using the Azure portal, Azure PowerShell, or the Azure CLI. You can also create and manage scale sets
with SDKs such as Python or Node.js, or directly with the REST APIs. Automatic scaling of VMs is accomplished by
applying metrics and rules to the scale set.
Configure autoscale for a scale set

Automatic scaling provides the right number of VMs to handle the load on your application. It enables you to add
VMs to handle increases in load and save money by removing VMs that are sitting idle. You specify a minimum
and maximum number of VMs to run based on a set of rules. Having a minimum makes sure your application is
always running even under no load. Having a maximum value limits your total possible hourly cost.
You can enable autoscale when you create the scale set using Azure PowerShell or Azure CLI. You can also enable
it after the scale set is created. You can create a scale set, install the extension, and configure autoscale using an
Azure Resource Manager template. In the Azure portal, enable autoscale from Azure Monitor, or enable autoscale
from the scale set settings.
Metrics
The autoscale feature of Azure Monitor enables you to scale the number of running VMs up or down based on
metrics. By default, VMs provide basic host-level metrics for disk, network, and CPU usage. When you configure
the collection of diagnostics data using the diagnostic extension, additional guest OS performance counters
become available for disk, CPU, and memory.
If your application needs to scale based on metrics that are not available through the host, then the VMs in the
scale set need to have either the Linux diagnostic extension or Windows diagnostics extension installed. If you
create a scale set using the Azure portal, you need to also use Azure PowerShell or the Azure CLI to install the
extension with the diagnostics configuration that you need.
Rules
Rules combine a metric with an action to be performed. When rule conditions are met, one or more autoscale
actions are triggered. For example, you might have a rule defined that increases the number of VMs by 1 if the
average CPU usage goes above 85 percent.
Notifications
You can set up triggers so that specific web URLs are called or emails are sent based on the autoscale rules that
you create. Webhooks allow you to route the Azure alert notifications to other systems for post-processing or
custom notifications.
Manually scale VMs in a scale set

Horizontal
You can add or remove VMs by changing the capacity of the scale set. In the Azure portal, you can decrease or
increase the number of VMs (shown as instance count) in the scale set by sliding the Override condition bar on
the Scaling screen left or right.
Using Azure PowerShell, you need to get the scale set object using Get-AzureRmVmss. You then set the
sku.capacity property to the number of VMs that you want and update the scale set with Update-AzureRmVmss.
Using Azure CLI, you change the capacity with the --new-capacity parameter for the az vmss scale command.
Vertical
You can manually change the size of the VMs in the Azure portal on the Size screen for the scale set. You can use
Azure PowerShell with Get-AzureRmVmss, setting the image reference sku property, and then using Update-
AzureRmVmss and Update-AzureRmVmssInstance.
Next steps
Learn more about scale sets in Design Considerations for Scale Sets.
Use infrastructure automation tools with virtual
machines in Azure
To create and manage Azure virtual machines (VMs) in a consistent manner at scale, some form of automation is
typically desired. There are many tools and solutions that allow you to automate the complete Azure infrastructure
deployment and management lifecycle. This article introduces some of the infrastructure automation tools that you
can use in Azure. These tools commonly fit in to one of the following approaches:
Automate the configuration of VMs
Tools include Ansible, Chef, and Puppet.
Tools specific to VM customization include cloud-init for Linux VMs, PowerShell Desired State
Configuration (DSC ), and the Azure Custom Script Extension for all Azure VMs.
Automate infrastructure management
Tools include Packer to automate custom VM image builds, and Terraform to automate the infrastructure
build process.
Azure Automation can perform actions across your Azure and on-premises infrastructure.
Automate application deployment and delivery
Examples include Azure DevOps Services and Jenkins.
Ansible
Ansible is an automation engine for configuration management, VM creation, or application deployment. Ansible
uses an agent-less model, typically with SSH keys, to authenticate and manage target machines. Configuration
tasks are defined in playbooks, with a number of Ansible modules available to carry out specific tasks. For more
information, see How Ansible works.
Learn how to:
Install and configure Ansible on Linux for use with Azure.
Create a Linux virtual machine.
Manage a Linux virtual machine.
Chef
Chef is an automation platform that helps define how your infrastructure is configured, deployed, and managed.
Additional components included Chef Habitat for application lifecycle automation rather than the infrastructure,
and Chef InSpec that helps automate compliance with security and policy requirements. Chef Clients are installed
on target machines, with one or more central Chef Servers that store and manage the configurations. For more
information, see An Overview of Chef.
Learn how to:
Deploy Chef Automate from the Azure Marketplace.
Install Chef on Windows and create Azure VMs.
Puppet
Puppet is an enterprise-ready automation platform that handles the application delivery and deployment process.
Agents are installed on target machines to allow Puppet Master to run manifests that define the desired
configuration of the Azure infrastructure and VMs. Puppet can integrate with other solutions such as Jenkins and
GitHub for an improved devops workflow. For more information, see How Puppet works.
Learn how to:
Deploy Puppet from the Azure Marketplace.
Cloud-init
Cloud-init is a widely used approach to customize a Linux VM as it boots for the first time. You can use cloud-init to
install packages and write files, or to configure users and security. Because cloud-init is called during the initial boot
process, there are no additional steps or required agents to apply your configuration. For more information on how
to properly format your #cloud-config files, see the cloud-init documentation site. #cloud-config files are text files
encoded in base64.
Cloud-init also works across distributions. For example, you don't use apt-get install or yum install to install a
package. Instead you can define a list of packages to install. Cloud-init automatically uses the native package
management tool for the distro you select.
We are actively working with our endorsed Linux distro partners in order to have cloud-init enabled images
available in the Azure marketplace. These images make your cloud-init deployments and configurations work
seamlessly with VMs and virtual machine scale sets. The following table outlines the current cloud-init enabled
images availability on the Azure platform:
PUBLISHER OFFER SKU VERSION CLOUD-INIT READY
Canonical UbuntuServer 16.04-LTS latest yes
Canonical UbuntuServer 14.04.5-LTS latest yes
CoreOS CoreOS Stable latest yes
OpenLogic CentOS 7-CI latest preview
RedHat RHEL 7-RAW-CI latest preview
Learn more details about cloud-init on Azure:

Cloud-init support for Linux virtual machines in Azure
Try a tutorial on automated VM configuration using cloud-init.
PowerShell DSC
PowerShell Desired State Configuration (DSC ) is a management platform to define the configuration of target
machines. DSC can also be used on Linux through the Open Management Infrastructure (OMI) server.
DSC configurations define what to install on a machine and how to configure the host. A Local Configuration
Manager (LCM ) engine runs on each target node that processes requested actions based on pushed configurations.
A pull server is a web service that runs on a central host to store the DSC configurations and associated resources.
The pull server communicates with the LCM engine on each target host to provide the required configurations and
report on compliance.
Learn how to:
Create a basic DSC configuration.
Configure a DSC pull server.
Use DSC for Linux.
Azure Custom Script Extension

The Azure Custom Script Extension for Linux or Windows downloads and executes scripts on Azure VMs. You can
use the extension when you create a VM, or any time after the VM is in use.
Scripts can be downloaded from Azure storage or any public location such as a GitHub repository. With the
Custom Script Extension, you can write scripts in any language that runs on the source VM. These scripts can be
used to install applications or configure the VM as desired. To secure credentials, sensitive information such as
passwords can be stored in a protected configuration. These credentials are only decrypted inside the VM.
Learn how to:
Create a Linux VM with the Azure CLI and use the Custom Script Extension.
Create a Windows VM with Azure PowerShell and use the Custom Script Extension.
Packer
Packer automates the build process when you create a custom VM image in Azure. You use Packer to define the
OS and run post-configuration scripts that customize the VM for your specific needs. Once configured, the VM is
then captured as a Managed Disk image. Packer automates the process to create the source VM, network and
storage resources, run configuration scripts, and then create the VM image.
Learn how to:
Use Packer to create a Linux VM image in Azure.
Use Packer to create a Windows VM image in Azure.
Terraform
Terraform is an automation tool that allows you to define and create an entire Azure infrastructure with a single
template format language - the HashiCorp Configuration Language (HCL ). With Terraform, you define templates
that automate the process to create network, storage, and VM resources for a given application solution. You can
use your existing Terraform templates for other platforms with Azure to ensure consistency and simplify the
infrastructure deployment without needing to convert to an Azure Resource Manager template.
Learn how to:
Install and configure Terraform with Azure.
Create an Azure infrastructure with Terraform.
Azure Automation
Azure Automation uses runbooks to process a set of tasks on the VMs you target. Azure Automation is used to
manage existing VMs rather than to create an infrastructure. Azure Automation can run across both Linux and
Windows VMs, as well as on-premises virtual or physical machines with a hybrid runbook worker. Runbooks can
be stored in a source control repository, such as GitHub. These runbooks can then run manually or on a defined
schedule.
Azure Automation also provides a Desired State Configuration (DSC ) service that allows you to create definitions
for how a given set of VMs should be configured. DSC then ensures that the required configuration is applied and
the VM stays consistent. Azure Automation DSC runs on both Windows and Linux machines.
Learn how to:
Create a PowerShell runbook.
Use Hybrid Runbook Worker to manage on-premises resources.
Use Azure Automation DSC.
Azure DevOps Services

Azure DevOps Services is a suite of tools that help you share and track code, use automated builds, and create a
complete continuous integration and development (CI/CD ) pipeline. Azure DevOps Services integrates with Visual
Studio and other editors to simplify usage. Azure DevOps Services can also create and configure Azure VMs and
then deploy code to them.
Learn how to:
Create a continuous integration pipeline with Azure DevOps Services.
Jenkins
Jenkins is a continuous integration server that helps deploy and test applications, and create automated pipelines
for code delivery. There are hundreds of plugins to extend the core Jenkins platform, and you can also integrate
with many other products and solutions through webhooks. You can manually install Jenkins on an Azure VM, run
Jenkins from within a Docker container, or use a pre-built Azure Marketplace image.
Learn how to:
Create a development infrastructure on a Linux VM in Azure with Jenkins, GitHub, and Docker.
Next steps
There are many different options to use infrastructure automation tools in Azure. You have the freedom to use the
solution that best fits your needs and environment. To get started and try some of the tools built-in to Azure, see
how to automate the customization of a Linux or Windows VM.
Secure and use policies on virtual machines in Azure
It’s important to keep your virtual machine (VM ) secure for the applications that you run. Securing your VMs can
include one or more Azure services and features that cover secure access to your VMs and secure storage of your
data. This article provides information that enables you to keep your VM and applications secure.
Antimalware
The modern threat landscape for cloud environments is dynamic, increasing the pressure to maintain effective
protection in order to meet compliance and security requirements. Microsoft Antimalware for Azure is a free real-
time protection capability that helps identify and remove viruses, spyware, and other malicious software. Alerts can
be configured to notify you when known malicious or unwanted software attempts to install itself or run on your
VM.
Azure Security Center

Azure Security Center helps you prevent, detect, and respond to threats to your VMs. Security Center provides
integrated security monitoring and policy management across your Azure subscriptions, helps detect threats that
might otherwise go unnoticed, and works with a broad ecosystem of security solutions.
Encryption
For enhanced Windows VM and Linux VM security and compliance, virtual disks in Azure can be encrypted.
Virtual disks on Windows VMs are encrypted at rest using Bitlocker. Virtual disks on Linux VMs are encrypted at
rest using dm-crypt.
There is no charge for encrypting virtual disks in Azure. Cryptographic keys are stored in Azure Key Vault using
software-protection, or you can import or generate your keys in Hardware Security Modules (HSMs) certified to
FIPS 140-2 level 2 standards. These cryptographic keys are used to encrypt and decrypt virtual disks attached to
your VM. You retain control of these cryptographic keys and can audit their use. An Azure Active Directory service
principal provides a secure mechanism for issuing these cryptographic keys as VMs are powered on and off.
Key Vault and SSH Keys

Secrets and certificates can be modeled as resources and provided by Key Vault. You can use Azure PowerShell to
create key vaults for Windows VMs and the Azure CLI for Linux VMs. You can also create keys for encryption.
Key vault access policies grant permissions to keys, secrets, and certificates separately. For example, you can give a
user access to only keys, but no permissions for secrets. However, permissions to access keys or secrets or
certificates are at the vault level. In other words, key vault access policy does not support object level permissions.
When you connect to VMs, you should use public-key cryptography to provide a more secure way to sign in to
them. This process involves a public and private key exchange using the secure shell (SSH) command to
authenticate yourself rather than a username and password. Passwords are vulnerable to brute-force attacks,
especially on Internet-facing VMs such as web servers. With a secure shell (SSH) key pair, you can create a Linux
VM that uses SSH keys for authentication, eliminating the need for passwords to log in. You can also use SSH keys
to connect from a Windows VM to a Linux VM.
Managed identities for Azure resources

A common challenge when building cloud applications is how to manage the credentials in your code for
authenticating to cloud services. Keeping the credentials secure is an important task. Ideally, the credentials never
appear on developer workstations and aren't checked into source control. Azure Key Vault provides a way to
securely store credentials, secrets, and other keys, but your code has to authenticate to Key Vault to retrieve them.
The managed identities for Azure resources feature in Azure Active Directory (Azure AD ) solves this problem. The
feature provides Azure services with an automatically managed identity in Azure AD. You can use the identity to
authenticate to any service that supports Azure AD authentication, including Key Vault, without any credentials in
your code. Your code that's running on a VM can request a token from two endpoints that are accessible only from
within the VM. For more detailed information about this service, review the managed identities for Azure resources
overview page.
Policies
Azure policies can be used to define the desired behavior for your organization's Windows VMs and Linux VMs. By
using policies, an organization can enforce various conventions and rules throughout the enterprise. Enforcement
of the desired behavior can help mitigate risk while contributing to the success of the organization.
Role-based access control

Using role-based access control (RBAC ), you can segregate duties within your team and grant only the amount of
access to users on your VM that they need to perform their jobs. Instead of giving everybody unrestricted
permissions on the VM, you can allow only certain actions. You can configure access control for the VM in the
Azure portal, using the Azure CLI, orAzure PowerShell.
Next steps
Walk through the steps to monitor virtual machine security by using Azure Security Center for Linux or
Windows.
Virtual machines lifecycle and states
Azure Virtual Machines (VMs) go through different states that can be categorized into provisioning and power
states. The purpose of this article is to describe these states and specifically highlight when customers are billed for
instance usage.
Power states
The power state represents the last known state of the VM.
The following table provides a description of each instance state and indicates whether it is billed for instance
usage or not.
STATE DESCRIPTION INSTANCE USAGE BILLING
Starting VM is starting up. Not billed

"statuses": [
{
"code": "PowerState/starting",
"level": "Info",
"displayStatus": "VM starting"
}
]
Running Normal working state for a VM Billed

"statuses": [
{
"code": "PowerState/running",
"level": "Info",
"displayStatus": "VM running"
}
]
Stopping This is a transitional state. When Billed
completed, it will show as Stopped.
"statuses": [
{
"code": "PowerState/stopping",
"level": "Info",
"displayStatus": "VM stopping"
}
]
Stopped The VM has been shut down from Not billed*

within the guest OS or using the
PowerOff APIs.
Hardware is still allocated to the VM
and it remains on the host.
"statuses": [
{
"code": "PowerState/stopped",
"level": "Info",
"displayStatus": "VM stopped"
}
]
Deallocating Transitional state. When completed, Not billed*

the VM will show as Deallocated.
"statuses": [
{
"code":
"PowerState/deallocating",
"level": "Info",
"displayStatus": "VM
deallocating"
}
]
Deallocated The VM has been stopped Not billed

successfully and removed from the
host.
"statuses": [
{
"code": "PowerState/deallocated",
"level": "Info",
"displayStatus": "VM deallocated"
}
]
*Some Azure resources, such as Disks and Networking, incur charges regardless of the instance's state.
Provisioning states
A provisioning state is the status of a user-initiated, control-plane operation on the VM. These states are separate
from the power state of a VM.
Create – VM creation.
Update – updates the model for an existing VM. Some non-model changes to VM such as Start/Restart
also fall under update.
Delete – VM deletion.
Deallocate – is where a VM is stopped and removed from the host. Deallocating a VM is considered an
update, so it will display provisioning states related to updating.
Here are the transitional operation states after the platform has accepted a user-initiated action:
States Description
Creating "statuses": [
{
"code": "ProvisioningState/creating",
"level": "Info",
"displayStatus": "Creating"
}
Updating "statuses": [
{
"code": "ProvisioningState/updating",
"level": "Info",
"displayStatus": "Updating"
}
]
Deleting "statuses": [
{
"code": "ProvisioningState/deleting",
"level": "Info",
"displayStatus": "Deleting"
}
]
OS provisioning states If a VM is created with an OS image and not with a specialized image, then following
substates can be observed:
1. OSProvisioningInprogress – The VM is running, and installation of guest OS is
in progress.
"statuses": [
{
"code": "ProvisioningState/creating/OSProvisioningInprogress",
"level": "Info",
"displayStatus": "OS Provisioning In progress"
}
]
2. OSProvisioningComplete – Short-lived state. The VM quickly transitions to

Success unless any extensions need to be installed. Installing extensions can take
time.
"statuses": [
{
"code": "ProvisioningState/creating/OSProvisioningComplete",
"level": "Info",
"displayStatus": "OS Provisioning Complete"
}
]
Note: OS Provisioning can transition to Failed if there is an OS failure or the OS

doesn't install in time. Customers will be billed for the deployed VM on the
infrastructure.
Once the operation is complete, the VM will transition into one of the following states:
Succeeded – the user-initiated actions have completed.
"statuses": [
{
"code": "ProvisioningState/succeeded",
"level": "Info",
"displayStatus": "Provisioning succeeded",
"time": "time"
}
]
Failed – represents a failed operation. Refer to the error codes to get more information and possible
solutions.
"statuses": [
{
"code": "ProvisioningState/failed/InternalOperationError",
"level": "Error",
"displayStatus": "Provisioning failed",
"message": "Operation abandoned due to internal error. Please try again later.",
"time": "time"
}
]
VM instance view
The instance view API provides VM running-state information. For more information, see the Virtual Machines -
Instance View API documentation.
Azure Resources explorer provides a simple UI for viewing the VM running state: Resource Explorer.
Provisioning states are visible on VM properties and instance view. Power states are available in instance view of
VM.
Next steps
To learn more about monitoring your VM, see How to monitor virtual machines in Azure.
How to monitor virtual machines in Azure
You can take advantage of many opportunities to monitor your VMs by collecting, viewing, and analyzing
diagnostic and log data. To do simple monitoring of your VM, you can use the Overview screen for the VM in the
Azure portal. You can use extensions to configure diagnostics on your VMs to collect additional metric data. You
can also use more advanced monitoring options, such as Application Insights and Log Analytics.
Diagnostics and metrics

You can set up and monitor the collection of diagnostics data using metrics in the Azure portal, the Azure CLI,
Azure PowerShell, and programming Applications Programming Interfaces (APIs). For example, you can:
Observe basic metrics for the VM. On the Overview screen of the Azure portal, the basic metrics shown
include CPU usage, network usage, total of disk bytes, and disk operations per second.
Enable the collection of boot diagnostics and view it using the Azure portal. When bringing your
own image to Azure or even booting one of the platform images, there can be many reasons why a VM gets
into a non-bootable state. You can easily enable boot diagnostics when you create a VM by clicking
Enabled for Boot Diagnostics under the Monitoring section of the Settings screen.
As VMs boot, the boot diagnostic agent captures boot output and stores it in Azure storage. This data can
be used to troubleshoot VM boot issues. Boot diagnostics are not automatically enabled when you create a
VM from command-line tools. Before enabling boot diagnostics, a storage account needs to be created for
storing boot logs. If you enable boot diagnostics in the Azure portal, a storage account is automatically
created for you.
If you didn’t enable boot diagnostics when the VM was created, you can always enable it later by using
Azure CLI, Azure PowerShell, or an Azure Resource Manager template.
Enable the collection of guest OS diagnostics data. When you create a VM, you have the opportunity
on the settings screen to enable guest OS diagnostics. When you do enable the collection of diagnostics
data, the IaaSDiagnostics extension for Linux or the IaaSDiagnostics extension for Windows is added to the
VM, which enables you to collect additional disk, CPU, and memory data.
Using the collected diagnostics data, you can configure autoscaling for your VMs. You can also configure
logs to store the data and set up alerts to let you know when performance isn't quite right.
Alerts
You can create alerts based on specific performance metrics. Examples of the issues you can be alerted about
include when average CPU usage exceeds a certain threshold, or available free disk space drops below a certain
amount. Alerts can be configured in the Azure portal, using Azure PowerShell, or the Azure CLI.
Azure Service Health

Azure Service Health provides personalized guidance and support when issues in Azure services affect you, and
helps you prepare for upcoming planned maintenance. Azure Service Health alerts you and your teams using
targeted and flexible notifications.
Azure Resource Health

Azure Resource health helps you diagnose and get support when an Azure issue impacts your resources. It
informs you about the current and past health of your resources and helps you mitigate issues. Resource health
provides technical support when you need help with Azure service issues.
Logs
The Azure Activity Log is a subscription log that provides insight into subscription-level events that have occurred
in Azure. The log includes a range of data, from Azure Resource Manager operational data to updates on Service
Health events. You can click Activity Log in the Azure portal to view the log for your VM.
Some of the things you can do with the activity log include:
Create an alert on an Activity Log event.
Stream it to an Event Hub for ingestion by a third-party service or custom analytics solution such as PowerBI.
Analyze it in PowerBI using the PowerBI content pack.
Save it to a storage account for archival or manual inspection. You can specify the retention time (in days) using
the Log Profile.
You can also access activity log data by using Azure PowerShell, the Azure CLI, or Monitor REST APIs.
Azure Diagnostic Logs are logs emitted by your VM that provide rich, frequent data about its operation. Diagnostic
logs differ from the activity log by providing insight about operations that were performed within the VM.
Some of the things you can do with diagnostics logs include:
Save them to a storage account for auditing or manual inspection. You can specify the retention time (in days)
using Resource Diagnostic Settings.
Stream them to Event Hubs for ingestion by a third-party service or custom analytics solution such as PowerBI.
Analyze them with OMS Log Analytics.
Advanced monitoring
Operations Management Suite (OMS ) provides monitoring, alerting, and alert remediation capabilities
across cloud and on-premises assets. You can install an extension on a Linux VM or a Windows VM that
installs the OMS agent, and enrolls the VM into an existing OMS workspace.
Log Analytics is a service in OMS that monitors your cloud and on-premises environments to maintain
their availability and performance. It collects data generated by resources in your cloud and on-premises
environments and from other monitoring tools to provide analysis across multiple sources.
For Windows and Linux VMs, the recommended method for collecting logs and metrics is by installing the
Log Analytics agent. The easiest way to install the Log Analytics agent on a VM is through the Log Analytics
VM Extension. Using the extension simplifies the installation process and automatically configures the agent
to send data to the Log Analytics workspace that you specify. The agent is also upgraded automatically,
ensuring that you have the latest features and fixes.
Network Watcher enables you to monitor your VM and its associated resources as they relate to the
network that they are in. You can install the Network Watcher Agent extension on a Linux VM or a Windows
VM.
Next steps
Walk through the steps in Monitor a Windows Virtual Machine with Azure PowerShell or Monitor a Linux
Virtual Machine with the Azure CLI.
Learn more about the best practices around Monitoring and diagnostics.
Backup and restore options for virtual machines in
Azure
You can protect your data by taking backups at regular intervals. There are several backup options available for
VMs, depending on your use-case.
Azure Backup
For backing up Azure VMs running production workloads, use Azure Backup. Azure Backup supports application-
consistent backups for both Windows and Linux VMs. Azure Backup creates recovery points that are stored in geo-
redundant recovery vaults. When you restore from a recovery point, you can restore the whole VM or just specific
files.
For a simple, hands-on introduction to Azure Backup for Azure VMs, see the "Back up Azure virtual machines"
tutorial for Linux or Windows.
For more information on how Azure Backup works, see Plan your VM backup infrastructure in Azure
Azure Site Recovery

Azure Site Recovery protects your VMs from a major disaster scenario, when a whole region experiences an
outage due to major natural disaster or widespread service interruption. You can configure Azure Site Recovery for
your VMs so that you can recover your application with a single click in matter of minutes. You can replicate to an
Azure region of your choice, it is not restricted to paired regions.
You can run disaster-recovery drills with on-demand test failovers, without affecting your production workloads or
ongoing replication. Create recovery plans to orchestrate failover and failback of the entire application running on
multiple VMs. The recovery plan feature is integrated with Azure automation runbooks.
You can get started by replicating your virtual machines.
Managed snapshots
In development and test environments, snapshots provide a quick and simple option for backing up VMs that use
Managed Disks. A managed snapshot is a read-only full copy of a managed disk. Snapshots exist independent of
the source disk and can be used to create new managed disks for rebuilding a VM. They are billed based on the
used portion of the disk. For example, if you create a snapshot of a managed disk with provisioned capacity of 64
GB and actual used data size of 10 GB, snapshot will be billed only for the used data size of 10 GB.
For more information on creating snapshots, see:
Create copy of VHD stored as a Managed Disk using Snapshots in Windows
Create copy of VHD stored as a Managed Disk using Snapshots in Linux
Next steps
You can try out Azure Backup by following the "Back up Windows virtual machines tutorial" for Linux or Windows.
HPC, Batch, and Big Compute solutions using Azure
VMs
Organizations have large-scale computing needs. These Big Compute workloads include engineering design and
analysis, financial risk calculations, image rendering, complex modeling, Monte Carlo simulations, and more.
Use the Azure cloud to efficiently run compute-intensive Linux and Windows workloads, from parallel batch jobs to
traditional HPC simulations. Run your HPC and batch workloads on Azure infrastructure, with your choice of
compute services, grid managers, Marketplace solutions, and vendor-hosted (SaaS ) applications. Azure provides
flexible solutions to distribute work and scale to thousands of VMs or cores and then scale down when you need
fewer resources.
Solution options
Do-it-yourself solutions
Set up your own cluster environment in Azure virtual machines or virtual machine scale sets.
Lift and shift an on-premises cluster, or deploy a new cluster in Azure for additional capacity.
Use Azure Resource Manager templates to deploy leading workload managers, infrastructure, and
applications.
Choose HPC and GPU VM sizes that include specialized hardware and network connections for MPI or
GPU workloads.
Add high performance storage for I/O -intensive workloads.
Hybrid solutions
Extend your on-premises solution to offload ("burst") peak workloads to Azure infrastructure
Use cloud compute on-demand with your existing workload manager.
Take advantage of HPC and GPU VM sizes for MPI or GPU workloads.
Big Compute solutions as a service
Develop custom Big Compute solutions and workflows using Azure Batch and related Azure services.
Run Azure-enabled engineering and simulation solutions from vendors including Altair, Rescale, and
Cycle Computing (now joined with Microsoft).
Use a Cray supercomputer as a service hosted in Azure.
Marketplace solutions
Use the scale of HPC applications and solutions offered in the Azure Marketplace.
The following sections provide more information about the supporting technologies and links to guidance.
Marketplace solutions
Visit the Azure Marketplace for Linux and Windows VM images and solutions designed for HPC. Examples include:
RogueWave CentOS -based HPC
SUSE Linux Enterprise Server for HPC
TIBCO Grid Server Engine
Azure Data Science VM for Windows and Linux
D3View
UberCloud
Intel Cloud Edition for Lustre
HPC applications
Run custom or commercial HPC applications in Azure. Several examples in this section are benchmarked to scale
efficiently with additional VMs or compute cores. Visit the Azure Marketplace for ready-to-deploy solutions.
NOTE
Check with the vendor of any commercial application for licensing or other restrictions for running in the cloud. Not all
vendors offer pay-as-you-go licensing. You might need a licensing server in the cloud for your solution, or connect to an on-
premises license server.
Engineering applications
Altair RADIOSS
ANSYS CFD
MATLAB Distributed Computing Server
StarCCM+
OpenFOAM
Graphics and rendering
Autodesk Maya, 3ds Max, and Arnold on Azure Batch
AI and deep learning
Batch AI training for deep learning models
Microsoft Cognitive Toolkit
Deep Learning VM
Batch Shipyard recipes for deep learning
HPC and GPU VM sizes

Azure offers a range of sizes for Linux and Windows VMs, including sizes designed for compute-intensive
workloads. For example, H16r and H16mr VMs can connect to a high throughput back-end RDMA network. This
cloud network can improve the performance of tightly coupled parallel applications running under Microsoft MPI
or Intel MPI.
N -series VMs feature NVIDIA GPUs designed for compute-intensive or graphics-intensive applications including
artificial intelligence (AI) learning and visualization.
Learn more:
High performance compute sizes for Linux and Windows VMs
GPU -enabled sizes for Linux and Windows VMs
Learn how to:
Set up a Linux RDMA cluster to run MPI applications
Set up a Windows RDMA cluster with Microsoft HPC Pack to run MPI applications
Use compute-intensive VMs in Batch pools
Azure Batch
Batch is a platform service for running large-scale parallel and high-performance computing (HPC ) applications
efficiently in the cloud. Azure Batch schedules compute-intensive work to run on a managed pool of virtual
machines, and can automatically scale compute resources to meet the needs of your jobs.
SaaS providers or developers can use the Batch SDKs and tools to integrate HPC applications or container
workloads with Azure, stage data to Azure, and build job execution pipelines.
Learn how to:
Get started developing with Batch
Use Azure Batch code samples
Use low -priority VMs with Batch
Run containerized HPC workloads with Batch Shipyard
Run parallel R workloads on Batch
Run on-demand Spark jobs on Batch
Workload managers
The following are examples of cluster and workload managers that can run in Azure infrastructure. Create stand-
alone clusters in Azure VMs or burst to Azure VMs from an on-premises cluster.
Alces Flight Compute
TIBCO DataSynapse GridServer
Bright Cluster Manager
IBM Spectrum Symphony and Symphony LSF
PBS Pro
Microsoft HPC Pack - see options to run in Windows and Linux VMs
HPC storage
Large-scale Batch and HPC workloads have demands for data storage and access that exceed the capabilities of
traditional cloud file systems. Implement parallel file system solutions in Azure such as Lustre and BeeGFS.
Learn more:
Parallel virtual file systems on Azure
High performance cloud storage solutions from Avere (now joined with Microsoft)
Related Azure services

Azure virtual machines, virtual machine scale sets, Batch, and related compute services are the foundation of most
Azure HPC solutions. However, your solution can take advantage of many related Azure services. Here is a partial
list:
Storage
Blob, table, and queue storage
File storage
Data and analytics
HDInsight
Data Factory
Data Lake Storage Gen1
Databricks
SQL Database
AI and machine learning
Machine Learning Services
Batch AI
Genomics
Networking
Virtual Network
ExpressRoute
Containers
Container Service
Azure Kubernetes Service (AKS )
Container Registry
Customer stories
Examples of customers that have solved business problems with Azure HPC solutions:
ANEO
AXA Global P&C
Axioma
d3View
EFS
Hymans Robertson
MetLife
Microsoft Research
Milliman
Mitsubishi UFJ Securities International
NeuroInitiative
Schlumberger
Towers Watson
Next steps
Learn more about Big Compute solutions for engineering simulation, rendering, banking and capital markets,
and genomics.
For the latest announcements, see the Microsoft HPC and Batch team blog and the Azure blog.
Use the managed and scalable Azure Batch service to run compute-intensive workloads, without managing
underlying infrastructure Learn more
Example Azure infrastructure walkthrough for
Windows VMs
This article walks through building out an example application infrastructure. We detail designing an infrastructure
for a simple online store that brings together all the guidelines and decisions around naming conventions,
availability sets, virtual networks and load balancers, and actually deploying your virtual machines (VMs).
Example workload
Adventure Works Cycles wants to build an online store application in Azure that consists of:
Two IIS servers running the client front-end in a web tier
Two IIS servers processing data and orders in an application tier
Two Microsoft SQL Server instances with AlwaysOn availability groups (two SQL Servers and a majority node
witness) for storing product data and orders in a database tier
Two Active Directory domain controllers for customer accounts and suppliers in an authentication tier
All the servers are located in two subnets:
a front-end subnet for the web servers
a back-end subnet for the application servers, SQL cluster, and domain controllers
Incoming secure web traffic must be load-balanced among the web servers as customers browse the online store.
Order processing traffic in the form of HTTP requests from the web servers must be balanced among the
application servers. Additionally, the infrastructure must be designed for high availability.
The resulting design must incorporate:
An Azure subscription and account
A single resource group
Azure Managed Disks
A virtual network with two subnets
Availability sets for the VMs with a similar role
Virtual machines
All the above follow these naming conventions:
Adventure Works Cycles uses [IT workload]-[location]-[Azure resource] as a prefix
For this example, "azos" (Azure Online Store) is the IT workload name and "use" (East US 2) is the
location
Virtual networks use AZOS -USE -VN [number]
Availability sets use azos-use-as-[role]
Virtual machine names use azos-use-vm-[vmname]
Azure subscriptions and accounts

Adventure Works Cycles is using their Enterprise subscription, named Adventure Works Enterprise Subscription, to
provide billing for this IT workload.
Storage
Adventure Works Cycles determined that they should use Azure Managed Disks. When creating VMs, both
available storage tiers are used:
Standard storage for the web servers, application servers, and domain controllers and their data disks.
Premium storage for the SQL Server VMs and their data disks.

Because the virtual network does not need ongoing connectivity to the Adventure Work Cycles on-premises
network, they decided on a cloud-only virtual network.
They created a cloud-only virtual network with the following settings using the Azure portal:
Name: AZOS -USE -VN01
Location: East US 2
Virtual network address space: 10.0.0.0/8
First subnet:
Name: FrontEnd
Address space: 10.0.1.0/24
Second subnet:
Name: BackEnd
Address space: 10.0.2.0/24
Availability sets
To maintain high availability of all four tiers of their online store, Adventure Works Cycles decided on four
availability sets:
azos-use-as-web for the web servers
azos-use-as-app for the application servers
azos-use-as-sql for the SQL Servers
azos-use-as-dc for the domain controllers
Virtual machines
Adventure Works Cycles decided on the following names for their Azure VMs:
azos-use-vm -web01 for the first web server
azos-use-vm -web02 for the second web server
azos-use-vm -app01 for the first application server
azos-use-vm -app02 for the second application server
azos-use-vm -sql01 for the first SQL Server server in the cluster
azos-use-vm -sql02 for the second SQL Server server in the cluster
azos-use-vm -dc01 for the first domain controller
azos-use-vm -dc02 for the second domain controller
Here is the resulting configuration.
This configuration incorporates:

A cloud-only virtual network with two subnets (FrontEnd and BackEnd)
Azure Managed Disks with both Standard and Premium disks
Four availability sets, one for each tier of the online store
The virtual machines for the four tiers
An external load balanced set for HTTPS -based web traffic from the Internet to the web servers
An internal load balanced set for unencrypted web traffic from the web servers to the application servers
A single resource group
Virtual machine vCPU quotas
The vCPU quotas for virtual machines and virtual machine scale sets are arranged in two tiers for each
subscription, in each region. The first tier is the Total Regional vCPUs, and the second tier is the various VM size
family cores such as the D -series vCPUs. Any time a new VM is deployed the vCPUs for the VM must not exceed
the vCPU quota for the VM size family or the total regional vCPU quota. If either of those quotas are exceeded, the
VM deployment will not be allowed. There is also a quota for the overall number of virtual machines in the region.
The details on each of these quotas can be seen in the Usage + quotas section of the Subscription page in the
Azure portal, or you can query for the values using PowerShell.
Check usage
You can use the Get-AzureRmVMUsage cmdlet to check on your quota usage.
Get-AzureRmVMUsage -Location "East US"
The output will look similar to this:
Name Current Value Limit Unit

---- ------------- ----- ----
Availability Sets 0 2000 Count
Total Regional vCPUs 4 260 Count
Virtual Machines 4 10000 Count
Virtual Machine Scale Sets 1 2000 Count
Standard B Family vCPUs 1 10 Count
Standard DSv2 Family vCPUs 1 100 Count
Standard Dv2 Family vCPUs 2 100 Count
Basic A Family vCPUs 0 100 Count
Standard A0-A7 Family vCPUs 0 250 Count
Standard A8-A11 Family vCPUs 0 100 Count
Standard D Family vCPUs 0 100 Count
Standard G Family vCPUs 0 100 Count
Standard DS Family vCPUs 0 100 Count
Standard GS Family vCPUs 0 100 Count
Standard F Family vCPUs 0 100 Count
Standard FS Family vCPUs 0 100 Count
Standard NV Family vCPUs 0 24 Count
Standard NC Family vCPUs 0 48 Count
Standard H Family vCPUs 0 8 Count
Standard Av2 Family vCPUs 0 100 Count
Standard LS Family vCPUs 0 100 Count
Standard Dv2 Promo Family vCPUs 0 100 Count
Standard DSv2 Promo Family vCPUs 0 100 Count
Standard MS Family vCPUs 0 0 Count
Standard Dv3 Family vCPUs 0 100 Count
Standard DSv3 Family vCPUs 0 100 Count
Standard Ev3 Family vCPUs 0 100 Count
Standard ESv3 Family vCPUs 0 100 Count
Standard FSv2 Family vCPUs 0 100 Count
Standard ND Family vCPUs 0 0 Count
Standard NCv2 Family vCPUs 0 0 Count
Standard NCv3 Family vCPUs 0 0 Count
Standard LSv2 Family vCPUs 0 0 Count
Standard Storage Managed Disks 2 10000 Count
Premium Storage Managed Disks 1 10000 Count
Reserved VM Instances
Reserved VM Instances, which are scoped to a single subscription without VM size flexibility, will add a new aspect
to the vCPU quotas. These values describe the number of instances of the stated size that must be deployable in
the subscription. They work as a placeholder in the quota system to ensure that quota is reserved to ensure
reserved VM instances are deployable in the subscription. For example, if a specific subscription has 10
Standard_D1 reserved VM instances the usages limit for Standard_D1 reserved VM instances will be 10. This will
cause Azure to ensure that there are always at least 10 vCPUs available in the Total Regional vCPUs quota to be
used for Standard_D1 instances and there are at least 10 vCPUs available in the Standard D Family vCPU quota to
be used for Standard_D1 instances.
If a quota increase is required to purchase a Single Subscription RI, you can request a quota increase on your
subscription.
Next steps
For more information about billing and quotas, see Azure subscription and service limits, quotas, and constraints.
Create and manage Windows VMs in Azure using C#
An Azure Virtual Machine (VM ) needs several supporting Azure resources. This article covers creating, managing,
and deleting VM resources using C#. You learn how to:
Create a Visual Studio project
Install the package
Create credentials
Create resources
Perform management tasks
Delete resources
Run the application
It takes about 20 minutes to do these steps.

1. If you haven't already, install Visual Studio. Select .NET desktop development on the Workloads page, and
then click Install. In the summary, you can see that .NET Framework 4 - 4.6 development tools is
automatically selected for you. If you have already installed Visual Studio, you can add the .NET workload using
the Visual Studio Launcher.
2. In Visual Studio, click File > New > Project.
3. In Templates > Visual C#, select Console App (.NET Framework), enter myDotnetProject for the name of
the project, select the location of the project, and then click OK.
Install the package

NuGet packages are the easiest way to install the libraries that you need to finish these steps. To get the libraries
that you need in Visual Studio, do these steps:
1. Click Tools > Nuget Package Manager, and then click Package Manager Console.
2. Type this command in the console:
Install-Package Microsoft.Azure.Management.Fluent
Create credentials
Before you start this step, make sure that you have access to an Active Directory service principal. You should also
record the application ID, the authentication key, and the tenant ID that you need in a later step.
Create the authorization file
1. In Solution Explorer, right-click myDotnetProject > Add > New Item, and then select Text File in Visual C#
Items. Name the file azureauth.properties, and then click Add.
2. Add these authorization properties:
subscription=<subscription-id>
client=<application-id>
key=<authentication-key>
tenant=<tenant-id>
managementURI=https://management.core.windows.net/
baseURL=https://management.azure.com/
authURL=https://login.windows.net/
graphURL=https://graph.windows.net/
Replace <subscription-id> with your subscription identifier, <application-id> with the Active Directory
application identifier, <authentication-key> with the application key, and <tenant-id> with the tenant
identifier.
3. Save the azureauth.properties file.
4. Set an environment variable in Windows named AZURE_AUTH_LOCATION with the full path to
authorization file that you created. For example, the following PowerShell command can be used:
[Environment]::SetEnvironmentVariable("AZURE_AUTH_LOCATION", "C:\Visual Studio

2017\Projects\myDotnetProject\myDotnetProject\azureauth.properties", "User")
Create the management client

1. Open the Program.cs file for the project that you created, and then add these using statements to the
existing statements at top of the file:
using Microsoft.Azure.Management.Compute.Fluent;
using Microsoft.Azure.Management.Compute.Fluent.Models;
using Microsoft.Azure.Management.Fluent;
using Microsoft.Azure.Management.ResourceManager.Fluent;
using Microsoft.Azure.Management.ResourceManager.Fluent.Core;
2. To create the management client, add this code to the Main method:
var credentials = SdkContext.AzureCredentialsFactory

.FromFile(Environment.GetEnvironmentVariable("AZURE_AUTH_LOCATION"));
var azure = Azure

.Configure()
.WithLogLevel(HttpLoggingDelegatingHandler.Level.Basic)
.Authenticate(credentials)
.WithDefaultSubscription();
Create resources
Create the resource group
All resources must be contained in a Resource group.
To specify values for the application and create the resource group, add this code to the Main method:
var groupName = "myResourceGroup";
var vmName = "myVM";
var location = Region.USWest;
Console.WriteLine("Creating resource group...");

var resourceGroup = azure.ResourceGroups.Define(groupName)
.WithRegion(location)
.Create();
Create the availability set

Availability sets make it easier for you to maintain the virtual machines used by your application.
To create the availability set, add this code to the Main method:
Console.WriteLine("Creating availability set...");

var availabilitySet = azure.AvailabilitySets.Define("myAVSet")
.WithExistingResourceGroup(groupName)
.WithSku(AvailabilitySetSkuTypes.Managed)
.Create();
Create the public IP address

A Public IP address is needed to communicate with the virtual machine.
To create the public IP address for the virtual machine, add this code to the Main method:
Console.WriteLine("Creating public IP address...");

var publicIPAddress = azure.PublicIPAddresses.Define("myPublicIP")
.WithDynamicIP()
.Create();
Create the virtual network

A virtual machine must be in a subnet of a Virtual network.
To create a subnet and a virtual network, add this code to the Main method:
Console.WriteLine("Creating virtual network...");

var network = azure.Networks.Define("myVNet")
.WithAddressSpace("10.0.0.0/16")
.WithSubnet("mySubnet", "10.0.0.0/24")
.Create();
Create the network interface

A virtual machine needs a network interface to communicate on the virtual network.
To create a network interface, add this code to the Main method:
Console.WriteLine("Creating network interface...");
var networkInterface = azure.NetworkInterfaces.Define("myNIC")
.WithExistingPrimaryNetwork(network)
.WithSubnet("mySubnet")
.WithPrimaryPrivateIPAddressDynamic()
.WithExistingPrimaryPublicIPAddress(publicIPAddress)
.Create();
Create the virtual machine

Now that you created all the supporting resources, you can create a virtual machine.
To create the virtual machine, add this code to the Main method:
Console.WriteLine("Creating virtual machine...");

azure.VirtualMachines.Define(vmName)
.WithExistingPrimaryNetworkInterface(networkInterface)
.WithLatestWindowsImage("MicrosoftWindowsServer", "WindowsServer", "2012-R2-Datacenter")
.WithAdminUsername("azureuser")
.WithAdminPassword("Azure12345678")
.WithComputerName(vmName)
.WithExistingAvailabilitySet(availabilitySet)
.WithSize(VirtualMachineSizeTypes.StandardDS1)
.Create();
NOTE
This tutorial creates a virtual machine running a version of the Windows Server operating system. To learn more about
selecting other images, see Navigate and select Azure virtual machine images with Windows PowerShell and the Azure CLI.
If you want to use an existing disk instead of a marketplace image, use this code:
var managedDisk = azure.Disks.Define("myosdisk")

.WithWindowsFromVhd("https://mystorage.blob.core.windows.net/vhds/myosdisk.vhd")
.WithSizeInGB(128)
.WithSku(DiskSkuTypes.PremiumLRS)
.Create();
azure.VirtualMachines.Define("myVM")
.WithExistingPrimaryNetworkInterface(networkInterface)
.WithSpecializedOSDisk(managedDisk, OperatingSystemTypes.Windows)
.WithExistingAvailabilitySet(availabilitySet)
.Create();

deleting a virtual machine. Additionally, you may want to create code to automate repetitive or complex tasks.
When you need to do anything with the VM, you need to get an instance of it:
var vm = azure.VirtualMachines.GetByResourceGroup(groupName, vmName);
Get information about the VM

To get information about the virtual machine, add this code to the Main method:
Console.WriteLine("Getting information about the virtual machine...");
Console.WriteLine("hardwareProfile");
Console.WriteLine(" vmSize: " + vm.Size);
Console.WriteLine("storageProfile");
Console.WriteLine(" imageReference");
Console.WriteLine(" publisher: " + vm.StorageProfile.ImageReference.Publisher);
Console.WriteLine(" offer: " + vm.StorageProfile.ImageReference.Offer);
Console.WriteLine(" sku: " + vm.StorageProfile.ImageReference.Sku);
Console.WriteLine(" version: " + vm.StorageProfile.ImageReference.Version);
Console.WriteLine(" osDisk");
Console.WriteLine(" osType: " + vm.StorageProfile.OsDisk.OsType);
Console.WriteLine(" name: " + vm.StorageProfile.OsDisk.Name);
Console.WriteLine(" createOption: " + vm.StorageProfile.OsDisk.CreateOption);
Console.WriteLine(" caching: " + vm.StorageProfile.OsDisk.Caching);
Console.WriteLine("osProfile");
Console.WriteLine(" computerName: " + vm.OSProfile.ComputerName);
Console.WriteLine(" adminUsername: " + vm.OSProfile.AdminUsername);
Console.WriteLine(" provisionVMAgent: " + vm.OSProfile.WindowsConfiguration.ProvisionVMAgent.Value);
Console.WriteLine(" enableAutomaticUpdates: " +
vm.OSProfile.WindowsConfiguration.EnableAutomaticUpdates.Value);
Console.WriteLine("networkProfile");
foreach (string nicId in vm.NetworkInterfaceIds)
{
Console.WriteLine(" networkInterface id: " + nicId);
}
Console.WriteLine("vmAgent");
Console.WriteLine(" vmAgentVersion" + vm.InstanceView.VmAgent.VmAgentVersion);
Console.WriteLine(" statuses");
foreach (InstanceViewStatus stat in vm.InstanceView.VmAgent.Statuses)
{
Console.WriteLine(" code: " + stat.Code);
Console.WriteLine(" level: " + stat.Level);
Console.WriteLine(" displayStatus: " + stat.DisplayStatus);
Console.WriteLine(" message: " + stat.Message);
Console.WriteLine(" time: " + stat.Time);
}
Console.WriteLine("disks");
foreach (DiskInstanceView disk in vm.InstanceView.Disks)
{
Console.WriteLine(" name: " + disk.Name);
Console.WriteLine(" statuses");
foreach (InstanceViewStatus stat in disk.Statuses)
{
Console.WriteLine(" time: " + stat.Time);
}
}
Console.WriteLine("VM general status");
Console.WriteLine(" provisioningStatus: " + vm.ProvisioningState);
Console.WriteLine(" id: " + vm.Id);
Console.WriteLine(" name: " + vm.Name);
Console.WriteLine(" type: " + vm.Type);
Console.WriteLine(" location: " + vm.Region);
Console.WriteLine("VM instance status");
foreach (InstanceViewStatus stat in vm.InstanceView.Statuses)
{
}
Console.WriteLine("Press enter to continue...");
Console.ReadLine();
Stop the VM
You can stop a virtual machine and keep all its settings, but continue to be charged for it, or you can stop a virtual
machine and deallocate it. When a virtual machine is deallocated, all resources associated with it are also
deallocated and billing ends for it.
To stop the virtual machine without deallocating it, add this code to the Main method:
Console.WriteLine("Stopping vm...");
vm.PowerOff();
Console.ReadLine();
If you want to deallocate the virtual machine, change the PowerOff call to this code:
vm.Deallocate();
Start the VM
To start the virtual machine, add this code to the Main method:
Console.WriteLine("Starting vm...");
vm.Start();
Console.ReadLine();
Resize the VM
Many aspects of deployment should be considered when deciding on a size for your virtual machine. For more
information, see VM sizes.
To change size of the virtual machine, add this code to the Main method:
Console.WriteLine("Resizing vm...");
vm.Update()
.Apply();
Console.ReadLine();
Add a data disk to the VM

To add a data disk to the virtual machine, add this code to the Main method to add a data disk that is 2 GB in size,
han a LUN of 0 and a caching type of ReadWrite:
Console.WriteLine("Adding data disk to vm...");

vm.Update()
.WithNewDataDisk(2, 0, CachingTypes.ReadWrite)
.Apply();
Console.WriteLine("Press enter to delete resources...");
Console.ReadLine();
Delete resources
Because you are charged for resources used in Azure, it is always good practice to delete resources that are no
longer needed. If you want to delete the virtual machines and all the supporting resources, all you have to do is
delete the resource group.
To delete the resource group, add this code to the Main method:
azure.ResourceGroups.DeleteByName(groupName);
Run the application

It should take about five minutes for this console application to run completely from start to finish.
1. To run the console application, click Start.
2. Before you press Enter to start deleting resources, you could take a few minutes to verify the creation of the
resources in the Azure portal. Click the deployment status to see information about the deployment.
Next steps
Take advantage of using a template to create a virtual machine by using the information in Deploy an Azure
Virtual Machine using C# and a Resource Manager template.
Learn more about using the Azure libraries for .NET.
Create a VM from a VHD by using the Azure portal
There are several ways to create a virtual machine (VM ) in Azure:

If you already have a virtual hard disk (VHD ) to use or you want to copy the VHD from an existing VM to
use, you can create a new VM by attaching the VHD to the new VM as an OS disk.
You can create a new VM from the VHD of a VM that has been deleted. For example, if you have an Azure
VM that isn't working correctly, you can delete the VM and use its VHD to create a new VM. You can either
reuse the same VHD or create a copy of the VHD by creating a snapshot and then creating a new managed
disk from the snapshot. Although creating a snapshot takes a few more steps, it preserves the original VHD
and provides you with a fallback.
You can create an Azure VM from an on-premises VHD by uploading the on-premises VHD and attaching it
to a new VM. You use PowerShell or another tool to upload the VHD to a storage account, and then you
create a managed disk from the VHD. For more information, see Upload a specialized VHD.
Don't use a specialized disk if you want to create multiple VMs. Instead, for larger deployments, create an image
and then use that image to create multiple VMs.
Copy a disk
Create a snapshot and then create a disk from the snapshot. This strategy allows you to keep the original VHD as a
fallback:
1. From the Azure portal, on the left menu, select All services.
2. In the All services search box, enter disks and then select Disks to display the list of available disks.
3. Select the disk that you would like to use. The Disk page for that disk appears.
4. From the menu at the top, select Create snapshot.
5. Enter a Name for the snapshot.
6. Choose a Resource group for the snapshot. You can use either an existing resource group or create a new one.
7. For Account type, choose either Standard (HDD ) or Premium (SSD ) storage.
8. When you're done, select Create to create the snapshot.
9. After the snapshot has been created, select Create a resource in the left menu.
10. In the search box, enter managed disk and then select Managed Disks from the list.
11. On the Managed Disks page, select Create.
12. Enter a Name for the disk.
13. Choose a Resource group for the disk. You can use either an existing resource group or create a new one. This
selection will also be used as the resource group where you create the VM from the disk.
14. For Account type, choose either Standard (HDD ) or Premium (SSD ) storage.
15. In Source type, ensure Snapshot is selected.
16. In the Source snapshot drop-down, select the snapshot you want to use.
17. Make any other adjustments as needed and then select Create to create the disk.
Create a VM from a disk

After you have the managed disk VHD that you want to use, you can create the VM in the portal:
1. From the Azure portal, on the left menu, select All services.
2. In the All services search box, enter disks and then select Disks to display the list of available disks.
3. Select the disk that you would like to use. The Disk page for that disk opens.
4. In the Overview page, ensure that DISK STATE is listed as Unattached. If it isn't, you might need to either
detach the disk from the VM or delete the VM to free up the disk.
5. In the menu at the top of the page, select Create VM.
6. On the Basics page for the new VM, enter a Virtual machine name and either select an existing Resource
group or create a new one.
7. For Size, select Change size to access the Size page.
8. Select a VM size row and then choose Select.
9. On the Networking page, you can either let the portal create all new resources or you can select an existing
Virtual network and Network security group. The portal always creates a new network interface and public
IP address for the new VM.
10. On the Management page, make any changes to the monitoring options.
11. On the Guest config page, add any extensions as needed.
12. When you're done, select Review + create.
13. If the VM configuration passes validation, select Create to start the deployment.
Next steps
You can also use PowerShell to upload a VHD to Azure and create a specialized VM.
Create a Windows VM from a specialized disk by
using PowerShell
Create a new VM by attaching a specialized managed disk as the OS disk. A specialized disk is a copy of a virtual
hard disk (VHD ) from an existing VM that contains the user accounts, applications, and other state data from your
original VM.
When you use a specialized VHD to create a new VM, the new VM retains the computer name of the original VM.
Other computer-specific information is also kept and, in some cases, this duplicate information could cause issues.
When copying a VM, be aware of what types of computer-specific information your applications rely on.
You have several options:
Use an existing managed disk. This option is useful if you have a VM that isn't working correctly. You can delete
the VM and then reuse the managed disk to create a new VM.
Upload a VHD
Copy an existing Azure VM by using snapshots
You can also use the Azure portal to create a new VM from a specialized VHD.
This article shows you how to use managed disks. If you have a legacy deployment that requires using a storage
account, see Create a VM from a specialized VHD in a storage account.
Before you begin

To use PowerShell, make sure that you have the latest version of the AzureRM.Compute PowerShell module.
Install-Module AzureRM -RequiredVersion 6.0.0
For more information, see Azure PowerShell.
Option 1: Use an existing disk

If you had a VM that you deleted and you want to reuse the OS disk to create a new VM, use Get-AzureRmDisk.
$resourceGroupName = 'myResourceGroup'
$osDiskName = 'myOsDisk'
$osDisk = Get-AzureRmDisk `
-ResourceGroupName $resourceGroupName `
-DiskName $osDiskName
You can now attach this disk as the OS disk to a new VM.
Option 2: Upload a specialized VHD

You can upload the VHD from a specialized VM created with an on-premises virtualization tool, like Hyper-V, or a
VM exported from another cloud.
Prepare the VM
Use the VHD as-is to create a new VM.
Prepare a Windows VHD to upload to Azure. Do not generalize the VM by using Sysprep.
Remove any guest virtualization tools and agents that are installed on the VM (such as VMware tools).
Make sure the VM is configured to get the IP address and DNS settings from DHCP. This ensures that the
server obtains an IP address within the virtual network when it starts up.
Get the storage account
You'll need a storage account in Azure to store the uploaded VHD. You can either use an existing storage account
or create a new one.
Show the available storage accounts.
Get-AzureRmStorageAccount
To use an existing storage account, proceed to the Upload the VHD section.
Create a storage account.
1. You'll need the name of the resource group where the storage account will be created. Use Get-
AzureRmResourceGroup see all the resource groups that are in your subscription.
Get-AzureRmResourceGroup
Create a resource group named myResourceGroup in the West US region.
New-AzureRmResourceGroup `
-Name myResourceGroup `
-Location "West US"
2. Create a storage account named mystorageaccount in the new resource group by using the New -
AzureRmStorageAccount cmdlet.
New-AzureRmStorageAccount `
-Name mystorageaccount `
-Location "West US" `
-SkuName "Standard_LRS" `
-Kind "Storage"
Upload the VHD to your storage account

Use the Add-AzureRmVhd cmdlet to upload the VHD to a container in your storage account. This example
uploads the file myVHD.vhd from "C:\Users\Public\Documents\Virtual hard disks" to a storage account named
mystorageaccount in the myResourceGroup resource group. The file is stored in the container named mycontainer
and the new file name will be myUploadedVHD.vhd.
$resourceGroupName = "myResourceGroup"
$urlOfUploadedVhd = "https://mystorageaccount.blob.core.windows.net/mycontainer/myUploadedVHD.vhd"
Add-AzureRmVhd -ResourceGroupName $resourceGroupName `
-Destination $urlOfUploadedVhd `
-LocalFilePath "C:\Users\Public\Documents\Virtual hard disks\myVHD.vhd"
If the commands are successful, you'll get a response that looks similar to this:
MD5 hash is being calculated for the file C:\Users\Public\Documents\Virtual hard disks\myVHD.vhd.
MD5 hash calculation is completed.
Elapsed time for the operation: 00:03:35
Creating new page blob of size 53687091712...
Elapsed time for upload: 01:12:49
LocalFilePath DestinationUri
------------- --------------
C:\Users\Public\Doc... https://mystorageaccount.blob.core.windows.net/mycontainer/myUploadedVHD.vhd
This command may take a while to complete, depending on your network connection and the size of your VHD
file.
Create a managed disk from the VHD
Create a managed disk from the specialized VHD in your storage account by using New -AzureRMDisk. This
example uses myOSDisk1 for the disk name, puts the disk in Standard_LRS storage, and uses
https://storageaccount.blob.core.windows.net/vhdcontainer/osdisk.vhd as the URI for the source VHD.
Create a new resource group for the new VM.
$destinationResourceGroup = 'myDestinationResourceGroup'
New-AzureRmResourceGroup -Location $location `
-Name $destinationResourceGroup
Create the new OS disk from the uploaded VHD.
$sourceUri = 'https://storageaccount.blob.core.windows.net/vhdcontainer/osdisk.vhd'
$osDisk = New-AzureRmDisk -DiskName $osDiskName -Disk `
(New-AzureRmDiskConfig -AccountType Standard_LRS `
-Location $location -CreateOption Import `
-SourceUri $sourceUri) `
-ResourceGroupName $destinationResourceGroup
Option 3: Copy an existing Azure VM

You can create a copy of a VM that uses managed disks by taking a snapshot of the VM, and then by using that
snapshot to create a new managed disk and a new VM.
Take a snapshot of the OS disk
You can take a snapshot of an entire VM (including all disks) or of just a single disk. The following steps show you
how to take a snapshot of just the OS disk of your VM with the New -AzureRmSnapshot cmdlet.
First, set some parameters.
$vmName = 'myVM'
$location = 'westus'
$snapshotName = 'mySnapshot'
Get the VM object.
$vm = Get-AzureRmVM -Name $vmName `

-ResourceGroupName $resourceGroupName
Get the OS disk name.
$disk = Get-AzureRmDisk -ResourceGroupName $resourceGroupName `

-DiskName $vm.StorageProfile.OsDisk.Name
Create the snapshot configuration.
$snapshotConfig = New-AzureRmSnapshotConfig `
-SourceUri $disk.Id `
-OsType Windows `
-CreateOption Copy `
-Location $location
Take the snapshot.
$snapShot = New-AzureRmSnapshot `
-Snapshot $snapshotConfig `
-SnapshotName $snapshotName `
To use this snapshot to create a VM that needs to be high-performing, add the parameter
-AccountType Premium_LRS to the New -AzureRmSnapshot command. This parameter creates the snapshot so that
it's stored as a Premium Managed Disk. Premium Managed Disks are more expensive than Standard, so be sure
you'll need Premium before using this parameter.
Create a new disk from the snapshot
Create a managed disk from the snapshot by using New -AzureRMDisk. This example uses myOSDisk for the disk
name.
Create a new resource group for the new VM.
$destinationResourceGroup = 'myDestinationResourceGroup'
New-AzureRmResourceGroup -Location $location `
-Name $destinationResourceGroup
Set the OS disk name.
Create the managed disk.
$osDisk = New-AzureRmDisk -DiskName $osDiskName -Disk `

(New-AzureRmDiskConfig -Location $location -CreateOption Copy `
-SourceResourceId $snapshot.Id) `
-ResourceGroupName $destinationResourceGroup
Create the new VM

Create networking and other VM resources to be used by the new VM.
Create the subnet and virtual network
Create the virtual network and subnet for the VM.
1. Create the subnet. This example creates a subnet named mySubNet, in the resource group
myDestinationResourceGroup, and sets the subnet address prefix to 10.0.0.0/24.
$subnetName = 'mySubNet'
$singleSubnet = New-AzureRmVirtualNetworkSubnetConfig `
-Name $subnetName `
2. Create the virtual network. This example sets the virtual network name to myVnetName, the location to
West US, and the address prefix for the virtual network to 10.0.0.0/16.
$vnetName = "myVnetName"
-Name $vnetName -ResourceGroupName $destinationResourceGroup `
-Subnet $singleSubnet
Create the network security group and an RDP rule

To be able to sign in to your VM with remote desktop protocol (RDP ), you'll need to have a security rule that
allows RDP access on port 3389. In our example, the VHD for the new VM was created from an existing
specialized VM, so you can use an account that existed on the source virtual machine for RDP.
This example sets the network security group (NSG ) name to myNsg and the RDP rule name to myRdpRule.
$nsgName = "myNsg"
$rdpRule = New-AzureRmNetworkSecurityRuleConfig -Name myRdpRule -Description "Allow RDP" `

-Access Allow -Protocol Tcp -Direction Inbound -Priority 110 `
-SourceAddressPrefix Internet -SourcePortRange * `
-DestinationAddressPrefix * -DestinationPortRange 3389
$nsg = New-AzureRmNetworkSecurityGroup `
-ResourceGroupName $destinationResourceGroup `
-Name $nsgName -SecurityRules $rdpRule
For more information about endpoints and NSG rules, see Opening ports to a VM in Azure by using PowerShell.
Create a public IP address and NIC
To enable communication with the virtual machine in the virtual network, you'll need a public IP address and a
network interface.
1. Create the public IP. In this example, the public IP address name is set to myIP.
$ipName = "myIP"
$pip = New-AzureRmPublicIpAddress `
-Name $ipName -ResourceGroupName $destinationResourceGroup `
-AllocationMethod Dynamic
2. Create the NIC. In this example, the NIC name is set to myNicName.
$nicName = "myNicName"
$nic = New-AzureRmNetworkInterface -Name $nicName `
-ResourceGroupName $destinationResourceGroup `
-Location $location -SubnetId $vnet.Subnets[0].Id `
-PublicIpAddressId $pip.Id `
-NetworkSecurityGroupId $nsg.Id
Set the VM name and size

This example sets the VM name to myVM and the VM size to Standard_A2.
$vmName = "myVM"
$vmConfig = New-AzureRmVMConfig -VMName $vmName -VMSize "Standard_A2"
Add the NIC
$vm = Add-AzureRmVMNetworkInterface -VM $vmConfig -Id $nic.Id
Add the OS disk

Add the OS disk to the configuration by using Set-AzureRmVMOSDisk. This example sets the size of the disk to
128 GB and attaches the managed disk as a Windows OS disk.
$vm = Set-AzureRmVMOSDisk -VM $vm -ManagedDiskId $osDisk.Id -StorageAccountType Standard_LRS `

-DiskSizeInGB 128 -CreateOption Attach -Windows
Complete the VM
Create the VM by using New -AzureRMVM with the configurations that we just created.
New-AzureRmVM -ResourceGroupName $destinationResourceGroup -Location $location -VM $vm
If this command is successful, you'll see output like this:
RequestId IsSuccessStatusCode StatusCode ReasonPhrase

--------- ------------------- ---------- ------------
True OK OK
Verify that the VM was created

You should see the newly created VM either in the Azure portal under Browse > Virtual machines, or by using
the following PowerShell commands.
$vmList = Get-AzureRmVM -ResourceGroupName $destinationResourceGroup

$vmList.Name
Next steps
Sign in to your new virtual machine. For more information, see How to connect and log on to an Azure virtual
machine running Windows.
Deploy an Azure Virtual Machine using C# and a
Resource Manager template
This article shows you how to deploy an Azure Resource Manager template using C#. The template that you create
deploys a single virtual machine running Windows Server in a new virtual network with a single subnet.
For a detailed description of the virtual machine resource, see Virtual machines in an Azure Resource Manager
template. For more information about all the resources in a template, see Azure Resource Manager template
walkthrough.

In this step, you make sure that Visual Studio is installed and you create a console application used to deploy the
template.
1. If you haven't already, install Visual Studio. Select .NET desktop development on the Workloads page, and
then click Install. In the summary, you can see that .NET Framework 4 - 4.6 development tools is
automatically selected for you. If you have already installed Visual Studio, you can add the .NET workload using
the Visual Studio Launcher.
2. In Visual Studio, click File > New > Project.
3. In Templates > Visual C#, select Console App (.NET Framework), enter myDotnetProject for the name of
the project, select the location of the project, and then click OK.
Install the packages

NuGet packages are the easiest way to install the libraries that you need to finish these steps. To get the libraries
that you need in Visual Studio, do these steps:
1. Click Tools > Nuget Package Manager, and then click Package Manager Console.
2. Type these commands in the console:
Install-Package Microsoft.Azure.Management.Fluent
Install-Package WindowsAzure.Storage
Create the files

In this step, you create a template file that deploys the resources and a parameters file that supplies parameter
values to the template. You also create an authorization file that is used to perform Azure Resource Manager
operations.
Create the template file
Items. Name the file CreateVMTemplate.json, and then click Add.
2. Add this JSON code to the file that you created:
{
"$schema": "https://schema.management.azure.com/schemas/2015-01-01/deploymentTemplate.json#",
"contentVersion": "1.0.0.0",
"parameters": {
"adminUsername": { "type": "string" },
"adminPassword": { "type": "securestring" }
},
"variables": {
"vnetID": "[resourceId('Microsoft.Network/virtualNetworks','myVNet')]",
"subnetRef": "[concat(variables('vnetID'),'/subnets/mySubnet')]",
},
"resources": [
{
"apiVersion": "2016-03-30",
"type": "Microsoft.Network/publicIPAddresses",
"name": "myPublicIPAddress",
"properties": {
"publicIPAllocationMethod": "Dynamic",
"dnsSettings": {
"domainNameLabel": "myresourcegroupdns1"
}
}
},
{
"apiVersion": "2016-03-30",
"type": "Microsoft.Network/virtualNetworks",
"name": "myVNet",
"properties": {
"addressSpace": { "addressPrefixes": [ "10.0.0.0/16" ] },
"subnets": [
{
"name": "mySubnet",
"properties": { "addressPrefix": "10.0.0.0/24" }
}
]
}
},
{
"apiVersion": "2016-03-30",
"type": "Microsoft.Network/networkInterfaces",
"name": "myNic",
"dependsOn": [
"[resourceId('Microsoft.Network/publicIPAddresses/', 'myPublicIPAddress')]",
"[resourceId('Microsoft.Network/virtualNetworks/', 'myVNet')]"
],
"properties": {
"ipConfigurations": [
{
"name": "ipconfig1",
"properties": {
"privateIPAllocationMethod": "Dynamic",
"publicIPAddress": { "id": "
[resourceId('Microsoft.Network/publicIPAddresses','myPublicIPAddress')]" },
"subnet": { "id": "[variables('subnetRef')]" }
}
}
]
}
},
{
"apiVersion": "2016-04-30-preview",
"name": "myVM",
"dependsOn": [
"[resourceId('Microsoft.Network/networkInterfaces/', 'myNic')]"
],
"properties": {
"properties": {
"hardwareProfile": { "vmSize": "Standard_DS1" },
"osProfile": {
"computerName": "myVM",
"adminUsername": "[parameters('adminUsername')]",
"adminPassword": "[parameters('adminPassword')]"
},
"storageProfile": {
"imageReference": {
"publisher": "MicrosoftWindowsServer",
"offer": "WindowsServer",
"sku": "2012-R2-Datacenter",
"version": "latest"
},
"osDisk": {
"name": "myManagedOSDisk",
"caching": "ReadWrite",
"createOption": "FromImage"
}
},
"networkProfile": {
"networkInterfaces": [
{
"id": "[resourceId('Microsoft.Network/networkInterfaces','myNic')]"
}
]
}
}
}
]
}
3. Save the CreateVMTemplate.json file.

Create the parameters file
To specify values for the resource parameters that are defined in the template, you create a parameters file that
contains the values.
Items. Name the file Parameters.json, and then click Add.
2. Add this JSON code to the file that you created:
{
"$schema": "https://schema.management.azure.com/schemas/2015-01-01/deploymentTemplate.json",
"parameters": {
"adminUserName": { "value": "azureuser" },
"adminPassword": { "value": "Azure12345678" }
}
}
3. Save the Parameters.json file.

Before you can deploy a template, make sure that you have access to an Active Directory service principal. From
the service principal, you acquire a token for authenticating requests to Azure Resource Manager. You should also
record the application ID, the authentication key, and the tenant ID that you need in the authorization file.
Items. Name the file azureauth.properties, and then click Add.
2. Add these authorization properties:
tenant=<tenant-id>
identifier.
3. Save the azureauth.properties file.
4. Set an environment variable in Windows named AZURE_AUTH_LOCATION with the full path to
authorization file that you created, for example the following PowerShell command can be used:
[Environment]::SetEnvironmentVariable("AZURE_AUTH_LOCATION", "C:\Visual Studio

2017\Projects\myDotnetProject\myDotnetProject\azureauth.properties", "User")

1. Open the Program.cs file for the project that you created, and then add these using statements to the
existing statements at top of the file:
using Microsoft.Azure.Management.Compute.Fluent;
using Microsoft.Azure.Management.Compute.Fluent.Models;
using Microsoft.Azure.Management.Fluent;
using Microsoft.Azure.Management.ResourceManager.Fluent;
using Microsoft.Azure.Management.ResourceManager.Fluent.Core;
using Microsoft.WindowsAzure.Storage;
using Microsoft.WindowsAzure.Storage.Blob;
2. To create the management client, add this code to the Main method:
var credentials = SdkContext.AzureCredentialsFactory

.FromFile(Environment.GetEnvironmentVariable("AZURE_AUTH_LOCATION"));
var azure = Azure

.Configure()
.WithLogLevel(HttpLoggingDelegatingHandler.Level.Basic)
.Authenticate(credentials)
.WithDefaultSubscription();

To specify values for the application, add code to the Main method:
var groupName = "myResourceGroup";

var location = Region.USWest;
var resourceGroup = azure.ResourceGroups.Define(groupName)

.Create();
Create a storage account
The template and parameters are deployed from a storage account in Azure. In this step, you create the account
and upload the files.
To create the account, add this code to the Main method:
string storageAccountName = SdkContext.RandomResourceName("st", 10);
Console.WriteLine("Creating storage account...");

var storage = azure.StorageAccounts.Define(storageAccountName)
.WithRegion(Region.USWest)
.WithExistingResourceGroup(resourceGroup)
.Create();
var storageKeys = storage.GetKeys();

string storageConnectionString = "DefaultEndpointsProtocol=https;"
+ "AccountName=" + storage.Name
+ ";AccountKey=" + storageKeys[0].Value
+ ";EndpointSuffix=core.windows.net";
var account = CloudStorageAccount.Parse(storageConnectionString);

var serviceClient = account.CreateCloudBlobClient();
Console.WriteLine("Creating container...");
var container = serviceClient.GetContainerReference("templates");
container.CreateIfNotExistsAsync().Wait();
var containerPermissions = new BlobContainerPermissions()
{ PublicAccess = BlobContainerPublicAccessType.Container };
container.SetPermissionsAsync(containerPermissions).Wait();
Console.WriteLine("Uploading template file...");

var templateblob = container.GetBlockBlobReference("CreateVMTemplate.json");
templateblob.UploadFromFileAsync("..\\..\\CreateVMTemplate.json").Result();
Console.WriteLine("Uploading parameters file...");

var paramblob = container.GetBlockBlobReference("Parameters.json");
paramblob.UploadFromFileAsync("..\\..\\Parameters.json").Result();
Deploy the template

Deploy the template and parameters from the storage account that was created.
To deploy the template, add this code to the Main method:
var templatePath = "https://" + storageAccountName + ".blob.core.windows.net/templates/CreateVMTemplate.json";

var paramPath = "https://" + storageAccountName + ".blob.core.windows.net/templates/Parameters.json";
var deployment = azure.Deployments.Define("myDeployment")
.WithTemplateLink(templatePath, "1.0.0.0")
.WithParametersLink(paramPath, "1.0.0.0")
.WithMode(Microsoft.Azure.Management.ResourceManager.Fluent.Models.DeploymentMode.Incremental)
.Create();
Console.WriteLine("Press enter to delete the resource group...");
Console.ReadLine();
Delete the resources

longer needed. You don’t need to delete each resource separately from a resource group. Delete the resource
group and all its resources are automatically deleted.
To delete the resource group, add this code to the Main method:
azure.ResourceGroups.DeleteByName(groupName);
Run the application

1. To run the console application, click Start.
Next steps
If there were issues with the deployment, a next step would be to look at Troubleshoot common Azure
deployment errors with Azure Resource Manager.
Learn how to deploy a virtual machine and its supporting resources by reviewing Deploy an Azure Virtual
Machine Using C#.
Automating Azure virtual machine deployment with
Chef
NOTE
Chef is a great tool for delivering automation and desired state configurations.
With the latest cloud API release, Chef provides seamless integration with Azure, giving you the ability to provision
and deploy configuration states through a single command.
In this article, you set up your Chef environment to provision Azure virtual machines and walk through creating a
policy or “CookBook” and then deploying this cookbook to an Azure virtual machine.
Let’s begin!
Chef basics
Before you begin, review the basic concepts of Chef.
The following diagram depicts the high-level Chef architecture.
Chef has three main architectural components: Chef Server, Chef Client (node), and Chef Workstation.
The Chef Server is the management point and there are two options for the Chef Server: a hosted solution or an
on-premises solution. We will be using a hosted solution for this tutorial.
The Chef Client (node) is the agent that sits on the servers you are managing.
The Chef Workstation is the admin workstation where we create policies and execute management commands. We
run the knife command from the Chef Workstation to manage the infrastructure.
There is also the concept of “Cookbooks” and “Recipes”. These are effectively the policies we define and apply to
the servers.
Preparing the workstation

First, lets prep the workstation. I’m using a standard Windows workstation. We need to create a directory to store
the config files and cookbooks.
First create a directory called C:\chef.
Then create a second directory called c:\chef\cookbooks.
We now need to download the Azure settings file so Chef can communicate with the Azure subscription.
Download your publish settings using the PowerShell Azure Get-AzurePublishSettingsFile command.
Save the publish settings file in C:\chef.
Creating a managed Chef account
Sign up for a hosted Chef account here.
During the signup process, you will be asked to create a new organization.
Once your organization is created, download the starter kit.
NOTE
If you receive a prompt warning you that your keys will be reset, it’s okay to proceed as we have no existing infrastructure
configured as yet.
This starter kit zip file contains your organization config files and keys.
Configuring the Chef workstation

Extract the content of the chef-starter.zip to C:\chef.
Copy all files under chef-starter\chef-repo.chef to your c:\chef directory.
Your directory should now look something like the following example.
You should now have four files including the Azure publishing file in the root of c:\chef.
The PEM files contain your organization and admin private keys for communication while the knife.rb file contains
your knife configuration. We will need to edit the knife.rb file.
Open the file in your editor of choice and modify the “cookbook_path” by removing the /../ from the path so it
appears as:
cookbook_path ["#{current_dir}/cookbooks"]
Also add the following line reflecting the name of your Azure publish settings file.
knife[:azure_publish_settings_file] = "yourfilename.publishsettings"
Your knife.rb file should now look similar to the following example.
These lines will ensure that Knife references the cookbooks directory under c:\chef\cookbooks, and also uses our
Azure Publish Settings file during Azure operations.
Installing the Chef Development Kit

Next, download and install the ChefDK (Chef Development Kit) to set up your Chef Workstation.
Install in the default location of c:\opscode. This install will take around 10 minutes.
Confirm your PATH variable contains entries for
C:\opscode\chefdk\bin;C:\opscode\chefdk\embedded\bin;c:\users\yourusername.chefdk\gem\ruby\2.0.0\bin
If they are not there, make sure you add these paths!
NOTE
The order of the path is important! If your opscode paths are not in the correct order you will have issues.
Reboot your workstation before you continue.

Next, we will install the Knife Azure extension. This provides Knife with the “Azure Plugin”.
Run the following command.
chef gem install knife-azure ––pre
NOTE
The –pre argument ensures you are receiving the latest RC version of the Knife Azure Plugin which provides access to the
latest set of APIs.
It’s likely that a number of dependencies will also be installed at the same time.
To ensure everything is configured correctly, run the following command.
knife azure image list
If everything is configured correctly, you will see a list of available Azure images scroll through.
Congratulations. The workstation is set up!
Creating a Cookbook
A Cookbook is used by Chef to define a set of commands that you wish to execute on your managed client.
Creating a Cookbook is straightforward and we use the chef generate cookbook command to generate the
Cookbook template. I will be calling my Cookbook web server as I would like a policy that automatically deploys
IIS.
Under your C:\Chef directory run the following command.
chef generate cookbook webserver
This will generate a set of files under the directory C:\Chef\cookbooks\webserver. We now need to define the set of
commands we would like the Chef client to execute on the managed virtual machine.
The commands are stored in the file default.rb. In this file, I’ll be defining a set of commands that installs IIS, starts
IIS and copies a template file to the wwwroot folder.
Modify the C:\chef\cookbooks\webserver\recipes\default.rb file and add the following lines.
powershell_script 'Install IIS' do

action :run
code 'add-windowsfeature Web-Server'
end
service 'w3svc' do
action [ :enable, :start ]
end
template 'c:\inetpub\wwwroot\Default.htm' do
source 'Default.htm.erb'
rights :read, 'Everyone'
end
Save the file once you are done.

Creating a template
As we mentioned previously, we need to generate a template file which will be used as the default.html page.
Run the following command to generate the template.
chef generate template webserver Default.htm
Now navigate to the C:\chef\cookbooks\webserver\templates\default\Default.htm.erb file. Edit the file by adding

some simple “Hello World” HTML code, and then save the file.
Upload the Cookbook to the Chef Server

In this step, we are taking a copy of the Cookbook that we have created on the local machine and uploading it to
the Chef Hosted Server. Once uploaded, the Cookbook will appear under the Policy tab.
knife cookbook upload webserver
Deploy a virtual machine with Knife Azure

We will now deploy an Azure virtual machine and apply the “Webserver” Cookbook which will install the IIS web
service and default web page.
In order to do this, use the knife azure server create command.
Am example of the command appears next.
knife azure server create --azure-dns-name 'diegotest01' --azure-vm-name 'testserver01' --azure-vm-size 'Small'
--azure-storage-account 'portalvhdsxxxx' --bootstrap-protocol 'cloud-api' --azure-source-image
'a699494373c04fc0bc8f2bb1389d6106__Windows-Server-2012-Datacenter-201411.01-en.us-127GB.vhd' --azure-service-
location 'Southeast Asia' --winrm-user azureuser --winrm-password 'myPassword123' --tcp-endpoints 80,3389 --r
'recipe[webserver]'
The parameters are self-explanatory. Substitute your particular variables and run.
NOTE
Through the command line, I’m also automating my endpoint network filter rules by using the –tcp-endpoints parameter. I’ve
opened up ports 80 and 3389 to provide access to my web page and RDP session.
Once you run the command, go to the Azure portal and you will see your machine begin to provision.
The command prompt appears next.

Once the deployment is complete, we should be able to connect to the web service over port 80 as we had opened
the port when we provisioned the virtual machine with the Knife Azure command. As this virtual machine is the
only virtual machine in my cloud service, I’ll connect it with the cloud service url.
As you can see, I got creative with my HTML code.

Don’t forget we can also connect through an RDP session from the Azure portal via port 3389.
I hope this has been helpful! Go and start your infrastructure as code journey with Azure today!
Create and manage Windows VMs in Azure using
Java
and deleting VM resources using Java. You learn how to:
Create a Maven project
Add dependencies
Create credentials
Create resources
Delete resources
Run the application
Create a Maven project

1. If you haven't already done so, install Java.
2. Install Maven.
3. Create a new folder and the project:
mkdir java-azure-test
cd java-azure-test
mvn archetype:generate -DgroupId=com.fabrikam -DartifactId=testAzureApp -DarchetypeArtifactId=maven-

archetype-quickstart -DinteractiveMode=false
Add dependencies
1. Under the testAzureApp folder, open the pom.xml file and add the build configuration to <project> to
enable the building of your application:
<build>
<plugins>
<plugin>
<groupId>org.codehaus.mojo</groupId>
<artifactId>exec-maven-plugin</artifactId>
<configuration>
<mainClass>com.fabrikam.testAzureApp.App</mainClass>
</configuration>
</plugin>
</plugins>
</build>
2. Add the dependencies that are needed to access the Azure Java SDK.
<dependency>
<groupId>com.microsoft.azure</groupId>
<artifactId>azure</artifactId>
<version>1.1.0</version>
</dependency>
<dependency>
<artifactId>azure-mgmt-compute</artifactId>
</dependency>
<dependency>
<artifactId>azure-mgmt-resources</artifactId>
</dependency>
<dependency>
<artifactId>azure-mgmt-network</artifactId>
</dependency>
<dependency>
<groupId>com.squareup.okio</groupId>
<artifactId>okio</artifactId>
</dependency>
<dependency>
<groupId>com.nimbusds</groupId>
<artifactId>nimbus-jose-jwt</artifactId>
<version>3.6</version>
</dependency>
<dependency>
<groupId>net.minidev</groupId>
<artifactId>json-smart</artifactId>
<version>1.0.6.3</version>
</dependency>
<dependency>
<groupId>javax.mail</groupId>
<artifactId>mail</artifactId>
</dependency>
3. Save the file.
Create credentials
Before you start this step, make sure that you have access to an Active Directory service principal. You should also
record the application ID, the authentication key, and the tenant ID that you need in a later step.
1. Create a file named azureauth.properties and add these properties to it:
tenant=<tenant-id>
identifier.
2. Save the file.
3. Set an environment variable named AZURE_AUTH_LOCATION in your shell with the full path to the
authentication file.
1. Open the App.java file under src\main\java\com\fabrikam and make sure this package statement is at the
top:
package com.fabrikam.testAzureApp;
2. Under the package statement, add these import statements:
import com.microsoft.azure.management.Azure;
import com.microsoft.azure.management.compute.AvailabilitySet;
import com.microsoft.azure.management.compute.AvailabilitySetSkuTypes;
import com.microsoft.azure.management.compute.CachingTypes;
import com.microsoft.azure.management.compute.InstanceViewStatus;
import com.microsoft.azure.management.compute.DiskInstanceView;
import com.microsoft.azure.management.compute.VirtualMachine;
import com.microsoft.azure.management.compute.VirtualMachineSizeTypes;
import com.microsoft.azure.management.network.PublicIPAddress;
import com.microsoft.azure.management.network.Network;
import com.microsoft.azure.management.network.NetworkInterface;
import com.microsoft.azure.management.resources.ResourceGroup;
import com.microsoft.azure.management.resources.fluentcore.arm.Region;
import com.microsoft.azure.management.resources.fluentcore.model.Creatable;
import com.microsoft.rest.LogLevel;
import java.io.File;
import java.util.Scanner;
3. To create the Active Directory credentials that you need to make requests, add this code to the main method
of the App class:
try {
final File credFile = new File(System.getenv("AZURE_AUTH_LOCATION"));
Azure azure = Azure.configure()
.withLogLevel(LogLevel.BASIC)
.authenticate(credFile)
.withDefaultSubscription();
} catch (Exception e) {
System.out.println(e.getMessage());
e.printStackTrace();
}
Create resources
Create the resource group
To specify values for the application and create the resource group, add this code to the try block in the main
method:
System.out.println("Creating resource group...");
ResourceGroup resourceGroup = azure.resourceGroups()
.define("myResourceGroup")
.withRegion(Region.US_EAST)
.create();
Create the availability set

To create the availability set, add this code to the try block in the main method:
System.out.println("Creating availability set...");

AvailabilitySet availabilitySet = azure.availabilitySets()
.define("myAvailabilitySet")
.withExistingResourceGroup("myResourceGroup")
.withSku(AvailabilitySetSkuTypes.MANAGED)
.create();
Create the public IP address

To create the public IP address for the virtual machine, add this code to the try block in the main method:
System.out.println("Creating public IP address...");

PublicIPAddress publicIPAddress = azure.publicIPAddresses()
.define("myPublicIP")
.withDynamicIP()
.create();
Create the virtual network

To create a subnet and a virtual network, add this code to the try block in the main method:
System.out.println("Creating virtual network...");

Network network = azure.networks()
.define("myVN")
.withAddressSpace("10.0.0.0/16")
.withSubnet("mySubnet","10.0.0.0/24")
.create();
Create the network interface

To create a network interface, add this code to the try block in the main method:
System.out.println("Creating network interface...");
NetworkInterface networkInterface = azure.networkInterfaces()
.define("myNIC")
.withExistingPrimaryNetwork(network)
.withSubnet("mySubnet")
.withPrimaryPrivateIPAddressDynamic()
.withExistingPrimaryPublicIPAddress(publicIPAddress)
.create();

To create the virtual machine, add this code to the try block in the main method:
System.out.println("Creating virtual machine...");

VirtualMachine virtualMachine = azure.virtualMachines()
.define("myVM")
.withExistingPrimaryNetworkInterface(networkInterface)
.withLatestWindowsImage("MicrosoftWindowsServer", "WindowsServer", "2012-R2-Datacenter")
.withAdminUsername("azureuser")
.withAdminPassword("Azure12345678")
.withComputerName("myVM")
.withExistingAvailabilitySet(availabilitySet)
.withSize("Standard_DS1")
.create();
Scanner input = new Scanner(System.in);
System.out.println("Press enter to get information about the VM...");
input.nextLine();
NOTE
This tutorial creates a virtual machine running a version of the Windows Server operating system. To learn more about
selecting other images, see Navigate and select Azure virtual machine images with Windows PowerShell and the Azure CLI.
If you want to use an existing disk instead of a marketplace image, use this code:
ManagedDisk managedDisk = azure.disks.define("myosdisk")

.withWindowsFromVhd("https://mystorage.blob.core.windows.net/vhds/myosdisk.vhd")
.withSizeInGB(128)
.withSku(DiskSkuTypes.PremiumLRS)
.create();
azure.virtualMachines.define("myVM")
.withExistingPrimaryNetworkInterface(networkInterface)
.withSpecializedOSDisk(managedDisk, OperatingSystemTypes.Windows)
.withExistingAvailabilitySet(availabilitySet)
.withSize(VirtualMachineSizeTypes.StandardDS1)
.create();

When you need to do anything with the VM, you need to get an instance of it. Add this code to the try block of the
main method:
VirtualMachine vm = azure.virtualMachines().getByResourceGroup("myResourceGroup", "myVM");

To get information about the virtual machine, add this code to the try block in the main method:
System.out.println("hardwareProfile");
System.out.println(" vmSize: " + vm.size());
System.out.println("storageProfile");
System.out.println(" imageReference");
System.out.println(" publisher: " + vm.storageProfile().imageReference().publisher());
System.out.println(" offer: " + vm.storageProfile().imageReference().offer());
System.out.println(" sku: " + vm.storageProfile().imageReference().sku());
System.out.println(" version: " + vm.storageProfile().imageReference().version());
System.out.println(" osDisk");
System.out.println(" osType: " + vm.storageProfile().osDisk().osType());
System.out.println(" name: " + vm.storageProfile().osDisk().name());
System.out.println(" createOption: " + vm.storageProfile().osDisk().createOption());
System.out.println(" caching: " + vm.storageProfile().osDisk().caching());
System.out.println("osProfile");
System.out.println(" computerName: " + vm.osProfile().computerName());
System.out.println(" adminUserName: " + vm.osProfile().adminUsername());
System.out.println(" provisionVMAgent: " + vm.osProfile().windowsConfiguration().provisionVMAgent());
System.out.println(" enableAutomaticUpdates: " +
vm.osProfile().windowsConfiguration().enableAutomaticUpdates());
System.out.println("networkProfile");
System.out.println(" networkInterface: " + vm.primaryNetworkInterfaceId());
System.out.println("vmAgent");
System.out.println(" vmAgentVersion: " + vm.instanceView().vmAgent().vmAgentVersion());
System.out.println(" statuses");
for(InstanceViewStatus status : vm.instanceView().vmAgent().statuses()) {
System.out.println(" code: " + status.code());
System.out.println(" displayStatus: " + status.displayStatus());
System.out.println(" message: " + status.message());
System.out.println(" time: " + status.time());
}
System.out.println("disks");
for(DiskInstanceView disk : vm.instanceView().disks()) {
System.out.println(" name: " + disk.name());
System.out.println(" statuses");
for(InstanceViewStatus status : disk.statuses()) {
System.out.println(" time: " + status.time());
}
}
System.out.println("VM general status");
System.out.println(" provisioningStatus: " + vm.provisioningState());
System.out.println(" id: " + vm.id());
System.out.println(" name: " + vm.name());
System.out.println(" type: " + vm.type());
System.out.println("VM instance status");
for(InstanceViewStatus status : vm.instanceView().statuses()) {
}
System.out.println("Press enter to continue...");
input.nextLine();
Stop the VM
To stop the virtual machine without deallocating it, add this code to the try block in the main method:
System.out.println("Stopping vm...");
vm.powerOff();
input.nextLine();
If you want to deallocate the virtual machine, change the PowerOff call to this code:
vm.deallocate();
Start the VM
To start the virtual machine, add this code to the try block in the main method:
System.out.println("Starting vm...");
vm.start();
input.nextLine();
Resize the VM
To change size of the virtual machine, add this code to the try block in the main method:
System.out.println("Resizing vm...");
vm.update()
.withSize(VirtualMachineSizeTypes.STANDARD_DS2)
.apply();
input.nextLine();

To add a data disk to the virtual machine that is 2 GB in size, has a LUN of 0, and a caching type of ReadWrite, add
this code to the try block in the main method:
System.out.println("Adding data disk...");

vm.update()
.withNewDataDisk(2, 0, CachingTypes.READ_WRITE)
.apply();
System.out.println("Press enter to delete resources...");
input.nextLine();
Delete resources
1. To delete the resource group, add this code to the try block in the main method:
System.out.println("Deleting resources...");
azure.resourceGroups().deleteByName("myResourceGroup");
1. Save the App.java file.
Run the application

1. To run the application, use this Maven command:
mvn compile exec:java
Next steps
Learn more about using the Azure libraries for Java.
Create and manage Windows VMs in Azure using
Python
and deleting VM resources using Python. You learn how to:
Install packages
Create credentials
Create resources
Delete resources
Run the application

1. If you haven't already, install Visual Studio. Select Python development on the Workloads page, and then click
Install. In the summary, you can see that Python 3 64-bit (3.6.0) is automatically selected for you. If you have
already installed Visual Studio, you can add the Python workload using the Visual Studio Launcher.
2. After installing and starting Visual Studio, click File > New > Project.
3. Click Templates > Python > Python Application, enter myPythonProject for the name of the project, select
the location of the project, and then click OK.
Install packages
1. In Solution Explorer, under myPythonProject, right-click Python Environments, and then select Add virtual
environment.
2. On the Add Virtual Environment screen, accept the default name of env, make sure that Python 3.6 (64 -bit) is
selected for the base interpreter, and then click Create.
3. Right-click the env environment that you created, click Install Python Package, enter azure in the search box,
and then press Enter.
You should see in the output windows that the azure packages were successfully installed.
Create credentials
Before you start this step, make sure that you have an Active Directory service principal. You should also record the
application ID, the authentication key, and the tenant ID that you need in a later step.
1. Open myPythonProject.py file that was created, and then add this code to enable your application to run:
if __name__ == "__main__":
2. To import the code that is needed, add these statements to the top of the .py file:
from azure.common.credentials import ServicePrincipalCredentials
from azure.mgmt.resource import ResourceManagementClient
from azure.mgmt.compute import ComputeManagementClient
from azure.mgmt.network import NetworkManagementClient
from azure.mgmt.compute.models import DiskCreateOption
3. Next in the .py file, add variables after the import statements to specify common values used in the code:
SUBSCRIPTION_ID = 'subscription-id'
GROUP_NAME = 'myResourceGroup'
LOCATION = 'westus'
VM_NAME = 'myVM'
Replace subscription-id with your subscription identifier.

4. To create the Active Directory credentials that you need to make requests, add this function after the
variables in the .py file:
def get_credentials():
credentials = ServicePrincipalCredentials(
client_id = 'application-id',
secret = 'authentication-key',
tenant = 'tenant-id'
)
return credentials
Replace application-id, authentication-key, and tenant-id with the values that you previously collected
when you created your Azure Active Directory service principal.
5. To call the function that you previously added, add this code under the if statement at the end of the .py file:
credentials = get_credentials()
Create resources
Initialize management clients
Management clients are needed to create and manage resources using the Python SDK in Azure. To create the
management clients, add this code under the if statement at then end of the .py file:
resource_group_client = ResourceManagementClient(
credentials,
SUBSCRIPTION_ID
)
network_client = NetworkManagementClient(
credentials,
SUBSCRIPTION_ID
)
compute_client = ComputeManagementClient(
credentials,
SUBSCRIPTION_ID
)
Create the VM and supporting resources

1. To create a resource group, add this function after the variables in the .py file:
def create_resource_group(resource_group_client):
resource_group_params = { 'location':LOCATION }
resource_group_result = resource_group_client.resource_groups.create_or_update(
GROUP_NAME,
resource_group_params
)
create_resource_group(resource_group_client)
input('Resource group created. Press enter to continue...')
1. To create an availability set, add this function after the variables in the .py file:
def create_availability_set(compute_client):
avset_params = {
'location': LOCATION,
'sku': { 'name': 'Aligned' },
'platform_fault_domain_count': 3
}
availability_set_result = compute_client.availability_sets.create_or_update(
GROUP_NAME,
'myAVSet',
avset_params
)
create_availability_set(compute_client)
print("------------------------------------------------------")
input('Availability set created. Press enter to continue...')

1. To create a public IP address for the virtual machine, add this function after the variables in the .py file:
def create_public_ip_address(network_client):
public_ip_addess_params = {
'public_ip_allocation_method': 'Dynamic'
}
creation_result = network_client.public_ip_addresses.create_or_update(
GROUP_NAME,
'myIPAddress',
public_ip_addess_params
)
return creation_result.result()
creation_result = create_public_ip_address(network_client)
print("------------------------------------------------------")
print(creation_result)
input('Press enter to continue...')

1. To create a virtual network, add this function after the variables in the .py file:
def create_vnet(network_client):
vnet_params = {
'address_space': {
'address_prefixes': ['10.0.0.0/16']
}
}
creation_result = network_client.virtual_networks.create_or_update(
GROUP_NAME,
'myVNet',
vnet_params
)
creation_result = create_vnet(network_client)
print("------------------------------------------------------")
3. To add a subnet to the virtual network, add this function after the variables in the .py file:
def create_subnet(network_client):
subnet_params = {
'address_prefix': '10.0.0.0/24'
}
creation_result = network_client.subnets.create_or_update(
GROUP_NAME,
'myVNet',
'mySubnet',
subnet_params
)
creation_result = create_subnet(network_client)
print("------------------------------------------------------")

1. To create a network interface, add this function after the variables in the .py file:
def create_nic(network_client):
subnet_info = network_client.subnets.get(
GROUP_NAME,
'myVNet',
'mySubnet'
)
publicIPAddress = network_client.public_ip_addresses.get(
GROUP_NAME,
'myIPAddress'
)
nic_params = {
'ip_configurations': [{
'name': 'myIPConfig',
'public_ip_address': publicIPAddress,
'subnet': {
'id': subnet_info.id
}
}]
}
creation_result = network_client.network_interfaces.create_or_update(
GROUP_NAME,
'myNic',
nic_params
)
creation_result = create_nic(network_client)
print("------------------------------------------------------")
1. To create the virtual machine, add this function after the variables in the .py file:
def create_vm(network_client, compute_client):
nic = network_client.network_interfaces.get(
GROUP_NAME,
'myNic'
)
avset = compute_client.availability_sets.get(
GROUP_NAME,
'myAVSet'
)
vm_parameters = {
'os_profile': {
'computer_name': VM_NAME,
'admin_username': 'azureuser',
'admin_password': 'Azure12345678'
},
'hardware_profile': {
'vm_size': 'Standard_DS1'
},
'storage_profile': {
'image_reference': {
'publisher': 'MicrosoftWindowsServer',
'offer': 'WindowsServer',
'sku': '2012-R2-Datacenter',
'version': 'latest'
}
},
'network_profile': {
'network_interfaces': [{
'id': nic.id
}]
},
'availability_set': {
'id': avset.id
}
}
creation_result = compute_client.virtual_machines.create_or_update(
GROUP_NAME,
VM_NAME,
vm_parameters
)
NOTE
This tutorial creates a virtual machine running a version of the Windows Server operating system. To learn more
about selecting other images, see Navigate and select Azure virtual machine images with Windows PowerShell and
the Azure CLI.
creation_result = create_vm(network_client, compute_client)

print("------------------------------------------------------")

1. To get information about the virtual machine, add this function after the variables in the .py file:
def get_vm(compute_client):
vm = compute_client.virtual_machines.get(GROUP_NAME, VM_NAME, expand='instanceView')
print("hardwareProfile")
print(" vmSize: ", vm.hardware_profile.vm_size)
print("\nstorageProfile")
print(" imageReference")
print(" publisher: ", vm.storage_profile.image_reference.publisher)
print(" offer: ", vm.storage_profile.image_reference.offer)
print(" sku: ", vm.storage_profile.image_reference.sku)
print(" version: ", vm.storage_profile.image_reference.version)
print(" osDisk")
print(" osType: ", vm.storage_profile.os_disk.os_type.value)
print(" name: ", vm.storage_profile.os_disk.name)
print(" createOption: ", vm.storage_profile.os_disk.create_option.value)
print(" caching: ", vm.storage_profile.os_disk.caching.value)
print("\nosProfile")
print(" computerName: ", vm.os_profile.computer_name)
print(" adminUsername: ", vm.os_profile.admin_username)
print(" provisionVMAgent: {0}".format(vm.os_profile.windows_configuration.provision_vm_agent))
print(" enableAutomaticUpdates:
{0}".format(vm.os_profile.windows_configuration.enable_automatic_updates))
print("\nnetworkProfile")
for nic in vm.network_profile.network_interfaces:
print(" networkInterface id: ", nic.id)
print("\nvmAgent")
print(" vmAgentVersion", vm.instance_view.vm_agent.vm_agent_version)
print(" statuses")
for stat in vm_result.instance_view.vm_agent.statuses:
print(" code: ", stat.code)
print(" displayStatus: ", stat.display_status)
print(" message: ", stat.message)
print(" time: ", stat.time)
print("\ndisks");
for disk in vm.instance_view.disks:
print(" name: ", disk.name)
print(" statuses")
for stat in disk.statuses:
print(" time: ", stat.time)
print("\nVM general status")
print(" provisioningStatus: ", vm.provisioning_state)
print(" id: ", vm.id)
print(" name: ", vm.name)
print(" type: ", vm.type)
print(" location: ", vm.location)
print("\nVM instance status")
for stat in vm.instance_view.statuses:
get_vm(compute_client)
print("------------------------------------------------------")
Stop the VM
1. To stop the virtual machine without deallocating it, add this function after the variables in the .py file:
def stop_vm(compute_client):
compute_client.virtual_machines.power_off(GROUP_NAME, VM_NAME)
If you want to deallocate the virtual machine, change the power_off call to this code:
compute_client.virtual_machines.deallocate(GROUP_NAME, VM_NAME)
stop_vm(compute_client)
Start the VM
1. To start the virtual machine, add this function after the variables in the .py file:
def start_vm(compute_client):
compute_client.virtual_machines.start(GROUP_NAME, VM_NAME)
start_vm(compute_client)
Resize the VM
1. To change the size of the virtual machine, add this function after the variables in the .py file:
def update_vm(compute_client):
vm = compute_client.virtual_machines.get(GROUP_NAME, VM_NAME)
vm.hardware_profile.vm_size = 'Standard_DS3'
update_result = compute_client.virtual_machines.create_or_update(
GROUP_NAME,
VM_NAME,
vm
)
return update_result.result()
update_result = update_vm(compute_client)
print("------------------------------------------------------")
print(update_result)

Virtual machines can have one or more data disks that are stored as VHDs.
1. To add a data disk to the virtual machine, add this function after the variables in the .py file:
def add_datadisk(compute_client):
disk_creation = compute_client.disks.create_or_update(
GROUP_NAME,
'myDataDisk1',
{
'disk_size_gb': 1,
'creation_data': {
'create_option': DiskCreateOption.empty
}
}
)
data_disk = disk_creation.result()
vm = compute_client.virtual_machines.get(GROUP_NAME, VM_NAME)
add_result = vm.storage_profile.data_disks.append({
'lun': 1,
'name': 'myDataDisk1',
'create_option': DiskCreateOption.attach,
'managed_disk': {
'id': data_disk.id
}
})
add_result = compute_client.virtual_machines.create_or_update(
GROUP_NAME,
VM_NAME,
vm)
return add_result.result()
add_result = add_datadisk(compute_client)
print("------------------------------------------------------")
print(add_result)
Delete resources
Because you are charged for resources used in Azure, it's always a good practice to delete resources that are no
1. To delete the resource group and all resources, add this function after the variables in the .py file:
def delete_resources(resource_group_client):
resource_group_client.resource_groups.delete(GROUP_NAME)
delete_resources(resource_group_client)
3. Save myPythonProject.py.
Run the application

1. To run the console application, click Start in Visual Studio.
2. Press Enter after the status of each resource is returned. In the status information, you should see a
Succeeded provisioning state. After the virtual machine is created, you have the opportunity to delete all
the resources that you create. Before you press Enter to start deleting resources, you could take a few
minutes to verify their creation in the Azure portal. If you have the Azure portal open, you might have to
refresh the blade to see new resources.
It should take about five minutes for this console application to run completely from start to finish. It may
take several minutes after the application has finished before all the resources and the resource group are
deleted.
Next steps
If there were issues with the deployment, a next step would be to look at Troubleshooting resource group
deployments with Azure portal
Learn more about the Azure Python Library
Create a Windows virtual machine from a Resource
Manager template
This article shows you how to deploy an Azure Resource Manager template using PowerShell. The template that
you create deploys a single virtual machine running Windows Server in a new virtual network with a single subnet.
For a detailed description of the virtual machine resource, see Virtual machines in an Azure Resource Manager
template. For more information about all the resources in a template, see Azure Resource Manager template
walkthrough.
It should take about five minutes to do the steps in this article.

5.3 or later. Run Get-Module -ListAvailable AzureRM to find the version. If you need to upgrade, see Install Azure

All resources must be deployed in a resource group.
1. Get a list of available locations where resources can be created.
Get-AzureRmLocation | sort DisplayName | Select DisplayName
2. Create the resource group in the location that you select. This example shows the creation of a resource
group named myResourceGroup in the West US location:
New-AzureRmResourceGroup -Name "myResourceGroup" -Location "West US"

Create the files
In this step, you create a template file that deploys the resources and a parameters file that supplies parameter
values to the template. You also create an authorization file that is used to perform Azure Resource Manager
operations.
1. Create a file named CreateVMTemplate.json and add this JSON code to it:
{
"parameters": {
"adminUsername": { "type": "string" },
"adminPassword": { "type": "securestring" }
},
"variables": {
"vnetID": "[resourceId('Microsoft.Network/virtualNetworks','myVNet')]",
"subnetRef": "[concat(variables('vnetID'),'/subnets/mySubnet')]"
},
"resources": [
{
"apiVersion": "2016-03-30",
"type": "Microsoft.Network/publicIPAddresses",
"name": "myPublicIPAddress",
"properties": {
"publicIPAllocationMethod": "Dynamic",
"dnsSettings": {
"domainNameLabel": "myresourcegroupdns1"
}
}
},
{
"apiVersion": "2016-03-30",
"type": "Microsoft.Network/virtualNetworks",
"name": "myVNet",
"properties": {
"addressSpace": { "addressPrefixes": [ "10.0.0.0/16" ] },
"subnets": [
{
"name": "mySubnet",
"properties": { "addressPrefix": "10.0.0.0/24" }
}
]
}
},
{
"apiVersion": "2016-03-30",
"type": "Microsoft.Network/networkInterfaces",
"name": "myNic",
"dependsOn": [
"[resourceId('Microsoft.Network/publicIPAddresses/', 'myPublicIPAddress')]",
"[resourceId('Microsoft.Network/virtualNetworks/', 'myVNet')]"
],
"properties": {
"ipConfigurations": [
{
"name": "ipconfig1",
"properties": {
"privateIPAllocationMethod": "Dynamic",
"publicIPAddress": { "id": "
[resourceId('Microsoft.Network/publicIPAddresses','myPublicIPAddress')]" },
"subnet": { "id": "[variables('subnetRef')]" }
}
}
}
]
}
},
{
"name": "myVM",
"dependsOn": [
"[resourceId('Microsoft.Network/networkInterfaces/', 'myNic')]"
],
"properties": {
"hardwareProfile": { "vmSize": "Standard_DS1" },
"osProfile": {
"computerName": "myVM",
},
"storageProfile": {
"imageReference": {
"sku": "2012-R2-Datacenter",
"version": "latest"
},
"osDisk": {
"name": "myManagedOSDisk",
}
},
"networkProfile": {
{
"id": "[resourceId('Microsoft.Network/networkInterfaces','myNic')]"
}
]
}
}
}
]
}
2. Create a file named Parameters.json and add this JSON code to it:
{
"$schema": "https://schema.management.azure.com/schemas/2015-01-01/deploymentParameters.json#",
"parameters": {
"adminUserName": { "value": "azureuser" },
"adminPassword": { "value": "Azure12345678" }
}
}
3. Create a new storage account and container:
$storageName = "st" + (Get-Random)

New-AzureRmStorageAccount -ResourceGroupName "myResourceGroup" -AccountName $storageName -Location "West
US" -SkuName "Standard_LRS" -Kind Storage
$accountKey = (Get-AzureRmStorageAccountKey -ResourceGroupName myResourceGroup -Name
$storageName).Value[0]
$context = New-AzureStorageContext -StorageAccountName $storageName -StorageAccountKey $accountKey
New-AzureStorageContainer -Name "templates" -Context $context -Permission Container
4. Upload the files to the storage account:
Set-AzureStorageBlobContent -File "C:\templates\CreateVMTemplate.json" -Context $context -Container

"templates"
Set-AzureStorageBlobContent -File "C:\templates\Parameters.json" -Context $context -Container templates
Change the -File paths to the location where you stored the files.
Create the resources

Deploy the template using the parameters:
$templatePath = "https://" + $storageName + ".blob.core.windows.net/templates/CreateVMTemplate.json"

$parametersPath = "https://" + $storageName + ".blob.core.windows.net/templates/Parameters.json"
New-AzureRmResourceGroupDeployment -ResourceGroupName "myResourceGroup" -Name "myDeployment" -TemplateUri
$templatePath -TemplateParameterUri $parametersPath
NOTE
You can also deploy templates and parameters from local files. To learn more, see Using Azure PowerShell with Azure Storage.
Next Steps
If there were issues with the deployment, you might take a look at Troubleshoot common Azure deployment
errors with Azure Resource Manager.
Learn how to create and manage a virtual machine in Create and manage Windows VMs with the Azure
PowerShell module.
How to connect and log on to an Azure virtual
machine running Windows
You'll use the Connect button in the Azure portal to start a Remote Desktop (RDP ) session from a Windows
desktop. First you connect to the virtual machine, and then you log on.
To connect to a Windows VM from a Mac, you will need to install an RDP client for Mac such as Microsoft
Remote Desktop.
Connect to the virtual machine

1. If you haven't already done so, sign in to the Azure portal.
2. On the left menu, select Virtual Machines.
3. Select the virtual machine from the list.
4. On the top of the page for the virtual machine, select Connect.
5. On the Connect to virtual machine page, select the appropriate options and select Download RDP file.
6. Open the downloaded RDP file and select Connect when prompted.
7. You will get a warning that the .rdp file is from an unknown publisher. This is expected. In the Remote
Desktop Connection window, select Connect to continue.
8. In the Windows Security window, select More choices and then Use a different account. Enter the
credentials for an account on the virtual machine and then select OK.
Local account: This is usually the local account user name and password that you specified when you
created the virtual machine. In this case, the domain is the name of the virtual machine and it is entered as
vmname\username.
Domain joined VM: If the VM belongs to a domain, enter the user name in the format
Domain\Username. The account also needs to either be in the Administrators group or have been granted
remote access privileges to the VM.
Domain controller: If the VM is a domain controller, enter the user name and password of a domain
administrator account for that domain.
9. Select Yes to verify the identity of the virtual machine and finish logging on.
TIP
If the Connect button in the portal is grayed-out and you are not connected to Azure via an Express Route or Site-to-Site
VPN connection, you will need to create and assign your VM a public IP address before you can use RDP. For more
information, see Public IP addresses in Azure.
Next steps
If you have difficulty connecting, see Troubleshoot Remote Desktop connections.
Azure Hybrid Benefit for Windows Server
For customers with Software Assurance, Azure Hybrid Benefit for Windows Server allows you to use your on-
premises Windows Server licenses and run Windows virtual machines on Azure at a reduced cost. You can use
Azure Hybrid Benefit for Windows Server to deploy new virtual machines with Windows OS. This article goes
over the steps on how to deploy new VMs with Azure Hybrid Benefit for Windows Server and how you can update
existing running VMs. For more information about Azure Hybrid Benefit for Windows Server licensing and cost
savings, see the Azure Hybrid Benefit for Windows Server licensing page.
IMPORTANT
Each 2-processor license or each set of 16-core licenses are entitled to two instances of up to 8 cores, or one instance of up
to 16 cores. The Azure Hybrid Benefit for Standard Edition licenses can only be used once either on-premises or in Azure.
Datacenter Edition benefits allow for simultaneous usage both on-premises and in Azure.
IMPORTANT
Using Azure Hybrid Benefit for Windows Server with any VMs running Windows Server OS are now supported in all regions,
including VMs with additional software such as SQL Server or third-party marketplace software.
NOTE
For classic VMs, only deploying new VM from on-prem custom images is supported. To take advantage of the capabilities
supported in this article, you must first migrate classic VMs to Resource Manager model.
Ways to use Azure Hybrid Benefit for Windows Server

There are few ways to use Windows virtual machines with the Azure Hybrid Benefit:
1. You can deploy VMs from one of the provided Windows Server images on the Azure Marketplace
2. You can upload a custom VM and deploy using a Resource Manager template or Azure PowerShell
3. You can toggle and convert existing VM between running with Azure Hybrid Benefit or pay on-demand cost for
Windows Server
4. You can also apply Azure Hybrid Benefit for Windows Server on virtual machine scale set as well
Create a VM with Azure Hybrid Benefit for Windows Server

All Windows Server OS based images are supported for Azure Hybrid Benefit for Windows Server. You can use
Azure platform support images or upload your own custom Windows Server images.
Portal
To create a VM with Azure Hybrid Benefit for Windows Server, use the toggle under the "Save money" section.
Powershell
New-AzureRmVm `
-Name "myVM" `
-ImageName "Win2016Datacenter" `
-LicenseType "Windows_Server"
CLI
az vm create \
--resource-group myResourceGroup \
--name myVM \
--location eastus \
--license-type Windows_Server
Template
Within your Resource Manager templates, an additional parameter licenseType must be specified. You can read
more about authoring Azure Resource Manager templates
"properties": {
"licenseType": "Windows_Server",
"hardwareProfile": {
"vmSize": "[variables('vmSize')]"
}
Convert an existing VM using Azure Hybrid Benefit for Windows Server

If you have an existing VM that you would like to convert to take advantage of Azure Hybrid Benefit for Windows
Server, you can update your VM's license type by following the instructions below.
NOTE
Changing the license type on the VM does not cause the system to reboot or cause a service interuption. It is simply an
update to a metadata flag.
Portal
From portal VM blade, you can update the VM to use Azure Hybrid Benefit by selecting "Configuration" option
and toggle the "Azure hybrid benefit" option
Powershell
Convert existing Windows Server VMs to Azure Hybrid Benefit for Windows Server
$vm = Get-AzureRmVM -ResourceGroup "rg-name" -Name "vm-name"

$vm.LicenseType = "Windows_Server"
Update-AzureRmVM -ResourceGroupName rg-name -VM $vm
Convert Windows Server VMs with benefit back to pay-as-you-go
$vm = Get-AzureRmVM -ResourceGroup "rg-name" -Name "vm-name"

$vm.LicenseType = "None"
Update-AzureRmVM -ResourceGroupName rg-name -VM $vm
CLI
Convert existing Windows Server VMs to Azure Hybrid Benefit for Windows Server
az vm update --resource-group myResourceGroup --name myVM --set licenseType=Windows_Server
How to verify your VM is utilizing the licensing benefit

Once you have deployed your VM through either PowerShell, Resource Manager template or portal, you can
verify the setting in the following methods.
Portal
From portal VM blade, you can view the toggle for Azure Hybrid Benefit for Windows Server by selecting
"Configuration" tab.
Powershell
The following example shows the license type for a single VM
Get-AzureRmVM -ResourceGroup "myResourceGroup" -Name "myVM"
Output:
Type : Microsoft.Compute/virtualMachines
Location : westus
LicenseType : Windows_Server
This output contrasts with the following VM deployed without Azure Hybrid Benefit for Windows Server licensing:
Location : westus
LicenseType :
CLI
az vm get-instance-view -g MyResourceGroup -n MyVM --query '[?licenseType==Windows_Server]' -o table
NOTE
Changing the license type on the VM does not cause the system to reboot or cause a service interuption. It is a metadata
licensing flag only.
List all VMs with Azure Hybrid Benefit for Windows Server in a
subscription
To see and count all virtual machines deployed with Azure Hybrid Benefit for Windows Server, you can run the
following command from your subscription:
Portal
From the Virtual Machine or Virtual machine scale sets resource blade, you can view a list of all your VM (s) and
licensing type by configuring the table column to include "Azure Hybrid Benefit". The VM setting can either be in
"Enabled", "Not enabled" or "Not supported" state.
Powershell
$vms = Get-AzureRMVM
$vms | ?{$_.LicenseType -like "Windows_Server"} | select ResourceGroupName, Name, LicenseType
CLI
az vm list --query '[?licenseType==Windows_Server]' -o table
Deploy a Virtual Machine Scale Set with Azure Hybrid Benefit for
Windows Server
Within your virtual machine scale set Resource Manager templates, an additional parameter licenseType must be
specified within your VirtualMachineProfile property. You can do this during create or update for your scale set
through ARM template, Powershell, Azure CLI or REST.
The following example uses ARM template with a Windows Server 2016 Datacenter image:
"virtualMachineProfile": {
"storageProfile": {
"osDisk": {
},
"imageReference": {
"sku": "2016-Datacenter",
"version": "latest"
}
},
"licenseType": "Windows_Server",
"osProfile": {
"computerNamePrefix": "[parameters('vmssName')]",
}
You can also learn more about how to Modify a virtual machine scale set for more ways to update your scale set.
Next steps
Read more about How to save money with the Azure Hybrid Benefit
Read more about Frequently asked questions for Azure Hybrid Benefit
Learn more about Azure Hybrid Benefit for Windows Server licensing detailed guidance
Learn more about Azure Hybrid Benefit for Windows Server and Azure Site Recovery make migrating
applications to Azure even more cost-effective
Learn more about Windows 10 on Azure with Multitenant Hosting Right
Learn more about Using Resource Manager templates
How to deploy Windows 10 on Azure with
Multitenant Hosting Rights
For customers with Windows 10 Enterprise E3/E5 per user or Windows Virtual Desktop Access per user (User
Subscription Licenses or Add-on User Subscription Licenses), Multitenant Hosting Rights for Windows 10 allows
you to bring your Windows 10 Licenses to the cloud and run Windows 10 Virtual Machines on Azure without
paying for another license. For more information, please see Multitenant Hosting for Windows 10.
NOTE
This article shows you to implement the licensing benefit for Windows 10 Pro Desktop images on Azure Marketplace.
For Windows 7, 8.1, 10 Enterprise (x64) images on Azure Marketplace for MSDN Subscriptions, please refer to Windows
client in Azure for dev/test scenarios
For Windows Server licensing benefits, please refer to Azure Hybrid use benefits for Windows Server images.
Deploying Windows 10 Image from Azure Marketplace

For Powershell, CLI and Azure Resource Manager template deployments, the Windows 10 image can be found
with the following publishername, offer, sku.
OS PUBLISHERNAME OFFER SKU
Windows 10 Pro MicrosoftWindowsDesktop Windows-10 RS2-Pro
Windows 10 Pro N MicrosoftWindowsDesktop Windows-10 RS2-ProN
Windows 10 Pro MicrosoftWindowsDesktop Windows-10 RS3-Pro
Windows 10 Pro N MicrosoftWindowsDesktop Windows-10 RS3-ProN
Uploading Windows 10 VHD to Azure

if you are uploading a generalized Windows 10 VHD, please note Windows 10 does not have built-in
administrator account enabled by default. To enable the built-in administrator account, include the following
command as part of the Custom Script extension.
Net user <username> /active:yes
The following powershell snippet is to mark all administrator accounts as active, including the built-in
administrator. This example is useful if the built-in administrator username is unknown.
$adminAccount = Get-WmiObject Win32_UserAccount -filter "LocalAccount=True" | ? {$_.SID -Like "S-1-5-21-*-500"}
if($adminAccount.Disabled)
{
$adminAccount.Disabled = $false
$adminAccount.Put()
}
For more information:

How to upload VHD to Azure
How to prepare a Windows VHD to upload to Azure
Deploying Windows 10 with Multitenant Hosting Rights

Make sure you have installed and configured the latest Azure PowerShell. Once you have prepared your VHD,
upload the VHD to your Azure Storage account using the Add-AzureRmVhd cmdlet as follows:
Add-AzureRmVhd -ResourceGroupName "myResourceGroup" -LocalFilePath "C:\Path\To\myvhd.vhd" `

-Destination "https://mystorageaccount.blob.core.windows.net/vhds/myvhd.vhd"
Deploy using Azure Resource Manager Template Deployment Within your Resource Manager templates, an
additional parameter for licenseType can be specified. You can read more about authoring Azure Resource
Manager templates. Once you have your VHD uploaded to Azure, edit you Resource Manager template to include
the license type as part of the compute provider and deploy your template as normal:
"properties": {
"licenseType": "Windows_Client",
"vmSize": "[variables('vmSize')]"
}
Deploy via PowerShell When deploying your Windows Server VM via PowerShell, you have an additional
parameter for -LicenseType . Once you have your VHD uploaded to Azure, you create a VM using New-AzureRmVM
and specify the licensing type as follows:
New-AzureRmVM -ResourceGroupName "myResourceGroup" -Location "West US" -VM $vm -LicenseType "Windows_Client"
Verify your VM is utilizing the licensing benefit

Once you have deployed your VM through either the PowerShell or Resource Manager deployment method, verify
the license type with Get-AzureRmVM as follows:
Get-AzureRmVM -ResourceGroup "myResourceGroup" -Name "myVM"
The output is similar to the following example for Windows 10 with correct license type:
Location : westus
LicenseType : Windows_Client
This output contrasts with the following VM deployed without Azure Hybrid Use Benefit licensing, such as a VM
deployed straight from the Azure Gallery:
Location : westus
LicenseType :
Additional Information about joining Azure AD

NOTE
Azure provisions all Windows VMs with built-in administrator account, which cannot be used to join AAD. For example,
Settings > Account > Access Work or School > +Connect will not work. You must create and log on as a second
administrator account to join Azure AD manually. You can also configure Azure AD using a provisioning package, use the link
is the Next Steps section to learn more.
Next Steps
Learn more about Configuring VDA for Windows 10
Learn more about Multitenant Hosting for Windows 10
How to encrypt virtual disks on a Windows VM
For enhanced virtual machine (VM ) security and compliance, virtual disks in Azure can be encrypted. Disks are
encrypted using cryptographic keys that are secured in an Azure Key Vault. You control these cryptographic keys
and can audit their use. This article details how to encrypt virtual disks on a Windows VM using Azure PowerShell.
You can also encrypt a Linux VM using the Azure CLI.
Overview of disk encryption

Virtual disks on Windows VMs are encrypted at rest using Bitlocker. There is no charge for encrypting virtual disks
in Azure. Cryptographic keys are stored in Azure Key Vault using software-protection, or you can import or
generate your keys in Hardware Security Modules (HSMs) certified to FIPS 140-2 level 2 standards. These
cryptographic keys are used to encrypt and decrypt virtual disks attached to your VM. You retain control of these
cryptographic keys and can audit their use. An Azure Active Directory service principal provides a secure
mechanism for issuing these cryptographic keys as VMs are powered on and off.
The process for encrypting a VM is as follows:
1. Create a cryptographic key in an Azure Key Vault.
2. Configure the cryptographic key to be usable for encrypting disks.
3. To read the cryptographic key from the Azure Key Vault, create an Azure Active Directory service principal with
the appropriate permissions.
4. Issue the command to encrypt your virtual disks, specifying the Azure Active Directory service principal and
appropriate cryptographic key to be used.
5. The Azure Active Directory service principal requests the required cryptographic key from Azure Key Vault.
6. The virtual disks are encrypted using the provided cryptographic key.
Encryption process
Disk encryption relies on the following additional components:
Azure Key Vault - used to safeguard cryptographic keys and secrets used for the disk encryption/decryption
process.
If one exists, you can use an existing Azure Key Vault. You do not have to dedicate a Key Vault to
encrypting disks.
To separate administrative boundaries and key visibility, you can create a dedicated Key Vault.
Azure Active Directory - handles the secure exchanging of required cryptographic keys and authentication for
requested actions.
You can typically use an existing Azure Active Directory instance for housing your application.
The service principal provides a secure mechanism to request and be issued the appropriate
cryptographic keys. You are not developing an actual application that integrates with Azure Active
Directory.
Requirements and limitations

Supported scenarios and requirements for disk encryption:
Enabling encryption on new Windows VMs from Azure Marketplace images or custom VHD image.
Enabling encryption on existing Windows VMs in Azure.
Enabling encryption on Windows VMs that are configured using Storage Spaces.
Disabling encryption on OS and data drives for Windows VMs.
All resources (such as Key Vault, Storage account, and VM ) must be in the same Azure region and subscription.
Standard tier VMs, such as A, D, DS, G, and GS series VMs.
Disk encryption is not currently supported in the following scenarios:
Basic tier VMs.
VMs created using the Classic deployment model.
Updating the cryptographic keys on an already encrypted VM.
Integration with on-prem Key Management Service.
Create Azure Key Vault and keys

Before you start, make sure that the latest version of the Azure PowerShell module has been installed. For more
information, see How to install and configure Azure PowerShell. Throughout the command examples, replace all
example parameters with your own names, location, and key values. The following examples use a convention of
myResourceGroup, myKeyVault, myVM, etc.
The first step is to create an Azure Key Vault to store your cryptographic keys. Azure Key Vault can store keys,
secrets, or passwords that allow you to securely implement them in your applications and services. For virtual disk
encryption, you create a Key Vault to store a cryptographic key that is used to encrypt or decrypt your virtual disks.
Enable the Azure Key Vault provider within your Azure subscription with Register-AzureRmResourceProvider, then
create a resource group with New -AzureRmResourceGroup. The following example creates a resource group
name myResourceGroup in the East US location:
$rgName = "myResourceGroup"
Register-AzureRmResourceProvider -ProviderNamespace "Microsoft.KeyVault"

New-AzureRmResourceGroup -Location $location -Name $rgName
The Azure Key Vault containing the cryptographic keys and associated compute resources such as storage and the
VM itself must reside in the same region. Create an Azure Key Vault with New -AzureRmKeyVault and enable the
Key Vault for use with disk encryption. Specify a unique Key Vault name for keyVaultName as follows:
$keyVaultName = "myUniqueKeyVaultName"
New-AzureRmKeyVault -Location $location `
-ResourceGroupName $rgName `
-VaultName $keyVaultName `
-EnabledForDiskEncryption
You can store cryptographic keys using software or Hardware Security Model (HSM ) protection. Using an HSM
requires a premium Key Vault. There is an additional cost to creating a premium Key Vault rather than standard Key
Vault that stores software-protected keys. To create a premium Key Vault, in the preceding step add the -Sku
"Premium" parameters. The following example uses software-protected keys since we created a standard Key Vault.
For both protection models, the Azure platform needs to be granted access to request the cryptographic keys when
the VM boots to decrypt the virtual disks. Create a cryptographic key in your Key Vault with Add-
AzureKeyVaultKey. The following example creates a key named myKey:
Add-AzureKeyVaultKey -VaultName $keyVaultName `
-Name "myKey" `
-Destination "Software"
Create the Azure Active Directory service principal

When virtual disks are encrypted or decrypted, you specify an account to handle the authentication and
exchanging of cryptographic keys from Key Vault. This account, an Azure Active Directory service principal, allows
the Azure platform to request the appropriate cryptographic keys on behalf of the VM. A default Azure Active
Directory instance is available in your subscription, though many organizations have dedicated Azure Active
Directory directories.
Create a service principal in Azure Active Directory with New -AzureRmADServicePrincipal. To specify a secure
password, follow the Password policies and restrictions in Azure Active Directory:
$appName = "My App"

$securePassword = ConvertTo-SecureString -String "P@ssw0rd!" -AsPlainText -Force
$app = New-AzureRmADApplication -DisplayName $appName `
-HomePage "https://myapp.contoso.com" `
-IdentifierUris "https://contoso.com/myapp" `
-Password $securePassword
New-AzureRmADServicePrincipal -ApplicationId $app.ApplicationId
To successfully encrypt or decrypt virtual disks, permissions on the cryptographic key stored in Key Vault must be
set to permit the Azure Active Directory service principal to read the keys. Set permissions on your Key Vault with
Set-AzureRmKeyVaultAccessPolicy:
Set-AzureRmKeyVaultAccessPolicy -VaultName $keyvaultName `

-ServicePrincipalName $app.ApplicationId `
-PermissionsToKeys "WrapKey" `
-PermissionsToSecrets "Set"

To test the encryption process, create a VM with New -AzureRmVm. The following example creates a VM named
myVM using a Windows Server 2016 Datacenter image. When prompted for credentials, enter the username and
password to be used for your VM:
New-AzureRmVm `
-Name "myVM" `
-Credential $cred
Encrypt virtual machine

To encrypt the virtual disks, you bring together all the previous components:
1. Specify the Azure Active Directory service principal and password.
2. Specify the Key Vault to store the metadata for your encrypted disks.
3. Specify the cryptographic keys to use for the actual encryption and decryption.
4. Specify whether you want to encrypt the OS disk, the data disks, or all.
Encrypt your VM with Set-AzureRmVMDiskEncryptionExtension using the Azure Key Vault key and Azure Active
Directory service principal credentials. The following example retrieves all the key information then encrypts the
VM named myVM:
$keyVault = Get-AzureRmKeyVault -VaultName $keyVaultName -ResourceGroupName $rgName;

$diskEncryptionKeyVaultUrl = $keyVault.VaultUri;
$keyVaultResourceId = $keyVault.ResourceId;
$keyEncryptionKeyUrl = (Get-AzureKeyVaultKey -VaultName $keyVaultName -Name myKey).Key.kid;
Set-AzureRmVMDiskEncryptionExtension -ResourceGroupName $rgName `

-VMName "myVM" `
-AadClientID $app.ApplicationId `
-AadClientSecret (New-Object PSCredential "user",$securePassword).GetNetworkCredential().Password `
-DiskEncryptionKeyVaultUrl $diskEncryptionKeyVaultUrl `
-DiskEncryptionKeyVaultId $keyVaultResourceId `
-KeyEncryptionKeyUrl $keyEncryptionKeyUrl `
-KeyEncryptionKeyVaultId $keyVaultResourceId
Accept the prompt to continue with the VM encryption. The VM restarts during the process. Once the encryption
process completes and the VM has rebooted, review the encryption status with Get-
AzureRmVmDiskEncryptionStatus:
Get-AzureRmVmDiskEncryptionStatus -ResourceGroupName $rgName -VMName "myVM"
The output is similar to the following example:
OsVolumeEncrypted : Encrypted
DataVolumesEncrypted : Encrypted
OsVolumeEncryptionSettings : Microsoft.Azure.Management.Compute.Models.DiskEncryptionSettings
ProgressMessage : OsVolume: Encrypted, DataVolumes: Encrypted
Next steps
For more information about managing Azure Key Vault, see Set up Key Vault for virtual machines.
For more information about disk encryption, such as preparing an encrypted custom VM to upload to Azure,
see Azure Disk Encryption.
Setting up WinRM access for Virtual Machines in
Azure Resource Manager
WinRM in Azure Service Management vs Azure Resource Manager

NOTE
article covers using the Resource Manager deployment model, which Microsoft recommends for new deployments instead of
the classic deployment model.
For an overview of the Azure Resource Manager, please see this article
For differences between Azure Service Management and Azure Resource Manager, please see this article
The key difference in setting up WinRM configuration between the two stacks is how the certificate gets installed
on the VM. In the Azure Resource Manager stack, the certificates are modeled as resources managed by the Key
Vault Resource Provider. Therefore, the user needs to provide their own certificate and upload it to a Key Vault
before using it in a VM.
Here are the steps you need to take to set up a VM with WinRM connectivity
1. Create a Key Vault
2. Create a self-signed certificate
3. Upload your self-signed certificate to Key Vault
4. Get the URL for your self-signed certificate in the Key Vault
5. Reference your self-signed certificates URL while creating a VM
Step 1: Create a Key Vault

You can use the below command to create the Key Vault
New-AzureRmKeyVault -VaultName "<vault-name>" -ResourceGroupName "<rg-name>" -Location "<vault-location>" -

EnabledForDeployment -EnabledForTemplateDeployment
Step 2: Create a self-signed certificate

You can create a self-signed certificate using this PowerShell script
$certificateName = "somename"
$thumbprint = (New-SelfSignedCertificate -DnsName $certificateName -CertStoreLocation Cert:\CurrentUser\My -

KeySpec KeyExchange).Thumbprint
$cert = (Get-ChildItem -Path cert:\CurrentUser\My\$thumbprint)
$password = Read-Host -Prompt "Please enter the certificate password." -AsSecureString
Export-PfxCertificate -Cert $cert -FilePath ".\$certificateName.pfx" -Password $password

Step 3: Upload your self-signed certificate to the Key Vault
Before uploading the certificate to the Key Vault created in step 1, it needs to converted into a format the
Microsoft.Compute resource provider will understand. The below PowerShell script will allow you do that
$fileName = "<Path to the .pfx file>"

$fileContentBytes = Get-Content $fileName -Encoding Byte
$fileContentEncoded = [System.Convert]::ToBase64String($fileContentBytes)
$jsonObject = @"
{
"data": "$filecontentencoded",
"dataType" :"pfx",
"password": "<password>"
}
"@
$jsonObjectBytes = [System.Text.Encoding]::UTF8.GetBytes($jsonObject)
$jsonEncoded = [System.Convert]::ToBase64String($jsonObjectBytes)
$secret = ConvertTo-SecureString -String $jsonEncoded -AsPlainText –Force

Set-AzureKeyVaultSecret -VaultName "<vault name>" -Name "<secret name>" -SecretValue $secret
Step 4: Get the URL for your self-signed certificate in the Key Vault
The Microsoft.Compute resource provider needs a URL to the secret inside the Key Vault while provisioning the
VM. This enables the Microsoft.Compute resource provider to download the secret and create the equivalent
certificate on the VM.
NOTE
The URL of the secret needs to include the version as well. An example URL looks like below
https://contosovault.vault.azure.net:443/secrets/contososecret/01h9db0df2cd4300a20ence585a6s7ve
Templates
You can get the link to the URL in the template using the below code
"certificateUrl": "[reference(resourceId(resourceGroup().name, 'Microsoft.KeyVault/vaults/secrets', '<vault-

name>', '<secret-name>'), '2015-06-01').secretUriWithVersion]"
PowerShell
You can get this URL using the below PowerShell command
$secretURL = (Get-AzureKeyVaultSecret -VaultName "<vault name>" -Name "<secret name>").Id
Step 5: Reference your self-signed certificates URL while creating a VM

Azure Resource Manager Templates
While creating a VM through templates, the certificate gets referenced in the secrets section and the winRM
section as below:
"osProfile": {
...
"secrets": [
{
"sourceVault": {
"id": "<resource id of the Key Vault containing the secret>"
},
"vaultCertificates": [
{
"certificateUrl": "<URL for the certificate you got in Step 4>",
"certificateStore": "<Name of the certificate store on the VM>"
}
]
}
],
"windowsConfiguration": {
...
"winRM": {
"listeners": [
{
"protocol": "http"
},
{
"protocol": "https",
"certificateUrl": "<URL for the certificate you got in Step 4>"
}
]
},
...
}
},
A sample template for the above can be found here at 201-vm-winrm-keyvault-windows

Source code for this template can be found on GitHub
PowerShell
$vm = New-AzureRmVMConfig -VMName "<VM name>" -VMSize "<VM Size>"

$credential = Get-Credential
$secretURL = (Get-AzureKeyVaultSecret -VaultName "<vault name>" -Name "<secret name>").Id
$vm = Set-AzureRmVMOperatingSystem -VM $vm -Windows -ComputerName "<Computer Name>" -Credential $credential -
WinRMHttp -WinRMHttps -WinRMCertificateUrl $secretURL
$sourceVaultId = (Get-AzureRmKeyVault -ResourceGroupName "<Resource Group name>" -VaultName "<Vault
Name>").ResourceId
$CertificateStore = "My"
$vm = Add-AzureRmVMSecret -VM $vm -SourceVaultId $sourceVaultId -CertificateStore $CertificateStore -
CertificateUrl $secretURL
Step 6: Connecting to the VM

Before you can connect to the VM you'll need to make sure your machine is configured for WinRM remote
management. Start PowerShell as an administrator and execute the below command to make sure you're set up.
Enable-PSRemoting -Force
NOTE
You might need to make sure the WinRM service is running if the above does not work. You can do that using
Get-Service WinRM
Once the setup is done, you can connect to the VM using the below command
Enter-PSSession -ConnectionUri https://<public-ip-dns-of-the-vm>:5986 -Credential $cred -SessionOption (New-

PSSessionOption -SkipCACheck -SkipCNCheck -SkipRevocationCheck) -Authentication Negotiate
What is role-based access control (RBAC)?
Access management for cloud resources is a critical function for any organization that is using the cloud. Role-
based access control (RBAC ) helps you manage who has access to Azure resources, what they can do with those
resources, and what areas they have access to.
RBAC is an authorization system built on Azure Resource Manager that provides fine-grained access management
of resources in Azure.
What can I do with RBAC?

Here are some examples of what you can do with RBAC:
Allow one user to manage virtual machines in a subscription and another user to manage virtual networks
Allow a DBA group to manage SQL databases in a subscription
Allow a user to manage all resources in a resource group, such as virtual machines, websites, and subnets
Allow an application to access all resources in a resource group
Best practice for using RBAC

Using RBAC, you can segregate duties within your team and grant only the amount of access to users that they
need to perform their jobs. Instead of giving everybody unrestricted permissions in your Azure subscription or
resources, you can allow only certain actions at a particular scope.
When planning your access control strategy, it's a best practice to grant users the least privilege to get their work
done. The following diagram shows a suggested pattern for using RBAC.
How RBAC works

The way you control access to resources using RBAC is to create role assignments. This is a key concept to
understand – it’s how permissions are enforced. A role assignment consists of three elements: security principal,
role definition, and scope.
Security principal
A security principal is an object that represents a user, group, or service principal that is requesting access to Azure
resources.
User - An individual who has a profile in Azure Active Directory. You can also assign roles to users in other
tenants. For information about users in other organizations, see Azure Active Directory B2B.
Group - A set of users created in Azure Active Directory. When you assign a role to a group, all users within
that group have that role.
Service principal - A security identity used by applications or services to access specific Azure resources. You
can think of it as a user identity (username and password or certificate) for an application.
Role definition
A role definition is a collection of permissions. It's sometimes just called a role. A role definition lists the operations
that can be performed, such as read, write, and delete. Roles can be high-level, like owner, or specific, like virtual
machine reader.
Azure includes several built-in roles that you can use. The following lists four fundamental built-in roles. The first
three apply to all resource types.
Owner - Has full access to all resources including the right to delegate access to others.
Contributor - Can create and manage all types of Azure resources but can’t grant access to others.
Reader - Can view existing Azure resources.
User Access Administrator - Lets you manage user access to Azure resources.
The rest of the built-in roles allow management of specific Azure resources. For example, the Virtual Machine
Contributor role allows a user to create and manage virtual machines. If the built-in roles don't meet the specific
needs of your organization, you can create your own custom roles.
Azure has introduced data operations (currently in preview ) that enable you to grant access to data within an
object. For example, if a user has read data access to a storage account, then they can read the blobs or messages
within that storage account. For more information, see Understand role definitions.
Scope
Scope is the boundary that the access applies to. When you assign a role, you can further limit the actions allowed
by defining a scope. This is helpful if you want to make someone a Website Contributor, but only for one resource
group.
In Azure, you can specify a scope at multiple levels: management group, subscription, resource group, or resource.
Scopes are structured in a parent-child relationship.
When you grant access at a parent scope, those permissions are inherited to the child scopes. For example:
If you assign the Owner role to a user at the management group scope, that user can manage everything in all
subscriptions in the management group.
If you assign the Reader role to a group at the subscription scope, the members of that group can view every
resource group and resource in the subscription.
If you assign the Contributor role to an application at the resource group scope, it can manage resources of all
types in that resource group, but not other resource groups in the subscription.
Role assignments
A role assignment is the process of binding a role definition to a user, group, or service principal at a particular
scope for the purpose of granting access. Access is granted by creating a role assignment, and access is revoked by
removing a role assignment.
The following diagram shows an example of a role assignment. In this example, the Marketing group has been
assigned the Contributor role for the pharma-sales resource group. This means that users in the Marketing group
can create or manage any Azure resource in the pharma-sales resource group. Marketing users do not have access
to resources outside the pharma-sales resource group, unless they are part of another role assignment.
You can create role assignments using the Azure portal, Azure CLI, Azure PowerShell, Azure SDKs, or REST APIs.
You can have up to 2000 role assignments in each subscription. To create and remove role assignments, you must
have Microsoft.Authorization/roleAssignments/* permission. This permission is granted through the Owner or
User Access Administrator roles.
Deny assignments
Previously, RBAC was an allow -only model with no deny, but now RBAC supports deny assignments in a limited
way. Similar to a role assignment, a deny assignment binds a set of deny actions to a user, group, or service
principal at a particular scope for the purpose of denying access. A role assignment defines a set of actions that are
allowed, while a deny assignment defines a set of actions that not allowed. In other words, deny assignments block
users from performing specified actions even if a role assignment grants them access. Deny assignments take
precedence over role assignments.
Currently, deny assignments are read-only and can only be set by Azure. Even though you can't create your own
deny assignments, you can list deny assignments because they could impact your effective permissions. To get
information about a deny assignment, you must have the Microsoft.Authorization/denyAssignments/read
permission, which is included in most built-in roles. For more information, see Understand deny assignments.
How RBAC determines if a user has access to a resource

The following are the high-level steps that RBAC uses to determine if you have access to a resource on the
management plane. This is helpful to understand if you are trying to troubleshoot an access issue.
1. A user (or service principal) acquires a token for Azure Resource Manager.
The token includes the user's group memberships (including transitive group memberships).
2. The user makes a REST API call to Azure Resource Manager with the token attached.
3. Azure Resource Manager retrieves all the role assignments and deny assignments that apply to the resource
upon which the action is being taken.
4. Azure Resource Manager narrows the role assignments that apply to this user or their group and
determines what roles the user has for this resource.
5. Azure Resource Manager determines if the action in the API call is included in the roles the user has for this
resource.
6. If the user doesn’t have a role with the action at the requested scope, access is not granted. Otherwise,
Azure Resource Manager checks if a deny assignment applies.
7. If a deny assignment applies, access is blocked. Otherwise access is granted.
Next steps
Quickstart: Grant access for a user using RBAC and the Azure portal
Manage access using RBAC and the Azure portal
Understand the different roles in Azure
Apply policies to Windows VMs with Azure Resource
Manager
By using policies, an organization can enforce various conventions and rules throughout the enterprise.
Enforcement of the desired behavior can help mitigate risk while contributing to the success of the organization. In
this article, we describe how you can use Azure Resource Manager policies to define the desired behavior for your
organization’s Virtual Machines.
For an introduction to policies, see What is Azure Policy?.
Permitted Virtual Machines

To ensure that virtual machines for your organization are compatible with an application, you can restrict the
permitted operating systems. In the following policy example, you allow only Windows Server 2012 R2 Datacenter
Virtual Machines to be created:
{
"if": {
"allOf": [
{
"field": "type",
"in": [
"Microsoft.Compute/disks",
"Microsoft.Compute/virtualMachines",
"Microsoft.Compute/VirtualMachineScaleSets"
]
},
{
"not": {
"allOf": [
{
"field": "Microsoft.Compute/imagePublisher",
"in": [
"MicrosoftWindowsServer"
]
},
{
"field": "Microsoft.Compute/imageOffer",
"in": [
"WindowsServer"
]
},
{
"field": "Microsoft.Compute/imageSku",
"in": [
"2012-R2-Datacenter"
]
},
{
"field": "Microsoft.Compute/imageVersion",
"in": [
"latest"
]
}
]
}
}
]
},
"then": {
"effect": "deny"
}
}
Use a wild card to modify the preceding policy to allow any Windows Server Datacenter image:
{
"like": "*Datacenter"
}
Use anyOf to modify the preceding policy to allow any Windows Server 2012 R2 Datacenter or higher image:
{
"anyOf": [
{
"like": "2012-R2-Datacenter*"
},
{
"like": "2016-Datacenter*"
}
]
}
For information about policy fields, see Policy aliases.
Managed disks
To require the use of managed disks, use the following policy:
{
"if": {
"anyOf": [
{
"allOf": [
{
"field": "type",
"equals": "Microsoft.Compute/virtualMachines"
},
{
"field": "Microsoft.Compute/virtualMachines/osDisk.uri",
"exists": true
}
]
},
{
"allOf": [
{
"field": "type",
"equals": "Microsoft.Compute/VirtualMachineScaleSets"
},
{
"anyOf": [
{
"field": "Microsoft.Compute/VirtualMachineScaleSets/osDisk.vhdContainers",
"exists": true
},
{
"field": "Microsoft.Compute/VirtualMachineScaleSets/osdisk.imageUrl",
"exists": true
}
]
}
]
}
]
},
"then": {
"effect": "deny"
}
}
Images for Virtual Machines

For security reasons, you can require that only approved custom images are deployed in your environment. You
can specify either the resource group that contains the approved images, or the specific approved images.
The following example requires images from an approved resource group:
{
"if": {
"allOf": [
{
"field": "type",
"in": [
"Microsoft.Compute/virtualMachines",
"Microsoft.Compute/VirtualMachineScaleSets"
]
},
{
"not": {
"field": "Microsoft.Compute/imageId",
"contains": "resourceGroups/CustomImage"
}
}
]
},
"then": {
"effect": "deny"
}
}
The following example specifies the approved image IDs:
{
"field": "Microsoft.Compute/imageId",
"in": ["{imageId1}","{imageId2}"]
}
Virtual Machine extensions

You may want to forbid usage of certain types of extensions. For example, an extension may not be compatible with
certain custom virtual machine images. The following example shows how to block a specific extension. It uses
publisher and type to determine which extension to block.
{
"if": {
"allOf": [
{
"field": "type",
"equals": "Microsoft.Compute/virtualMachines/extensions"
},
{
"field": "Microsoft.Compute/virtualMachines/extensions/publisher",
"equals": "Microsoft.Compute"
},
{
"field": "Microsoft.Compute/virtualMachines/extensions/type",
"equals": "{extension-type}"
}
]
},
"then": {
"effect": "deny"
}
}
Azure Hybrid Use Benefit

When you have an on-premises license, you can save the license fee on your virtual machines. When you don't
have the license, you should forbid the option. The following policy forbids usage of Azure Hybrid Use Benefit
(AHUB ):
{
"if": {
"allOf": [
{
"field": "type",
"in":[ "Microsoft.Compute/virtualMachines","Microsoft.Compute/VirtualMachineScaleSets"]
},
{
"field": "Microsoft.Compute/licenseType",
"exists": true
}
]
},
"then": {
"effect": "deny"
}
}
Next steps
After defining a policy rule (as shown in the preceding examples), you need to create the policy definition and
assign it to a scope. The scope can be a subscription, resource group, or resource. To assign policies, see Use
Azure portal to assign and manage resource policies, Use PowerShell to assign policies, or Use Azure CLI to
assign policies.
For an introduction to resource policies, see What is Azure Policy?.
For guidance on how enterprises can use Resource Manager to effectively manage subscriptions, see Azure
enterprise scaffold - prescriptive subscription governance.
Set up Key Vault for virtual machines in Azure
Resource Manager
NOTE
article covers using the Resource Manager deployment model, which Microsoft recommends for new deployments instead of
the classic deployment model.
In Azure Resource Manager stack, secrets/certificates are modeled as resources that are provided by the resource
provider of Key Vault. To learn more about Key Vault, see What is Azure Key Vault?
NOTE
1. In order for Key Vault to be used with Azure Resource Manager virtual machines, the EnabledForDeployment property
on Key Vault must be set to true. You can do this in various clients.
2. The Key Vault needs to be created in the same subscription and location as the Virtual Machine.
Use PowerShell to set up Key Vault

To create a key vault by using PowerShell, see Get started with Azure Key Vault.
For new key vaults, you can use this PowerShell cmdlet:
New-AzureRmKeyVault -VaultName 'ContosoKeyVault' -ResourceGroupName 'ContosoResourceGroup' -Location 'East

Asia' -EnabledForDeployment
For existing key vaults, you can use this PowerShell cmdlet:
Set-AzureRmKeyVaultAccessPolicy -VaultName 'ContosoKeyVault' -EnabledForDeployment
Use CLI to set up Key Vault

To create a key vault by using the command-line interface (CLI), see Manage Key Vault using CLI.
For CLI, you have to create the key vault before you assign the deployment policy. You can do this by using the
following command:
azure keyvault set-policy ContosoKeyVault –enabled-for-deployment true
Use templates to set up Key Vault

While you use a template, you need to set the enabledForDeployment property to true for the Key Vault resource.
{
"type": "Microsoft.KeyVault/vaults",
"name": "ContosoKeyVault",
"apiVersion": "2015-06-01",
"location": "<location-of-key-vault>",
"properties": {
"enabledForDeployment": "true",
....
....
}
}
For other options that you can configure when you create a key vault by using templates, see Create a key vault.
Overview of the features in Azure Backup
Azure Backup is the Azure-based service you can use to back up (or protect) and restore your data in the Microsoft
cloud. Azure Backup replaces your existing on-premises or off-site backup solution with a cloud-based solution
that is reliable, secure, and cost-competitive. Azure Backup offers multiple components that you download and
deploy on the appropriate computer, server, or in the cloud. The component, or agent, that you deploy depends on
what you want to protect. All Azure Backup components (no matter whether you're protecting data on-premises or
in the cloud) can be used to back up data to a Recovery Services vault in Azure. See the Azure Backup components
table (later in this article) for information about which component to use to protect specific data, applications, or
workloads.
Watch a video overview of Azure Backup
Why use Azure Backup?

Traditional backup solutions have evolved to treat the cloud as an endpoint, or static storage destination, similar to
disks or tape. While this approach is simple, it is limited and doesn't take full advantage of an underlying cloud
platform, which translates to an expensive, inefficient solution. Other solutions are expensive because you end up
paying for the wrong type of storage, or storage that you don't need. Other solutions are often inefficient because
they don't offer you the type or amount of storage you need, or administrative tasks require too much time. In
contrast, Azure Backup delivers these key benefits:
Automatic storage management - Hybrid environments often require heterogeneous storage - some on-
premises and some in the cloud. With Azure Backup, there is no cost for using on-premises storage devices. Azure
Backup automatically allocates and manages backup storage, and it uses a pay-as-you-use model. Pay-as-you-use
means that you only pay for the storage that you consume. For more information, see the Azure pricing article.
Unlimited scaling - Azure Backup uses the underlying power and unlimited scale of the Azure cloud to deliver
high-availability - with no maintenance or monitoring overhead. You can set up alerts to provide information about
events, but you don't need to worry about high-availability for your data in the cloud.
Multiple storage options - An aspect of high-availability is storage replication. Azure Backup offers two types of
replication: locally redundant storage and geo-redundant storage. Choose the backup storage option based on
need:
Locally redundant storage (LRS ) replicates your data three times (it creates three copies of your data) in a
storage scale unit in a datacenter. All copies of the data exist within the same region. LRS is a low -cost
option for protecting your data from local hardware failures.
Geo-redundant storage (GRS ) is the default and recommended replication option. GRS replicates your data
to a secondary region (hundreds of miles away from the primary location of the source data). GRS costs
more than LRS, but GRS provides a higher level of durability for your data, even if there is a regional
outage.
Unlimited data transfer - Azure Backup does not limit the amount of inbound or outbound data you transfer.
Azure Backup also does not charge for the data that is transferred. However, if you use the Azure Import/Export
service to import large amounts of data, there is a cost associated with inbound data. For more information about
this cost, see Offline-backup workflow in Azure Backup. Outbound data refers to data transferred from a Recovery
Services vault during a restore operation.
Data encryption - Data encryption allows for secure transmission and storage of your data in the public cloud.
You store the encryption passphrase locally, and it is never transmitted or stored in Azure. If it is necessary to
restore any of the data, only you have encryption passphrase, or key.
Application-consistent backup - An application-consistent backup means a recovery point has all required data
to restore the backup copy. Azure Backup provides application-consistent backups, which ensure additional fixes
are not required to restore the data. Restoring application-consistent data reduces the restoration time, allowing
you to quickly return to a running state.
Long-term retention - You can use Recovery Services vaults for short-term and long-term data retention. Azure
doesn't limit the length of time data can remain in a Recovery Services vault. You can keep data in a vault for as
long as you like. Azure Backup has a limit of 9999 recovery points per protected instance. See the Backup and
retention section in this article for an explanation of how this limit may impact your backup needs.
Which Azure Backup components should I use?

Use the following table for information about what you can protect with each Azure Backup component.
WHERE ARE BACKUPS

COMPONENT BENEFITS LIMITS WHAT IS PROTECTED? STORED?
Azure Backup (MARS) Back up files and Backup 3x per day Files, Recovery Services
agent folders on physical or Not application Folders, vault
virtual Windows OS aware; file, folder, and System State
(VMs can be on- volume-level restore
premises or in Azure) only,
No separate No support for
backup server Linux.
required.
System Center DPM Application-aware Cannot back up Files, Recovery Services

snapshots (VSS) Oracle workload. Folders, vault,
Full flexibility for Volumes, Locally attached
when to take backups VMs, disk,
Recovery Applications, Tape (on-premises
granularity (all) Workloads only)
Can use Recovery System State
Services vault
Linux support on
Hyper-V and VMware
VMs
Back up and
restore VMware VMs
using DPM 2012 R2
Azure Backup Server Application-aware Cannot back up Files, Recovery Services

snapshots (VSS) Oracle workload. Folders, vault,
Full flexibility for Always requires live Volumes, Locally attached
when to take backups Azure subscription VMs, disk
Recovery No support for Applications,
granularity (all) tape backup Workloads,
Can use Recovery System State
Services vault
Linux support on
Hyper-V and VMware
VMs
Back up and
restore VMware VMs
Does not require a
System Center license
WHERE ARE BACKUPS
COMPONENT BENEFITS LIMITS WHAT IS PROTECTED? STORED?
Azure IaaS VM Application-aware Back up VMs once- VMs, Recovery Services

Backup snapshots (VSS) a-day All disks (using vault
Native backups for Restore VMs only PowerShell)
Windows/Linux at disk level
No specific agent Cannot back up
installation required on-premises
Fabric-level backup
with no backup
infrastructure needed
What are the deployment scenarios for each component?

CAN BE DEPLOYED ON-
COMPONENT CAN BE DEPLOYED IN AZURE? PREMISES? TARGET STORAGE SUPPORTED
Azure Backup (MARS) agent Yes Yes Recovery Services vault

The Azure Backup agent The Backup agent can be
can be deployed on any deployed on any
Windows Server VM Windows Server VM or
that runs in Azure. physical machine.
System Center DPM Yes Yes Locally attached disk,

Learn more about how Learn more about how Recovery Services vault,
to protect workloads in to protect workloads
Azure by using System and VMs in your tape (on-premises only)
Center DPM. datacenter.
Azure Backup Server Yes Yes Locally attached disk,

Learn more about how Learn more about how Recovery Services vault
to protect workloads in to protect workloads in
Azure by using Azure Azure by using Azure
Backup Server. Backup Server.
Azure IaaS VM Backup Yes No Recovery Services vault

Part of Azure fabric Use System Center DPM
to back up virtual
Specialized for backup of machines in your
Azure infrastructure as a datacenter.
service (IaaS) virtual
machines.
Which applications and workloads can be backed up?

The following table provides a matrix of the data and workloads that can be protected using Azure Backup. The
Azure Backup solution column has links to the deployment documentation for that solution.
DATA OR WORKLOAD SOURCE ENVIRONMENT AZURE BACKUP SOLUTION

DATA OR WORKLOAD SOURCE ENVIRONMENT AZURE BACKUP SOLUTION
Files and folders Windows Server Azure Backup agent,

System Center DPM (+ the Azure
Backup agent),
Azure Backup Server (includes the
Azure Backup agent)
Files and folders Windows computer Azure Backup agent,

System Center DPM (+ the Azure
Backup agent),
Azure Backup agent)
Hyper-V virtual machine (Windows) Windows Server System Center DPM (+ the Azure
Backup agent),
Azure Backup agent)
Hyper-V virtual machine (Linux) Windows Server System Center DPM (+ the Azure
Backup agent),
Azure Backup agent)
VMware virtual machine Windows Server System Center DPM (+ the Azure
Backup agent),
Azure Backup agent)
Microsoft SQL Server Windows Server System Center DPM (+ the Azure
Backup agent),
Azure Backup agent)
Microsoft SharePoint Windows Server System Center DPM (+ the Azure

Backup agent),
Azure Backup agent)
Microsoft Exchange Windows Server System Center DPM (+ the Azure

Backup agent),
Azure Backup agent)
Azure IaaS VMs (Windows) running in Azure Azure Backup (VM extension)
Azure IaaS VMs (Linux) running in Azure Azure Backup (VM extension)
Linux support
The following table shows the Azure Backup components that have support for Linux.
COMPONENT LINUX (AZURE ENDORSED) SUPPORT
Azure Backup (MARS) agent No (Only Windows based agent)
System Center DPM File-consistent backup of Linux Guest VMs on Hyper-V and
VMWare
VM restore of Hyper-V and VMWare Linux Guest VMs
File-consistent backup not available for Azure VM
Azure Backup Server File-consistent backup of Linux Guest VMs on Hyper-V and
VMWare
VM restore of Hyper-V and VMWare Linux Guest VMs
File-consistent backup not available for Azure VM
Azure IaaS VM Backup Application-consistent backup using pre-script and post-script

framework
Granular file recovery
Restore all VM disks
VM restore
Using Premium Storage VMs with Azure Backup

Azure Backup protects Premium Storage VMs. Azure Premium Storage is solid-state drive (SSD )-based storage
designed to support I/O -intensive workloads. Premium Storage is attractive for virtual machine (VM ) workloads.
For more information about Premium Storage, see the article, Premium Storage: High-Performance Storage for
Azure Virtual Machine Workloads.
Back up Premium Storage VMs
While backing up Premium Storage VMs, the Backup service creates a temporary staging location, named
"AzureBackup-", in the Premium Storage account. The size of the staging location is equal to the size of the
recovery point snapshot. Be sure the Premium Storage account has adequate free space to accommodate the
temporary staging location. For more information, see the article, premium storage limitations. Once the backup
job finishes, the staging location is deleted. The price of storage used for the staging location is consistent with all
Premium storage pricing.
NOTE
Do not modify or edit the staging location.
Restore Premium Storage VMs

You can restore Premium Storage VMs to either Premium Storage or to Standard Storage. Restoring a Premium
Storage VM recovery point back to Premium Storage is the typical process. However, it can be cost effective to
restore a Premium Storage VM recovery point to Standard Storage if you need a subset of files from the VM.
Using managed disk VMs with Azure Backup

Azure Backup protects managed disk VMs. Managed disks free you from managing storage accounts of virtual
machines and greatly simplify VM provisioning.
Back up managed disk VMs
Backing up VMs on managed disks is no different than backing up Resource Manager VMs. In the Azure portal,
you can configure the backup job directly from the Virtual Machine view or from the Recovery Services vault view.
You can back up VMs on managed disks through RestorePoint collections built on top of managed disks. Azure
Backup also supports backing up managed disk VMs encrypted using Azure Disk encryption(ADE ).
Restore managed disk VMs
Azure Backup allows you to restore a complete VM with managed disks, or restore managed disks to a storage
account. Azure manages the managed disks during the restore process. You (the customer) manage the storage
account created as part of the restore process. When restoring managed encrypted VMs, the VM's keys and
secrets should exist in the key vault prior to starting the restore operation.
What are the features of each Backup component?

The following sections provide tables that summarize the availability or support of various features in each Azure
Backup component. See the information following each table for additional support or details.
Storage
AZURE IAAS VM
FEATURE AZURE BACKUP AGENT SYSTEM CENTER DPM AZURE BACKUP SERVER BACKUP
Recovery Services
vault
Disk storage
Tape storage
Compression
(in Recovery Services
vault)
Incremental backup
Disk deduplication
The Recovery Services vault is the preferred storage target across all components. System Center DPM and Azure
Backup Server also provide the option to have a local disk copy. However, only System Center DPM provides the
option to write data to a tape storage device.
Compression
Backups are compressed to reduce the required storage space. The only component that does not use compression
is the VM extension. The VM extension copies all backup data from your storage account to the Recovery Services
vault in the same region. No compression is used when transferring the data. Transferring the data without
compression slightly inflates the storage used. However, storing the data without compression allows for faster
restoration, should you need that recovery point.
Disk Deduplication
You can take advantage of deduplication when you deploy System Center DPM or Azure Backup Server on a
Hyper-V virtual machine. Windows Server performs data deduplication (at the host level) on virtual hard disks
(VHDs) that are attached to the virtual machine as backup storage.
NOTE
Deduplication is not available in Azure for any Backup component. When System Center DPM and Backup Server are
deployed in Azure, the storage disks attached to the VM cannot be deduplicated.
Incremental backup explained

Every Azure Backup component supports incremental backup regardless of the target storage (disk, tape, Recovery
Services vault). Incremental backup ensures that backups are storage and time efficient, by transferring only those
changes made since the last backup.
Comparing Full, Differential and Incremental backup
Storage consumption, recovery time objective (RTO ), and network consumption varies for each type of backup
method. To keep the backup total cost of ownership (TCO ) down, you need to understand how to choose the best
backup solution. The following image compares Full Backup, Differential Backup, and Incremental Backup. In the
image, data source A is composed of 10 storage blocks A1-A10, which are backed up monthly. Blocks A2, A3, A4,
and A9 change in the first month, and block A5 changes in the next month.
With Full Backup, each backup copy contains the entire data source. Full backup consumes a large amount of
network bandwidth and storage, each time a backup copy is transferred.
Differential backup stores only the blocks that changed since the initial full backup, which results in a smaller
amount of network and storage consumption. Differential backups don't retain redundant copies of unchanged
data. However, because the data blocks that remain unchanged between subsequent backups are transferred and
stored, differential backups are inefficient. In the second month, changed blocks A2, A3, A4, and A9 are backed up.
In the third month, these same blocks are backed up again, along with changed block A5. The changed blocks
continue to be backed up until the next full backup happens.
Incremental Backup achieves high storage and network efficiency by storing only the blocks of data that changed
since the previous backup. With incremental backup, there is no need to take regular full backups. In the example,
after taking the full backup in the first month, blocks A2, A3, A4, and A9 are marked as changed, and transferred to
the second month. In the third month, only changed block A5 is marked and transferred. Moving less data saves
storage and network resources, which decreases TCO.
Security
AZURE IAAS VM
BACKUP
FEATURE AZURE BACKUP AGENT SYSTEM CENTER DPM AZURE BACKUP SERVER
Network security
(to Azure)
Data security
(in Azure)
Network security
All backup traffic from your servers to the Recovery Services vault is encrypted using Advanced Encryption
Standard 256. The backup data is sent over a secure HTTPS link. The backup data is also stored in the Recovery
Services vault in encrypted form. Only you, the Azure customer, have the passphrase to unlock this data. Microsoft
cannot decrypt the backup data at any point.
WARNING
Once you establish the Recovery Services vault, only you have access to the encryption key. Microsoft never maintains a copy
of your encryption key, and does not have access to the key. If the key is misplaced, Microsoft cannot recover the backup
data.
Data security
Backing up Azure VMs requires setting up encryption within the virtual machine. Azure Backup supports Azure
Disk Encryption, which uses BitLocker on Windows virtual machines and dm -crypt on Linux virtual machines. On
the back end, Azure Backup uses Azure Storage Service encryption, which protects data at rest.
Network
AZURE IAAS VM
FEATURE AZURE BACKUP AGENT SYSTEM CENTER DPM AZURE BACKUP SERVER BACKUP
Network compression
(to backup server)
Network compression
(to Recovery
Services vault)
Network protocol TCP TCP

(to backup server)
Network protocol HTTPS HTTPS HTTPS HTTPS

(to Recovery
Services vault)
The VM extension (on the IaaS VM ) reads the data directly from the Azure storage account over the storage
network, so it is not necessary to compress this traffic.
If you use a System Center DPM server or Azure Backup Server as a secondary backup server, compress the data
going from the primary server to the backup server. Compressing data before back up to DPM or Azure Backup
Server, saves bandwidth.
Network Throttling
The Azure Backup agent offers network throttling, which allows you to control how network bandwidth is used
during data transfer. Throttling can be helpful if you need to back up data during work hours but do not want the
backup process to interfere with other internet traffic. Throttling for data transfer applies to back up and restore
activities.
Backup and retention

Azure Backup has a limit of 9999 recovery points, also known as backup copies or snapshots, per protected
instance. A protected instance is a computer, server (physical or virtual), or workload configured to back up data to
Azure. For more information, see the section, What is a protected instance. An instance is protected once a backup
copy of data has been saved. The backup copy of data is the protection. If the source data was lost or became
corrupt, the backup copy could restore the source data. The following table shows the maximum backup frequency
for each component. Your backup policy configuration determines how quickly you consume the recovery points.
For example, if you create a recovery point each day, then you can retain recovery points for 27 years before you
run out. If you take a monthly recovery point, you can retain recovery points for 833 years before you run out. The
Backup service does not set an expiration time limit on a recovery point.
AZURE IAAS VM
AZURE BACKUP AGENT SYSTEM CENTER DPM AZURE BACKUP SERVER BACKUP
Backup frequency Three backups per Two backups per day Two backups per day One backup per day
(to Recovery Services day
vault)
Backup frequency Not applicable Every 15 minutes Every 15 minutes Not applicable
(to disk) for SQL Server for SQL Server
Every hour for Every hour for
other workloads other workloads
Retention options Daily, weekly, Daily, weekly, Daily, weekly, monthly, Daily, weekly, monthly,
monthly, yearly monthly, yearly yearly yearly
Maximum recovery 9999 9999 9999 9999

points per protected
instance
Maximum retention Depends on backup Depends on backup Depends on backup Depends on backup
period frequency frequency frequency frequency
Recovery points on Not applicable 64 for File Servers, 64 for File Servers, Not applicable
local disk 448 for Application 448 for Application
Servers Servers
Recovery points on Not applicable Unlimited Not applicable Not applicable

tape
What is a protected instance

A protected instance is a generic reference to a Windows computer, a server (physical or virtual), or SQL database
that has been configured to back up to Azure. An instance is protected once you configure a backup policy for the
computer, server, or database, and create a backup copy of the data. Subsequent copies of the backup data for that
protected instance (which are called recovery points), increase the amount of storage consumed. You can create up
to 9999 recovery points for a protected instance. If you delete a recovery point from storage, it does not count
against the 9999 recovery point total. Some common examples of protected instances are virtual machines,
application servers, databases, and personal computers running the Windows operating system. For example:
A virtual machine running the Hyper-V or Azure IaaS hypervisor fabric. The guest operating systems for the
virtual machine can be Windows Server or Linux.
An application server: The application server can be a physical or virtual machine running Windows Server and
workloads with data that needs to be backed up. Common workloads are Microsoft SQL Server, Microsoft
Exchange server, Microsoft SharePoint server, and the File Server role on Windows Server. To back up these
workloads you need System Center Data Protection Manager (DPM ) or Azure Backup Server.
A personal computer, workstation, or laptop running the Windows operating system.
What is a Recovery Services vault?

A Recovery Services vault is an online storage entity in Azure used to hold data such as backup copies, recovery
points, and backup policies. You can use Recovery Services vaults to hold backup data for Azure services and on-
premises servers and workstations. Recovery Services vaults make it easy to organize your backup data, while
minimizing management overhead. Within each Azure subscription, you can create up to 500 Recovery Services
vaults per Azure region. When considering where to store your data, not all regions are the same. See Geo-
redundant storage for information about region pairings and additional storage considerations.
Backup vaults, which were based on Azure Service Manager, were the first version of the vault. Recovery Services
vaults, which add the Azure Resource Manager model features, are the second version of the vault. See the
Recovery Services vault overview article for a full description of the feature differences. You can no longer create
Backup vaults, and all existing Backup vaults have been upgraded to Recovery Services vaults. You can use the
Azure portal to manage the vaults that were upgraded to Recovery Services vaults.
How does Azure Backup differ from Azure Site Recovery?

Azure Backup and Azure Site Recovery are related in that both services back up data and can restore that data.
However, these services serve different purposes in providing business continuity and disaster recovery in your
business. Use Azure Backup to protect and restore data at a more granular level. For example, if a presentation on a
laptop became corrupted, you would use Azure Backup to restore the presentation. If you wanted to replicate the
configuration and data on a VM across another datacenter, use Azure Site Recovery.
Azure Backup protects data on-premises and in the cloud. Azure Site Recovery coordinates virtual-machine and
physical-server replication, failover, and failback. Both services are important because your disaster recovery
solution needs to keep your data safe and recoverable (Backup) and keep your workloads available (Site Recovery)
when outages occur.
The following concepts can help you make important decisions around backup and disaster recovery.
CONCEPT DETAILS BACKUP DISASTER RECOVERY (DR)
Recovery point objective The amount of acceptable Backup solutions have wide Disaster recovery solutions
(RPO) data loss if a recovery needs variability in their acceptable have low RPOs. The DR copy
to be done. RPO. Virtual machine can be behind by a few
backups usually have an seconds or a few minutes.
RPO of one day, while
database backups have
RPOs as low as 15 minutes.
CONCEPT DETAILS BACKUP DISASTER RECOVERY (DR)
Recovery time objective The amount of time that it Because of the larger RPO, Disaster recovery solutions
(RTO) takes to complete a recovery the amount of data that a have smaller RTOs because
or restore. backup solution needs to they are more in sync with
process is typically much the source. Fewer changes
higher, which leads to longer need to be processed.
RTOs. For example, it can
take days to restore data
from tapes, depending on
the time it takes to transport
the tape from an off-site
location.
Retention How long data needs to be For scenarios that require Disaster recovery needs only
stored operational recovery (data operational recovery data,
corruption, inadvertent file which typically takes a few
deletion, OS failure), backup hours or up to a day.
data is typically retained for Because of the fine-grained
30 days or less. data capture used in DR
From a compliance solutions, using DR data for
standpoint, data might need long-term retention is not
to be stored for months or recommended.
even years. Backup data is
ideally suited for archiving in
such cases.
Next steps
Use one of the following tutorials for detailed, step-by-step, instructions for protecting data on Windows Server, or
protecting a virtual machine (VM ) in Azure:
Back up Files and Folders
Backup Azure Virtual Machines
For details about protecting other workloads, try one of these articles:
Back up your Windows Server
Back up application workloads
Backup Azure IaaS VMs
Back up a virtual machine in Azure
Azure backups can be created through the Azure portal. This method provides a browser-based user interface to
create and configure Azure backups and all related resources. You can protect your data by taking backups at
regular intervals. Azure Backup creates recovery points that can be stored in geo-redundant recovery vaults. This
article details how to back up a virtual machine (VM ) with the Azure portal.
This quickstart enables backup on an existing Azure VM. If you need to create a VM, you can create a VM with the
Azure portal.
Log in to Azure
Log in to the Azure portal at http://portal.azure.com.
Select a VM to back up
Create a simple scheduled daily backup to a Recovery Services Vault.
1. In the menu on the left, select Virtual machines.
2. From the list, choose a VM to back up. If you used the sample VM quickstart commands, the VM is named
myVM in the myResourceGroup resource group.
3. In the Operations section, choose Backup. The Enable backup window opens.
Enable backup on a VM
A Recovery Services vault is a logical container that stores the backup data for each protected resource, such as
Azure VMs. When the backup job for a protected resource runs, it creates a recovery point inside the Recovery
Services vault. You can then use one of these recovery points to restore data to a given point in time.
1. Select Create new and provide a name for the new vault, such as myRecoveryServicesVault.
2. If not already selected, choose Use existing, then select the resource group of your VM from the drop-
down menu.
By default, the vault is set for Geo-Redundant storage. To further protect your data, this storage redundancy
level ensures that your backup data is replicated to a secondary Azure region that is hundreds of miles away
from the primary region.
You create and use policies to define when a backup job runs and how long the recovery points are stored.
The default protection policy runs a backup job each day and retains recovery points for 30 days. You can
use these default policy values to quickly protect your VM.
3. To accept the default backup policy values, select Enable Backup.
It takes a few moments to create the Recovery Services vault.
Start a backup job

You can start a backup now rather than wait for the default policy to run the job at the scheduled time. This first
backup job creates a full recovery point. Each backup job after this initial backup creates incremental recovery
points. Incremental recovery points are storage and time-efficient, as they only transfer changes made since the
last backup.
1. On the Backup window for your VM, select Backup now.
2. To accept the backup retention policy of 30 days, leave the default Retain Backup Till date. To start the job,
select Backup.
Monitor the backup job

In the Backup window for your VM, the status of the backup and number of completed restore points are shown.
Once the VM backup job is complete, information on the Last backup time, Latest restore point, and Oldest
restore point is shown on the right-hand side of the Overview window.
Clean up deployment
When no longer needed, you can disable protection on the VM, remove the restore points and Recovery Services
vault, then delete the resource group and associated VM resources
If you are going to continue on to a Backup tutorial that explains how to restore data for your VM, skip the steps in
this section and go to Next steps.
1. Select the Backup option for your VM.
2. Select ...More to show additional options, then choose Stop backup.
3. Select Delete Backup Data from the drop-down menu.
4. In the Type the name of the Backup item dialog, enter your VM name, such as myVM. Select Stop
Backup
Once the VM backup has been stopped and recovery points removed, you can delete the resource group. If
you used an existing VM, you may wish to leave the resource group and VM in place.
5. In the menu on the left, select Resource groups.
6. From the list, choose your resource group. If you used the sample VM quickstart commands, the resource
group is named myResourceGroup.
7. Select Delete resource group. To confirm, enter the resource group name, then select Delete.
Next steps
In this quickstart, you created a Recovery Services vault, enabled protection on a VM, and created the initial
recovery point. To learn more about Azure Backup and Recovery Services, continue to the tutorials.
Back up multiple Azure VMs
Back up a virtual machine in Azure with PowerShell
The Azure PowerShell module is used to create and manage Azure resources from the command line or in scripts.
You can protect your data by taking backups at regular intervals. Azure Backup creates recovery points that can be
stored in geo-redundant recovery vaults. This article details how to back up a virtual machine (VM ) with the Azure
PowerShell module. You can also perform these steps with the Azure CLI or Azure portal.
This quickstart enables backup on an existing Azure VM. If you need to create a VM, you can create a VM with
Azure PowerShell.
This quickstart requires the Azure PowerShell module version 4.4 or later. Run Get-Module -ListAvailable AzureRM
to find the version. If you need to install or upgrade, see Install Azure PowerShell module.
Log in to Azure
Log in to your Azure subscription with the Connect-AzureRmAccount command and follow the on-screen directions.
Connect-AzureRmAccount
The first time you use Azure Backup, you must register the Azure Recovery Service provider in your subscription
with Register-AzureRmResourceProvider.
Register-AzureRmResourceProvider -ProviderNamespace "Microsoft.RecoveryServices"
Create a recovery services vaults

A Recovery Services vault is a logical container that stores the backup data for each protected resource, such as
Azure VMs. When the backup job for a protected resource runs, it creates a recovery point inside the Recovery
Services vault. You can then use one of these recovery points to restore data to a given point in time.
Create a Recovery Services vault with New -AzureRmRecoveryServicesVault. Specify the same resource group and
location as the VM you wish to protect. If you used the sample script to create your VM, the resource group is
named myResourceGroup, the VM is named myVM, and the resources are in the WestEurope location.
New-AzureRmRecoveryServicesVault `
-Name "myRecoveryServicesVault" `
-Location "WestEurope"
By default, the vault is set for Geo-Redundant storage. To further protect your data, this storage redundancy level
ensures that your backup data is replicated to a secondary Azure region that is hundreds of miles away from the
primary region.
To use this vault with the remaining steps, set the vault context with Set-AzureRmRecoveryServicesVaultContext
Get-AzureRmRecoveryServicesVault `
-Name "myRecoveryServicesVault" | Set-AzureRmRecoveryServicesVaultContext
Enable backup for an Azure VM
You create and use policies to define when a backup job runs and how long the recovery points are stored. The
default protection policy runs a backup job each day, and retains recovery points for 30 days. You can use these
default policy values to quickly protect your VM. First, set the default policy with Get-
AzureRmRecoveryServicesBackupProtectionPolicy:
$policy = Get-AzureRmRecoveryServicesBackupProtectionPolicy -Name "DefaultPolicy"
To enable backup protection for a VM, use Enable-AzureRmRecoveryServicesBackupProtection. Specify the policy
to use, then the resource group and VM to protect:
Enable-AzureRmRecoveryServicesBackupProtection `
-Name "myVM" `
-Policy $policy
Start a backup job

To start a backup now rather than wait for the default policy to run the job at the scheduled time, use Backup-
AzureRmRecoveryServicesBackupItem. This first backup job creates a full recovery point. Each backup job after
this initial backup creates incremental recovery points. Incremental recovery points are storage and time-efficient,
as they only transfer changes made since the last backup.
In the following set of commands, you specify a container in the Recovery Services vault that holds your backup
data with Get-AzureRmRecoveryServicesBackupContainer. Each VM to back up is treated as an item. To start a
backup job, obtain information on your VM item with Get-AzureRmRecoveryServicesBackupItem.
$backupcontainer = Get-AzureRmRecoveryServicesBackupContainer `
-ContainerType "AzureVM" `
-FriendlyName "myVM"
$item = Get-AzureRmRecoveryServicesBackupItem `
-Container $backupcontainer `
-WorkloadType "AzureVM"
Backup-AzureRmRecoveryServicesBackupItem -Item $item
As this first backup job creates a full recovery point, the process can take up to 20 minutes.
Monitor the backup job

To monitor the status of backup jobs, use Get-AzureRmRecoveryservicesBackupJob:
Get-AzureRmRecoveryservicesBackupJob
The output is similar to the following example, which shows the backup job is InProgress:
WorkloadName Operation Status StartTime EndTime JobID

------------ --------- ------ --------- ------- -----
myvm Backup InProgress 9/18/2017 9:38:02 PM 9f9e8f14
myvm ConfigureBackup Completed 9/18/2017 9:33:18 PM 9/18/2017 9:33:51 PM fe79c739
When the Status of the backup job reports Completed, your VM is protected with Recovery Services and has a full
recovery point stored.
Clean up deployment
When no longer needed, you can disable protection on the VM, remove the restore points and Recovery Services
vault, then delete the resource group and associated VM resources. If you used an existing VM, you can skip the
final Remove-AzureRmResourceGroup cmdlet to leave the resource group and VM in place.
If you are going to continue on to a Backup tutorial that explains how to restore data for your VM, skip the steps in
this section and go to Next steps.
Disable-AzureRmRecoveryServicesBackupProtection -Item $item -RemoveRecoveryPoints

$vault = Get-AzureRmRecoveryServicesVault -Name "myRecoveryServicesVault"
Remove-AzureRmRecoveryServicesVault -Vault $vault
Remove-AzureRmResourceGroup -Name "myResourceGroup"
Next steps
In this quickstart, you created a Recovery Services vault, enabled protection on a VM, and created the initial
recovery point. To learn more about Azure Backup and Recovery Services, continue to the tutorials.
Back up multiple Azure VMs
Use Azure portal to back up multiple virtual
machines
When you back up data in Azure, you store that data in an Azure resource called a Recovery Services vault. The
Recovery Services vault resource is available from the Settings menu of most Azure services. The benefit of having
the Recovery Services vault integrated into the Settings menu of most Azure services makes it very easy to back
up data. However, individually working with each database or virtual machine in your business is tedious. What if
you want to back up the data for all virtual machines in one department, or in one location? It is easy to back up
multiple virtual machines by creating a backup policy and applying that policy to the desired virtual machines. This
tutorial explains how to:
Create a Recovery Services vault
Define a backup policy
Apply the backup policy to protect multiple virtual machines
Trigger an on-demand backup job for the protected virtual machines
Log in to the Azure portal

Log in to the Azure portal.

The Recovery Services vault contains the backup data, and the backup policy applied to the protected virtual
machines. Backing up virtual machines is a local process. You cannot back up a virtual machine from one location
to a Recovery Services vault in another location. So, for each Azure location that has virtual machines to be backed
up, at least one Recovery Services vault must exist in that location.
1. On the left-hand menu, select All services and in the services list, type Recovery Services. As you type, the
list of resources filters. When you see Recovery Services vaults in the list, select it to open the Recovery
Services vaults menu.
2. In the Recovery Services vaults menu, click Add to open the Recovery Services vault menu.
3. In the Recovery Services vault menu,
Type myRecoveryServicesVault in Name.
The current subscription ID appears in Subscription. If you have additional subscriptions, you could
choose another subscription for the new vault.
For Resource group select Use existing and choose myResourceGroup. If myResourceGroup doesn't
exist, select Create new and type myResourceGroup.
From the Location drop-down menu, choose West Europe.
Click Create to create your Recovery Services vault.
A Recovery Services vault must be in the same location as the virtual machines being protected. If you have virtual
machines in multiple regions, create a Recovery Services vault in each region. This tutorial creates a Recovery
Services vault in West Europe because that is where myVM (the virtual machine created with the quickstart) was
created.
It can take several minutes for the Recovery Services vault to be created. Monitor the status notifications in the
upper right-hand area of the portal. Once your vault is created, it appears in the list of Recovery Services vaults.
When you create a Recovery Services vault, by default the vault has geo-redundant storage. To provide data
resiliency, geo-redundant storage replicates the data multiple times across two Azure regions.
Set backup policy to protect VMs

After creating the Recovery Services vault, the next step is to configure the vault for the type of data, and to set the
backup policy. Backup policy is the schedule for how often and when recovery points are taken. Policy also
includes the retention range for the recovery points. For this tutorial let's assume your business is a sports
complex with a hotel, stadium, and restaurants and concessions, and you are protecting the data on the virtual
machines. The following steps create a backup policy for the financial data.
1. From the list of Recovery Services vaults, select myRecoveryServicesVault to open its dashboard.
2. On the vault dashboard menu, click Backup to open the Backup menu.
3. On the Backup Goal menu, in the Where is your workload running drop-down menu, choose Azure.
From the What do you want to backup drop-down, choose Virtual machine, and click Backup.
These actions prepare the Recovery Services vault for interacting with a virtual machine. Recovery Services
vaults have a default policy that creates a restore point each day, and retains the restore points for 30 days.
4. To create a new policy, on the Backup policy menu, from the Choose backup policy drop-down menu,
select Create New.
5. In the Backup policy menu, for Policy Name type Finance. Enter the following changes for the Backup
policy:
For Backup frequency set the timezone for Central Time. Since the sports complex is in Texas, the
owner wants the timing to be local. Leave the backup frequency set to Daily at 3:30AM.
For Retention of daily backup point, set the period to 90 days.
For Retention of weekly backup point, use the Monday restore point and retain it for 52 weeks.
For Retention of monthly backup point, use the restore point from First Sunday of the month, and
retain it for 36 months.
Deselect the Retention of yearly backup point option. The leader of Finance doesn't want to keep
data longer than 36 months.
Click OK to create the backup policy.
After creating the backup policy, associate the policy with the virtual machines.
6. In the Select virtual machines dialog select myVM and click OK to deploy the backup policy to the virtual
machines.
All virtual machines that are in the same location, and are not already associated with a backup policy,
appear. myVMH1 and myVMR1 are selected to be associated with the Finance policy.
When the deployment completes, you receive a notification that deployment successfully completed.
Initial backup
You have enabled backup for the Recovery Services vaults, but an initial backup has not been created. It is a
disaster recovery best practice to trigger the first backup, so that your data is protected.
To run an on-demand backup job:
1. On the vault dashboard, click 3 under Backup Items, to open the Backup Items menu.
The Backup Items menu opens.
2. On the Backup Items menu, click Azure Virtual Machine to open the list of virtual machines associated
with the vault.
The Backup Items list opens.
3. On the Backup Items list, click the ellipses ... to open the Context menu.
4. On the Context menu, select Backup now.
The Backup Now menu opens.

5. On the Backup Now menu, enter the last day to retain the recovery point, and click Backup.
Deployment notifications let you know the backup job has been triggered, and that you can monitor the
progress of the job on the Backup jobs page. Depending on the size of your virtual machine, creating the
initial backup may take a while.
When the initial backup job completes, you can see its status in the Backup job menu. The on-demand
backup job created the initial restore point for myVM. If you want to back up other virtual machines, repeat
these steps for each virtual machine.
Clean up resources
If you plan to continue on to work with subsequent tutorials, do not clean up the resources created in this tutorial.
If you do not plan to continue, use the following steps to delete all resources created by this tutorial in the Azure
portal.
1. On the myRecoveryServicesVault dashboard, click 3 under Backup Items, to open the Backup Items
menu.
2. On the Backup Items menu, click Azure Virtual Machine to open the list of virtual machines associated
with the vault.
The Backup Items list opens.

3. In the Backup Items menu, click the ellipsis to open the Context menu.
4. On the context menu select Stop backup to open Stop Backup menu.
5. In the Stop Backup menu, select the upper drop-down menu and choose Delete Backup Data.
6. In the Type the name of the Backup item dialog, type myVM.
7. Once the backup item is verified (a checkmark appears), Stop backup button is enabled. Click Stop
Backup to stop the policy and delete the restore points.
8. In the myRecoveryServicesVault menu, click Delete.

Once the vault is deleted, you return to the list of Recovery Services vaults.
Next steps
In this tutorial you used the Azure portal to:
Set the vault to protect virtual machines
Create a custom backup and retention policy
Assign the policy to protect multiple virtual machines
Trigger an on-demand back up for virtual machines
Continue to the next tutorial to restore an Azure virtual machine from disk.
Restore VMs using CLI
Understanding Azure virtual machine usage
By analyzing your Azure usage data, powerful consumption insights can be gained – insights that can enable better
cost management and allocation throughout your organization. This document provides a deep dive into your
Azure Compute consumption details. For more details on general Azure usage, navigate to Understanding your
bill.
Download your usage details

To begin, download your usage details. The table below provides the definition and example values of usage for
Virtual Machines deployed via the Azure Resource Manager. This document does not contain detailed information
for VMs deployed via our classic model.
FIELDS MEANING EXAMPLE VALUES
Usage Date The date when the resource was used. “11/23/2017”
Meter ID Identifies the top-level service for which “Virtual Machines”

this usage belongs to.
Meter Sub-Category The billed meter identifier. “2005544f-659d-49c9-9094-

For Compute Hour usage, there 8e0aea1be3a5”
is a meter for each VM Size +
OS (Windows, Non-Windows) +
Region.
For Premium software usage,
there is a meter for each
software type. Most premium
software images have different
meters for each core size. For
more information, visit the
Compute Pricing Page.
Meter Name This is specific for each service in Azure. “Compute Hours”
For compute, it is always “Compute
Hours”.
Meter Region Identifies the location of the datacenter “JA East”

for certain services that are priced
based on datacenter location.
Unit Identifies the unit that the service is “Hours”

charged in. Compute resources are
billed per hour.
Consumed The amount of the resource that has “1”, “0.5”

been consumed for that day. For
Compute, we bill for each minute the
VM ran for a given hour (up to 6
decimals of accuracy).
FIELDS MEANING EXAMPLE VALUES
Resource Location Identifies the datacenter where the “JA East”

resource is running.
Consumed Service The Azure platform service that you "Microsoft.Compute"

used.
Resource Group The resource group in which the "MyRG"

deployed resource is running in. For
more information, see Azure Resource
Manager overview.
Instance ID The identifier for the resource. The "/subscriptions/xxxxxxxx-xxxx-xxxx-xxxx-

identifier contains the name you specify xxxxxxxxxxxx/
for the resource when it was created. resourceGroups/MyRG/providers/Micro
For VMs, the Instance ID will contain soft.Compute/virtualMachines/MyVM1”
the SubscriptionId,
ResourceGroupName, and VMName (or or
scale set name for scale set usage).
"/subscriptions/xxxxxxxx-xxxx-xxxx-xxxx-
xxxxxxxxxxxx/
resourceGroups/MyRG/providers/Micro
soft.Compute/virtualMachineScaleSets/
MyVMSS1”
Tags Tag you assign to the resource. Use "

tags to group billing records. Learn how {"myDepartment":"RD","myUser":"myNa
to tag your Virtual Machines. This is me"}"
available for Resource Manager VMs
only.
Additional Info Service-specific metadata. For VMs, we Virtual Machines

populate the following in the additional {"ImageType":"Canonical","ServiceType":"
info field: Standard_DS1_v2","VMName":"",
Image Type- specific image that "UsageType":"ComputeHR"}
you ran. Find the full list of
supported strings below under Virtual Machine Scale Sets
Image Types. {"ImageType":"Canonical","ServiceType":"
Service Type: the size that you Standard_DS1_v2","VMName":"myVM1"
deployed. , "UsageType":"ComputeHR"}
VMName: name of your VM.
This is only populated for scale Premium Software
set VMs. If you need your VM {"ImageType":"","ServiceType":"Standard
Name for scale set VMs, you can _DS1_v2","VMName":"",
find that in the Instance ID "UsageType":"ComputeHR_SW"}
string above.
UsageType: This specifies the
type of usage this represents.
ComputeHR is the
Compute Hour usage for
the underlying VM, like
Standard_D1_v2.
ComputeHR_SW is the
premium software charge
if the VM is using
premium software, like
Microsoft R Server.
Image Type
For some images in the Azure gallery, the image type is populated in the Additional Info field. This enables users to
understand and track what they have deployed on their Virtual Machine. The values that are populated in this field
based on the image you have deployed are the following:
BitRock
Canonical
FreeBSD
Open Logic
Oracle
SLES for SAP
SQL Server 14 Preview on Windows Server 2012 R2 Preview
SUSE
SUSE Premium
StorSimple Cloud Appliance
Red Hat
Red Hat for SAP Business Applications
Red Hat for SAP HANA
Windows Client BYOL
Windows Server BYOL
Windows Server Preview
Service Type
The service type field in the Additional Info field corresponds to the exact VM size you deployed. Premium storage
VMs (SSD -based) and non-premium storage VMs (HDD -based) are priced the same. If you deploy an SSD -based
size, like Standard_DS2_v2, you see the non-SSD size (‘Standard_D2_v2 VM’) in the Meter Sub-Category column
and the SSD -size (‘Standard_DS2_v2’) in the Additional Info field.
Region Names
The region name populated in the Resource Location field in the usage details varies from the region name used in
the Azure Resource Manager. Here is a mapping between the region values:
RESOURCE MANAGER REGION NAME RESOURCE LOCATION IN USAGE DETAILS
australiaeast AU East
australiasoutheast AU Southeast
brazilsouth BR South
CanadaCentral CA Central
CanadaEast CA East
CentralIndia IN Central
centralus Central US
chinaeast China East
chinanorth China North
eastasia East Asia
eastus East US
eastus2 East US 2
GermanyCentral DE Central
GermanyNortheast DE Northeast
japaneast JA East
japanwest JA West
KoreaCentral KR Central
KoreaSouth KR South
northcentralus North Central US
northeurope North Europe
southcentralus South Central US
southeastasia Southeast Asia
SouthIndia IN South
UKNorth US North
uksouth UK South
UKSouth2 UK South 2
ukwest UK West
USDoDCentral US DoD Central
USDoDEast US DoD East
USGovArizona USGov Arizona
usgoviowa USGov Iowa
USGovTexas USGov Texas

usgovvirginia USGov Virginia
westcentralus US West Central
westeurope West Europe
WestIndia IN West
westus West US
westus2 US West 2
Virtual machine usage FAQ

What resources are charged when deploying a VM?
VMs acquire costs for the VM itself, any premium software running on the VM, the storage account\managed disk
associated with the VM, and the networking bandwidth transfers from the VM.
How can I tell if a VM is using Azure Hybrid Benefit in the Usage CSV?
If you deploy using the Azure Hybrid Benefit, you are charged the Non-Windows VM rate since you are bringing
your own license to the cloud. In your bill, you can distinguish which Resource Manager VMs are running Azure
Hybrid Benefit because they have either “Windows_Server BYOL” or “Windows_Client BYOL” in the ImageType
column.
How are Basic vs. Standard VM Types differentiated in the Usage CSV?
Both Basic and Standard A-Series VMs are offered. If you deploy a Basic VM, in the Meter Sub Category, it has the
string “Basic.” If you deploy a Standard A-Series VM, then the VM size appears as “A1 VM” since Standard is the
default. To learn more about the differences between Basic and Standard, see the Pricing Page.
What are ExtraSmall, Small, Medium, Large, and ExtraLarge sizes?
ExtraSmall - ExtraLarge are the legacy names for Standard_A0 – Standard_A4. In classic VM usage records, you
might see this convention used if you have deployed these sizes.
What is the difference between Meter Region and Resource Location?
The Meter Region is associated with the meter. For some Azure services who use one price for all regions, the
Meter Region field could be blank. However, since VMs have dedicated prices per region for Virtual Machines, this
field is populated. Similarly, the Resource Location for Virtual Machines is the location where the VM is deployed.
The Azure region in both fields are the same, although they might have a different string convention for the region
name.
Why is the ImageType value blank in the Additional Info field?
The ImageType field is only populated for a subset of images. If you did not deploy one of the images above, the
ImageType is blank.
Why is the VMName blank in the Additional Info?
The VMName is only populated in the Additional Info field for VMs in a scale set. The InstanceID field contains the
VM name for non-scale set VMs.
What does ComputeHR mean in the UsageType field in the Additional Info?
ComputeHR stands for Compute Hour which represents the usage event for the underlying infrasturcture cost. If
the UsageType is ComputeHR_SW, the usage event represents the premium software charge for the VM.
How do I know if I am charged for premium software?
When exploring which VM Image best fits your needs, be sure to check out the Azure Marketplace. The image has
the software plan rate. If you see “Free” for the rate, there is no additional cost for the software.
What is the difference between Microsoft.ClassicCompute and Microsoft.Compute in the Consumed service?
Microsoft.ClassicCompute represents classic resources deployed via the Azure Service Manager. If you deploy via
the Resource Manager, then Microsoft.Compute is populated in the consumed service. Learn more about the
Azure Deployment models.
Why is the InstanceID field blank for my Virtual Machine usage?
If you deploy via the classic deployment model, the InstanceID string is not available.
Why are the tags for my VMs not flowing to the usage details?
Tags only flow to you the Usage CSV for Resource Manager VMs only. Classic resource tags are not available in
the usage details.
How can the consumed quantity be more than 24 hours one day?
In the Classic model, billing for resources is aggregated at the Cloud Service level. If you have more than one VM
in a Cloud Service that uses the same billing meter, your usage is aggregated together. VMs deployed via Resource
Manager are billed at the VM level, so this aggregation will not apply.
Why is pricing not available for DS/FS/GS/LS sizes on the pricing page?
Premium storage capable VMs are billed at the same rate as non-premium storage capable VMs. Only your
storage costs differ. Visit the storage pricing page for more information.
Next steps
To learn more about your usage details, see Understand your bill for Microsoft Azure.
Common PowerShell commands for creating and
managing Azure Virtual Machines
This article covers some of the Azure PowerShell commands that you can use to create and manage virtual
machines in your Azure subscription. For more detailed help with specific command-line switches and options, you
can use the Get-Help command.
See How to install and configure Azure PowerShell for information about installing the latest version of Azure
PowerShell, selecting your subscription, and signing in to your account.
These variables might be useful for you if running more than one of the commands in this article:
$location - The location of the virtual machine. You can use Get-AzureRmLocation to find a geographical region
that works for you.
$myResourceGroup - The name of the resource group that contains the virtual machine.
$myVM - The name of the virtual machine.
Create a VM - simplified
TASK COMMAND
Create a simple VM New-AzureRmVM -Name $myVM
New-AzureRMVM has a set of simplified parameters, where

all that is required is a single name. The value for -Name will
be used as the name for all of the resources required for
creating a new VM. You can specify more, but this is all that is
required.
Create a VM from a custom image New-AzureRmVm -ResourceGroupName $myResourceGroup

-Name $myVM ImageName "myImage" -Location $location
You need to have already created your own managed image.

You can use an image to make multiple, identical VMs.
Create a VM configuration
TASK COMMAND
Create a VM configuration $vm = New-AzureRmVMConfig -VMName $myVM -VMSize

"Standard_D1_v1"
The VM configuration is used to define or update settings for

the VM. The configuration is initialized with the name of the
VM and its size.
TASK COMMAND
Add configuration settings $vm = Set-AzureRmVMOperatingSystem -VM $vm -Windows

-ComputerName $myVM -Credential $cred -
ProvisionVMAgent -EnableAutoUpdate
Operating system settings including credentials are added to

the configuration object that you previously created using
New-AzureRmVMConfig.
Add a network interface $vm = Add-AzureRmVMNetworkInterface -VM $vm -Id

$nic.Id
A VM must have a network interface to communicate in a

virtual network. You can also use Get-
AzureRmNetworkInterface to retrieve an existing network
interface object.
Specify a platform image $vm = Set-AzureRmVMSourceImage -VM $vm -

PublisherName "publisher_name" -Offer "publisher_offer" -
Skus "product_sku" -Version "latest"
Image information is added to the configuration object that

you previously created using New-AzureRmVMConfig. The
object returned from this command is only used when you set
the OS disk to use a platform image.
Create a VM New-AzureRmVM -ResourceGroupName $myResourceGroup

-Location $location -VM $vm
All resources are created in a resource group. Before you run

this command, run New-AzureRmVMConfig, Set-
AzureRmVMOperatingSystem, Set-AzureRmVMSourceImage,
Add-AzureRmVMNetworkInterface, and Set-
AzureRmVMOSDisk.
Update a VM Update-AzureRmVM -ResourceGroupName

$myResourceGroup -VM $vm
Get the current VM configuration using Get-AzureRmVM,

change configuration settings on the VM object, and then run
this command.
Get information about VMs

TASK COMMAND
List VMs in a subscription Get-AzureRmVM
List VMs in a resource group Get-AzureRmVM -ResourceGroupName $myResourceGroup
To get a list of resource groups in your subscription, use Get-

AzureRmResourceGroup.
Get information about a VM Get-AzureRmVM -ResourceGroupName $myResourceGroup -

Name $myVM
Manage VMs
TASK COMMAND
Start a VM Start-AzureRmVM -ResourceGroupName $myResourceGroup

-Name $myVM
Stop a VM Stop-AzureRmVM -ResourceGroupName $myResourceGroup

-Name $myVM
Restart a running VM Restart-AzureRmVM -ResourceGroupName

$myResourceGroup -Name $myVM
Delete a VM Remove-AzureRmVM -ResourceGroupName

$myResourceGroup -Name $myVM
Next steps
See the basic steps for creating a virtual machine in Create a Windows VM using Resource Manager and
PowerShell.
Move a Windows VM to another Azure subscription
or resource group
This article walks you through how to move a Windows virtual machine (VM ) between resource groups or
subscriptions. Moving between subscriptions can be handy if you originally created a VM in a personal
subscription and now want to move it to your company's subscription to continue your work.
IMPORTANT
You cannot move Azure Managed Disks at this time.
New resource IDs are created as part of the move. After the VM has been moved, you will need to update your tools and
scripts to use the new resource IDs.
Use the Azure portal to move a VM to a different subscription

You can move a VM and its associated resources to a different subscription by using the Azure portal.
1. Open the Azure portal.
2. Click Browse > Resource groups and select the resource group containing the VM that you would like to
move.
3. At the top of the page for the resource group, select Move and then select Move to another subscription.
The Move resources page opens.
4. Select each of the resources to move. In most cases, you should move all of the related resources that are listed.
5. Select the Subscription where you want the VM to be moved.
6. Select an existing Resource group, or enter a name to have a new resource group created.
7. When you are done, select that you understand that new resource IDs will be created and that the new IDs will
need to be used with the VM after it is moved, and then select OK.
Use the Azure portal to move a VM to another resource group

You can move a VM and its associated resources to another resource group by using the Azure portal.
2. Click Browse > Virtual machines and select the VM you would like to move from the list.
3. In the page for the VM, next to the label for resource group, select Change. The Move resources page opens.
4. Select each of the resources to move. In most cases, you should move all of the related resources that are listed.
5. Select an existing Resource group, or enter a name to have a new resource group created.
6. When you are done, select that you understand that new resource IDs will be created and that the new IDs will
need to be used with the VM after it is moved, and then select OK.
Use Powershell to move a VM

To move a virtual machine to another resource group, you need to make sure that you also move all of the
dependent resources. To get a list with the resource ID of each of these resources, use the Get-AzureRMResource
cmdlet.
Get-AzureRMResource -ResourceGroupName <sourceResourceGroupName> | Format-table -Property ResourceId
You can use the output of the previous command as a comma-separated list of resource IDs to Move-
AzureRMResource to move each resource to the destination.
Move-AzureRmResource -DestinationResourceGroupName "<myDestinationResourceGroup>" `

-ResourceId <myResourceId,myResourceId,myResourceId>
To move the resources to different subscription, include the -DestinationSubscriptionId parameter.
Move-AzureRmResource -DestinationSubscriptionId "<myDestinationSubscriptionID>" `

-DestinationResourceGroupName "<myDestinationResourceGroup>" `
-ResourceId <myResourceId,myResourceId,myResourceId>
When you are asked to confirm that you want to move the specified resources, enter Y to confirm.
Next steps
You can move many different types of resources between resource groups and subscriptions. For more
information, see Move resources to a new resource group or subscription.
Resize a Windows VM
This article shows you how to move a VM to a different VM size using Azure Powershell.
After you create a virtual machine (VM ), you can scale the VM up or down by changing the VM size. In some cases,
you must deallocate the VM first. This can happen if the new size is not available on the hardware cluster that is
currently hosting the VM.
If your VM uses Premium Storage, make sure that you choose an s version of the size to get Premium Storage
support. For example, choose Standard_E4s_v3 instead of Standard_E4_v3.
Resize a Windows VM not in an availability set

Set some variables. Replace the values with your own information.
$resourceGroup = "myResourceGroup"
$vmName = "myVM"
List the VM sizes that are available on the hardware cluster where the VM is hosted.
Get-AzureRmVMSize -ResourceGroupName $resourceGroup -VMName $vmName
If the size you want is listed, run the following commands to resize the VM. If the desired size is not listed, go on to
step 3.
$vm = Get-AzureRmVM -ResourceGroupName $resourceGroup -VMName $vmName

$vm.HardwareProfile.VmSize = "<newVMsize>"
Update-AzureRmVM -VM $vm -ResourceGroupName $resourceGroup
If the size you want is not listed, run the following commands to deallocate the VM, resize it, and restart the VM.
Replace with the size you want.
Stop-AzureRmVM -ResourceGroupName $resourceGroup -Name $vmName -Force

$vm.HardwareProfile.VmSize = "<newVMSize>"
Start-AzureRmVM -ResourceGroupName $resourceGroup -Name $vmName
WARNING
Deallocating the VM releases any dynamic IP addresses assigned to the VM. The OS and data disks are not affected.
Resize a Windows VM in an availability set

If the new size for a VM in an availability set is not available on the hardware cluster currently hosting the VM, then
all VMs in the availability set will need to be deallocated to resize the VM. You also might need to update the size of
other VMs in the availability set after one VM has been resized. To resize a VM in an availability set, perform the
following steps.
$vmName = "myVM"
List the VM sizes that are available on the hardware cluster where the VM is hosted.
Get-AzureRmVMSize -ResourceGroupName $resourceGroup -VMName $vmName
If the desired size is listed, run the following commands to resize the VM. If it is not listed, go to the next section.

$vm.HardwareProfile.VmSize = "<newVmSize>"
If the size you want is not listed, continue with the following steps to deallocate all VMs in the availability set, resize
VMs, and restart them.
Stop all VMs in the availability set.
$as = Get-AzureRmAvailabilitySet -ResourceGroupName $resourceGroup

$vmIds = $as.VirtualMachinesReferences
foreach ($vmId in $vmIDs){
$string = $vmID.Id.Split("/")
$vmName = $string[8]
Stop-AzureRmVM -ResourceGroupName $resourceGroup -Name $vmName -Force
}
Resize and restart the VMs in the availability set.
$newSize = "<newVmSize>"
$as = Get-AzureRmAvailabilitySet -ResourceGroupName $resourceGroup
$vmIds = $as.VirtualMachinesReferences
foreach ($vmId in $vmIDs){
$string = $vmID.Id.Split("/")
$vmName = $string[8]
$vm = Get-AzureRmVM -ResourceGroupName $resourceGroup -Name $vmName
$vm.HardwareProfile.VmSize = $newSize
Update-AzureRmVM -ResourceGroupName $resourceGroup -VM $vm
Start-AzureRmVM -ResourceGroupName $resourceGroup -Name $vmName
}
Next steps
For additional scalability, run multiple VM instances and scale out. For more information, see Automatically scale
Windows machines in a Virtual Machine Scale Set.
Change the OS disk used by an Azure VM using
PowerShell
If you have an existing VM, but you want to swap the disk for a backup disk or another OS disk, you can use Azure
PowerShell to swap the OS disks. You don't have to delete and recreate the VM. You can even use a managed disk
in another resource group, as long as it isn't already in use.
The VM does need to be stopped\deallocated, then the resource ID of the managed disk can be replaced with the
resource ID of a different managed disk.
Make sure that the VM size and storage type are compatible with the disk you want to attach. For example, if the
disk you want to use is in Premium Storage, then the VM needs to be capable of Premium Storage (like a DS -
series size).
Get a list of disks in a resource group using Get-AzureRmDisk
Get-AzureRmDisk -ResourceGroupName myResourceGroup | Format-Table -Property Name
When you have the name of the disk that you would like to use, set that as the OS disk for the VM. This example
stop\deallocates the VM named myVM and assigns the disk named newDisk as the new OS disk.
# Get the VM
$vm = Get-AzureRmVM -ResourceGroupName myResourceGroup -Name myVM
# Make sure the VM is stopped\deallocated

Stop-AzureRmVM -ResourceGroupName myResourceGroup -Name $vm.Name -Force
# Get the new disk that you want to swap in

$disk = Get-AzureRmDisk -ResourceGroupName myResourceGroup -Name newDisk
# Set the VM configuration to point to the new disk

Set-AzureRmVMOSDisk -VM $vm -ManagedDiskId $disk.Id -Name $disk.Name
# Update the VM with the new OS disk

Update-AzureRmVM -ResourceGroupName myResourceGroup -VM $vm
# Start the VM
Start-AzureRmVM -Name $vm.Name -ResourceGroupName myResourceGroup
Next steps
To create a copy of a disk, see Snapshot a disk.
How to tag a Windows virtual machine in Azure
This article describes different ways to tag a Windows virtual machine in Azure through the Resource Manager
deployment model. Tags are user-defined key/value pairs which can be placed directly on a resource or a resource
group. Azure currently supports up to 15 tags per resource and resource group. Tags may be placed on a resource
at the time of creation or added to an existing resource. Please note that tags are supported for resources created
via the Resource Manager deployment model only. If you want to tag a Linux virtual machine, see How to tag a
Linux virtual machine in Azure.
Tagging a Virtual Machine through Templates

First, let’s look at tagging through templates. This template places tags on the following resources: Compute
(Virtual Machine), Storage (Storage Account), and Network (Public IP Address, Virtual Network, and Network
Interface). This template is for a Windows VM but can be adapted for Linux VMs.
Click the Deploy to Azure button from the template link. This will navigate to the Azure portal where you can
deploy this template.
This template includes the following tags: Department, Application, and Created By. You can add/edit these tags
directly in the template if you would like different tag names.
As you can see, the tags are defined as key/value pairs, separated by a colon (:). The tags must be defined in this
format:
“tags”: {
“Key1” : ”Value1”,
“Key2” : “Value2”
}
Save the template file after you finish editing it with the tags of your choice.
Next, in the Edit Parameters section, you can fill out the values for your tags.
Click Create to deploy this template with your tag values.
Tagging through the Portal

After creating your resources with tags, you can view, add, and delete tags in the portal.
Select the tags icon to view your tags:
Add a new tag through the portal by defining your own Key/Value pair, and save it.
Your new tag should now appear in the list of tags for your resource.
Tagging with PowerShell
To create, add, and delete tags through PowerShell, you first need to set up your PowerShell environment with
Azure Resource Manager. Once you have completed the setup, you can place tags on Compute, Network, and
Storage resources at creation or after the resource is created via PowerShell. This article will concentrate on
viewing/editing tags placed on Virtual Machines.
First, navigate to a Virtual Machine through the Get-AzureRmVM cmdlet.
PS C:\> Get-AzureRmVM -ResourceGroupName "MyResourceGroup" -Name "MyTestVM"
If your Virtual Machine already contains tags, you will then see all the tags on your resource:
Tags : {
"Application": "MyApp1",
"Created By": "MyName",
"Department": "MyDepartment",
"Environment": "Production"
}
If you would like to add tags through PowerShell, you can use the Set-AzureRmResource command. Note when
updating tags through PowerShell, tags are updated as a whole. So if you are adding one tag to a resource that
already has tags, you will need to include all the tags that you want to be placed on the resource. Below is an
example of how to add additional tags to a resource through PowerShell Cmdlets.
This first cmdlet sets all of the tags placed on MyTestVM to the $tags variable, using the Get-AzureRmResource and
Tags property.
PS C:\> $tags = (Get-AzureRmResource -ResourceGroupName MyResourceGroup -Name MyTestVM).Tags
The second command displays the tags for the given variable.
PS C:\> $tags
Name Value
---- -----
Value MyDepartment
Name Department
Value MyApp1
Name Application
Value MyName
Name Created By
Value Production
Name Environment
The third command adds an additional tag to the $tags variable. Note the use of the += to append the new
key/value pair to the $tags list.
PS C:\> $tags += @{Name="Location";Value="MyLocation"}
The fourth command sets all of the tags defined in the $tags variable to the given resource. In this case, it is
MyTestVM.
PS C:\> Set-AzureRmResource -ResourceGroupName MyResourceGroup -Name MyTestVM -ResourceType

"Microsoft.Compute/VirtualMachines" -Tag $tags
The fifth command displays all of the tags on the resource. As you can see, Location is now defined as a tag with
MyLocation as the value.
PS C:\> (Get-AzureRmResource -ResourceGroupName MyResourceGroup -Name MyTestVM).Tags
Name Value
---- -----
Value MyDepartment
Name Department
Value MyApp1
Name Application
Value MyName
Name Created By
Value Production
Name Environment
Value MyLocation
Name Location
To learn more about tagging through PowerShell, check out the Azure Resource Cmdlets.
Viewing your tags in the usage details

Tags placed on Compute, Network, and Storage resources in the Resource Manager deployment model will be
populated in your usage details in the billing portal.
Click on Download usage details to view the usage details in your subscription.
Select your billing statement and the Version 2 usage details:
From the usage details, you can see all of the tags in the Tags column:
By analyzing these tags along with usage, organizations will be able to gain new insights into their consumption
data.
Next steps
To learn more about tagging your Azure resources, see Azure Resource Manager Overview and Using Tags to
organize your Azure Resources.
To see how tags can help you manage your use of Azure resources, see Understanding your Azure Bill and Gain
insights into your Microsoft Azure resource consumption.
Time sync for Windows VMs in Azure
Time sync is important for security and event correlation. Sometimes it is used for distributed transactions
implementation. Time accuracy between multiple computer systems is achieved through synchronization.
Synchronization can be affected by multiple things, including reboots and network traffic between the time source
and the computer fetching the time.
Azure is now backed by infrastructure running Windows Server 2016. Windows Server 2016 has improved
algorithms used to correct time and condition the local clock to synchronize with UTC. Windows Server 2016 also
improved the VMICTimeSync service that governs how VMs sync with the host for accurate time. Improvements
include more accurate initial time on VM start or VM restore and interrupt latency correction for samples provided
to Windows Time (W32time).
NOTE
For a quick overview of Windows Time service, take a look at this high-level overview video.
For more information, see Accurate time for Windows Server 2016.
Overview
Accuracy for a computer clock is gauged on how close the computer clock is to the Coordinated Universal Time
(UTC ) time standard. UTC is defined by a multinational sample of precise atomic clocks that can only be off by one
second in 300 years. But, reading UTC directly requires specialized hardware. Instead, time servers are synced to
UTC and are accessed from other computers to provide scalability and robustness. Every computer has time
synchronization service running that knows what time servers to use and periodically checks if computer clock
needs to be corrected and adjusts time if needed.
Azure hosts are synchronized to internal Microsoft time servers that take their time from Microsoft-owned Stratum
1 devices, with GPS antennas. Virtual machines in Azure can either depend on their host to pass the accurate time
(host time) on to the VM or the VM can directly get time from a time server, or a combination of both.
Virtual machine interactions with the host can also affect the clock. During memory preserving maintenance, VMs
are paused for up to 30 seconds. For example, before maintenance begins the VM clock shows 10:00:00 AM and
lasts 28 seconds. After the VM resumes, the clock on the VM would still show 10:00:00 AM, which would be 28
seconds off. To correct for this, the VMICTimeSync service monitors what is happening on the host and prompts
for changes to happen on the VMs to compensate.
Without time synchronization working, the clock on the VM would accumulate errors. When there is only one VM,
the effect might not be significant unless the workload requires highly accurate timekeeping. But in most cases, we
have multiple, interconnected VMs that use time to track transactions and the time needs to be consistent
throughout the entire deployment. When time between VMs is different, you could see the following effects:
Authentication will fail. Security protocols like Kerberos or certificate-dependent technology rely on time being
consistent across the systems.
It's very hard to figure out what have happened in a system if logs (or other data) don't agree on time. The same
event would look like it occurred at different times, making correlation difficult.
If clock is off, the billing could be calculated incorrectly.
The best results for Windows deployments are achieved by using Windows Server 2016 as the guest operating
system, which ensures you can use the latest improvements in time synchronization.
Configuration options
There are three options for configuring time sync for your Windows VMs hosted in Azure:
Host time and time.windows.com. This is the default configuration used in Azure Marketplace images.
Host-only.
Use another, external time server with or without using host time.
Use the default
By default Windows OS VM images are configured for w32time to sync from two sources:
The NtpClient provider, which gets information from time.windows.com.
The VMICTimeSync service, used to communicate the host time to the VMs and make corrections after the VM
is paused for maintenance. Azure hosts use Microsoft-owned Stratum 1 devices to keep accurate time.
w32time would prefer the time provider in the following order of priority: stratum level, root delay, root dispersion,
time offset. In most cases, w32time would prefer time.windows.com to the host because time.windows.com reports
lower stratum.
For domain joined machines the domain itself establishes time sync hierarchy, but the forest root still needs to take
time from somewhere and the following considerations would still hold true.
Host-only
Because time.windows.com is a public NTP server, syncing time with it requires sending traffic over the internet,
varying packet delays can negatively affect quality of the time sync. Removing time.windows.com by switching to
host-only sync can sometimes improve your time sync results.
Switching to host-only time sync makes sense if you experience time sync issues using the default configuration.
Try out the host-only sync to see if that would improve the time sync on VM.
External time server
If you have specific time sync requirements, there is also an option of using external time servers. External time
servers can provide specific time, which can be useful for test scenarios, ensuring time uniformity with machines
hosted in non-Microsoft datacenters, or handling leap seconds in a special way.
You can combine external servers with the VMICTimeSync service and VMICTimeProvider to provide results
similar to the default configuration.
Check your configuration

Check if the NtpClient time provider is configured to use explicit NTP servers (NTP ) or domain time sync (NT5DS ).
w32tm /dumpreg /subkey:Parameters | findstr /i "type"
If the VM is using NTP, you will see the following output:
Value Name Value Type Value Data

Type REG_SZ NTP
To see what time server the NtpClient time provider is using, at an elevated command prompt type:
w32tm /dumpreg /subkey:Parameters | findstr /i "ntpserver"

If the VM is using the default, the output will look like this:
NtpServer REG_SZ time.windows.com,0x8
To see what time provider is being used currently.
w32tm /query /source
Here is the output you could see and what it would mean:
time.windows.com - in the default configuration, w32time would get time from time.windows.com. The time
sync quality depends on internet connectivity to it and is affected by packet delays. This is the usual output from
the default setup.
VM IC Time Synchronization Provider - the VM is syncing time from the host. This usually is the result if
you opt-in for host-only time sync or the NtpServer is not available at the moment.
Your domain server - the current machine is in a domain and the domain defines the time sync hierarchy.
Some other server - w32time was explicitly configured to get the time from that another server. Time sync
quality depends on this time server quality.
Local CMOS Clock - clock is unsynchronized. You can get this output if w32time hasn't had enough time to
start after a reboot or when all the configured time sources are not available.
Opt-in for host-only time sync

Azure is constantly working on improving time sync on hosts and can guarantee that all the time sync
infrastructure is collocated in Microsoft-owned datacenters. If you have time sync issues with the default setup that
prefers to use time.windows.com as the primary time source, you can use the following commands to opt-in to
host-only time sync.
Mark the VMIC provider as enabled.
reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\w32time\TimeProviders\VMICTimeProvider /v Enabled

/t REG_DWORD /d 1 /f
Mark the NTPClient provider as disabled.
reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\w32time\TimeProviders\NtpClient /v Enabled /t

REG_DWORD /d 0 /f
Restart the w32time Service.
net stop w32time && net start w32time
Windows Server 2012 and R2 VMs

Windows Server 2012 and Windows Server 2012 R2 have different default settings for time sync. The w32time by
default is configured in a way that prefers low overhead of the service over to precise time.
If you want to move your Windows Server 2012 and 2012 R2 deployments to use the newer defaults that prefer
precise time, you can apply the following settings.
Update the w32time poll and update intervals to match Windows Server 2016 settings.
reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\w32time\Config /v MinPollInterval /t REG_DWORD /d
6 /f
reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\w32time\Config /v MaxPollInterval /t REG_DWORD /d
10 /f
reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\w32time\Config /v UpdateInterval /t REG_DWORD /d
100 /f
w32tm /config /update
For w32time to be able to use the new poll intervals, the NtpServers be marked as using them. If servers are
annotated with 0x1 bitflag mask, that would override this mechanism and w32time would use SpecialPollInterval
instead. Make sure that specified NTP servers are either using 0x8 flag or no flag at all:
Check what flags are being used for the used NTP servers.
w32tm /dumpreg /subkey:Parameters | findstr /i "ntpserver"
Next steps
Below are links to more details about the time sync:
Windows Time Service Tools and Settings
Windows Server 2016 Improvements
Accurate Time for Windows Server 2016
Support boundary to configure the Windows Time service for high-accuracy environments
Run scripts in your Windows VM
To automate tasks or troubleshoot issues, you may need to run commands in a VM. The following article gives a
brief overview of the features that are available to run scripts and commands within your VMs.

The Custom Script Extension is primarily used for post deployment configuration and software installation.
Download and run scripts in Azure virtual machines.
Can be run using Azure Resource Manager templates, Azure CLI, REST API, PowerShell, or Azure portal.
Script files can be downloaded from Azure storage or GitHub, or provided from your PC when run from the
Azure portal.
Run PowerShell script in Windows machines and Bash script in Linux machines.
Useful for post deployment configuration, software installation, and other configuration or management tasks.
Run command
The Run Command feature enables virtual machine and application management and troubleshooting using
scripts, and is available even when the machine is not reachable, for example if the guest firewall doesn't have the
RDP or SSH port open.
Run scripts in Azure virtual machines.
Can be run using Azure portal, REST API, Azure CLI, or PowerShell
Quickly run a script and view output and repeat as needed in the Azure portal.
Script can be typed directly or you can run one of the built-in scripts.
Run PowerShell script in Windows machines and Bash script in Linux machines.
Useful for virtual machine and application management and for running scripts in virtual machines that are
unreachable.
Hybrid Runbook Worker

The Hybrid Runbook Worker provides general machine, application, and environment management with user's
custom scripts stored in an Automation account.
Run scripts in Azure and non-Azure machines.
Can be run using Azure portal, Azure CLI, REST API, PowerShell, webhook.
Scripts stored and managed in an Automation Account.
Run PowerShell, PowerShell Workflow, Python, or Graphical runbooks
No time limit on script run time.
Multiple scripts can run concurrently.
Full script output is returned and stored.
Job history available for 90 days.
Scripts can run as Local System or with user-supplied credentials.
Requires manual installation
Serial console
The Serial console provides direct access to a VM, similar to having a keyboard connected to the VM.
Run commands in Azure virtual machines.
Can be run using a text-based console to the machine in the Azure portal.
Login to the machine with a local user account.
Useful when access to the virtual machine is needed regardless of the machine's network or operating system
state.
Next steps
Learn more about the different features that are available to run scripts and commands within your VMs.
Run Command
Hybrid Runbook Worker
Serial console
Custom Script Extension for Windows
The Custom Script Extension downloads and executes scripts on Azure virtual machines. This extension is useful for post deployment configuration, software
installation, or any other configuration / management task. Scripts can be downloaded from Azure storage or GitHub, or provided to the Azure portal at extension run
time. The Custom Script extension integrates with Azure Resource Manager templates, and can also be run using the Azure CLI, PowerShell, Azure portal, or the
Azure Virtual Machine REST API.
This document details how to use the Custom Script Extension using the Azure PowerShell module, Azure Resource Manager templates, and details troubleshooting
steps on Windows systems.
Prerequisites
NOTE
Do not use Custom Script Extension to run Update-AzureRmVM with the same VM as its parameter, since it will wait on itself.
Operating System
The Custom Script Extension for Linux will run on the extension supported extension OS's, for more information, see this article.
Script Location
You can use the extension to use your Azure Blob storage credentials, to access Azure Blob storage. Alternatively, the script location can be any where, as long as the
VM can route to that end point, such as GitHub, internal file server etc.
Internet Connectivity
If you need to download a script externally such as GitHub or Azure Storage, then additional firewall/Network Security Group ports need to be opened. For example if
your script is located in Azure Storage, you can allow access using Azure NSG Service Tags for Storage.
If your script is on a local server, then you may still need additional firewall/Network Security Group ports need to be opened.
Tips and Tricks
The highest failure rate for this extension is due to syntax errors in the script, test the script runs without error, and also put in additional logging into the script to
make it easier to find where it failed.
Write scripts that are idempotent, so if they get run again more than once accidentally, it will not cause system changes.
Ensure the scripts do not require user input when they run.
There is 90 mins allowed for the script to run, anything longer will result in a failed provision of the extension.
Do not put reboots inside the script, this will cause issues with other extensions that are being installed, and post reboot, the extension will not continue after the
restart.
If you have a script that will cause a reboot, then install applications and run scripts etc. You should schedule the reboot using a Windows Scheduled Task, or using
tools such as DSC, or Chef, Puppet extensions.
The extension will only run a script once, if you want to run a script on every boot, then you need to use the extension to create a Windows Scheduled Task.
If you want to schedule when a script will run, you should use the extension to create a Windows Scheduled Task.
When the script is running, you will only see a 'transitioning' extension status from the Azure portal or CLI. If you want more frequent status updates of a running
script, you will need to create your own solution.
Custom Script extension does not natively support proxy servers, however you can use a file transfer tool that supports proxy servers within your script, such as
Curl
Be aware of non default directory locations that your scripts or commands may rely on, have logic to handle this.
Extension schema
The Custom Script Extension configuration specifies things like script location and the command to be run. You can store this configuration in configuration files,
specify it on the command line, or specify it in an Azure Resource Manager template.
You can store sensitive data in a protected configuration, which is encrypted and only decrypted inside the virtual machine. The protected configuration is useful when
the execution command includes secrets such as a password.
These items should be treated as sensitive data and specified in the extensions protected setting configuration. Azure VM extension protected setting data is
encrypted, and only decrypted on the target virtual machine.
{
"apiVersion": "2015-06-15",
"type": "Microsoft.Compute/virtualMachines/extensions",
"name": "config-app",
"dependsOn": [
"[concat('Microsoft.Compute/virtualMachines/', variables('vmName'),copyindex())]",
"[variables('musicstoresqlName')]"
],
"tags": {
"displayName": "config-app"
},
"properties": {
"publisher": "Microsoft.Compute",
"type": "CustomScriptExtension",
"typeHandlerVersion": "1.9",
"autoUpgradeMinorVersion": true,
"settings": {
"fileUris": [
"script location"
]
},
"protectedSettings": {
"commandToExecute": "myExecutionCommand",
"storageAccountName": "myStorageAccountName",
"storageAccountKey": "myStorageAccountKey"
}
}
}
Note - Only one version of an extension can be installed on a VM at a point in time, specifying custom script twice in the same Resource Manager template for the
same VM will fail.
Property values
NAME VALUE / EXAMPLE DATA TYPE
apiVersion 2015-06-15 date
publisher Microsoft.Compute string
type CustomScriptExtension string
typeHandlerVersion 1.9 int
fileUris (e.g) https://raw.githubusercontent.com/Microsoft/dotnet-core- array

sample-templates/master/dotnet-core-music-
windows/scripts/configure-music-app.ps1
commandToExecute (e.g) powershell -ExecutionPolicy Unrestricted -File configure- string

music-app.ps1
storageAccountName (e.g) examplestorageacct string
storageAccountKey (e.g) TmJK/1N3AbAZ3q/+hOXoi/l73zOqsaxXDhqa9Y83/v5UpXQ string

p2DQIBuv2Tifp60cE/OaHsJZmQZ7teQfczQj8hg==
NOTE
These property names are case-sensitive. To avoid deployment problems, use the names as shown here.
Property value details

commandToExecute : (required, string) the entry point script to execute. Use this field instead if your command contains secrets such as passwords, or your fileUris
are sensitive.
fileUris : (optional, string array) the URLs for file(s) to be downloaded.
storageAccountName : (optional, string) the name of storage account. If you specify storage credentials, all fileUris must be URLs for Azure Blobs.
storageAccountKey : (optional, string) the access key of storage account
The following values can be set in either public or protected settings, the extension will reject any configuration where the values below are set in both public and
protected settings.
commandToExecute
Using public settings maybe useful for debugging, but it is strongly recommended that you use protected settings.
Public settings are sent in clear text to the VM where the script will be executed. Protected settings are encrypted using a key known only to the Azure and the VM.
The settings are saved to the VM as they were sent, i.e. if the settings were encrypted they are saved encrypted on the VM. The certificate used to decrypt the
encrypted values is stored on the VM, and used to decrypt settings (if necessary) at runtime.
Template deployment
Azure VM extensions can be deployed with Azure Resource Manager templates. The JSON schema detailed in the previous section can be used in an Azure Resource
Manager template to run the Custom Script Extension during an Azure Resource Manager template deployment. A sample template that includes the Custom Script
Extension can be found here, GitHub.
PowerShell deployment
The Set-AzureRmVMCustomScriptExtension command can be used to add the Custom Script extension to an existing virtual machine. For more information, see Set-
AzureRmVMCustomScriptExtension.
Set-AzureRmVMCustomScriptExtension -ResourceGroupName myResourceGroup `

-VMName myVM `
-Location myLocation `
-FileUri myURL `
-Run 'myScript.ps1' `
-Name DemoScriptExtension
Further Examples
Using Multiple Script
In this example, you have three scripts that are used to build your server, the 'commandToExecute' calls the first script, then you have options on how the others are
called, for example, you can have a master script that controls the execution, with the right error handling, logging, and state management.
$fileUri = @("https://xxxxxxx.blob.core.windows.net/buildServer1/1_Add_Tools.ps1",
"https://xxxxxxx.blob.core.windows.net/buildServer1/2_Add_Features.ps1",
"https://xxxxxxx.blob.core.windows.net/buildServer1/3_CompleteInstall.ps1")
$Settings = @{"fileUris" = $fileUri};
$storageaccname = "xxxxxxx"
$storagekey = "1234ABCD"
$ProtectedSettings = @{"storageAccountName" = $storageaccname; "storageAccountKey" = $storagekey; "commandToExecute" = "powershell -ExecutionPolicy Unrestricted
-File 1_Add_Tools.ps1"};
#run command
Set-AzureRmVMExtension -ResourceGroupName myRG `
-VMName myVM `
-Name "buildserver1" `
-TypeHandlerVersion "1.9" `
-Settings $Settings `
-ProtectedSettings $ProtectedSettings `
Running scripts from a local share

In this example you may want to use a local SMB server for your script location, note, you do not need pass in another other settings, except commandToExecute.
$ProtectedSettings = @{"commandToExecute" = "powershell -ExecutionPolicy Unrestricted -File \\filesvr\build\serverUpdate1.ps1"};
Set-AzureRmVMExtension -ResourceGroupName myRG

-VMName myVM `
-Name "serverUpdate"
-TypeHandlerVersion "1.9" `
-ProtectedSettings $ProtectedSettings
How to run custom script more than once with CLI

If you want to run the custom script extension more than once, you can only do this under these conditions:
1. The extension 'Name' parameter is the same as the previous deployment of the extension.
2. You must updated the configuration otherwise the command will not re-executed, for example, you could add in a dynamic property in to the command, such as a
timestamp.
Troubleshoot and support

Troubleshoot
Data about the state of extension deployments can be retrieved from the Azure portal, and by using the Azure PowerShell module. To see the deployment state of
extensions for a given VM, run the following command:
Get-AzureRmVMExtension -ResourceGroupName myResourceGroup -VMName myVM -Name myExtensionName
Extension execution output is logged to files found under the following directory on the target virtual machine.
C:\WindowsAzure\Logs\Plugins\Microsoft.Compute.CustomScriptExtension
The specified files are downloaded into the following directory on the target virtual machine.
C:\Packages\Plugins\Microsoft.Compute.CustomScriptExtension\1.*\Downloads\<n>
where <n> is a decimal integer which may change between executions of the extension. The 1.* value matches the actual, current typeHandlerVersion value of the
extension. For example, the actual directory could be C:\Packages\Plugins\Microsoft.Compute.CustomScriptExtension\1.8\Downloads\2 .
When executing the commandToExecute command, the extension sets this directory (e.g., ...\Downloads\2 ) as the current working directory. This enables the use of
relative paths to locate the files downloaded via the fileURIs property. See the table below for examples.
Since the absolute download path may vary over time, it is better to opt for relative script/file paths in the commandToExecute string, whenever possible. For example:
"commandToExecute": "powershell.exe . . . -File \"./scripts/myscript.ps1\""
Path information after the first URI segment is retained for files downloaded via the fileUris property list. As shown in the table below, downloaded files are
mapped into download subdirectories to reflect the structure of the fileUris values.
Examples of Downloaded Files
URI IN FILEURIS RELATIVE DOWNLOADED LOCATION ABSOLUTE DOWNLOADED LOCATION *
https://someAcct.blob.core.windows.net/aContainer/scripts/myscript.ps1
./scripts/myscript.ps1 C:\Packages\Plugins\Microsoft.Compute.CustomScriptExtension\1.8\Downloa
https://someAcct.blob.core.windows.net/aContainer/topLevel.ps1
./topLevel.ps1 C:\Packages\Plugins\Microsoft.Compute.CustomScriptExtension\1.8\Downloa
* As above, the absolute directory paths change over the lifetime of the VM, but not within a single execution of the CustomScript extension.
Support
If you need more help at any point in this article, you can contact the Azure experts on the MSDN Azure and Stack Overflow forums. Alternatively, you can file an
Azure support incident. Go to the Azure support site and select Get support. For information about using Azure Support, read the Microsoft Azure support FAQ.
Run PowerShell scripts in your Windows VM with
Run Command
Run Command uses the VM agent to run PowerShell scripts within an Azure Windows VM. These scripts can be
used for general machine or application management, and can be used to quickly diagnose and remediate VM
access and network issues and get the VM back to a good state.
Benefits
There are multiple options that can be used to access your virtual machines. Run Command can run scripts on
your virtual machines remotely using the VM agent. Run Command can be used through the Azure portal, REST
API, or PowerShell for Windows VMs.
This capability is useful in all scenarios where you want to run a script within a virtual machines, and is one of the
only ways to troubleshoot and remediate a virtual machine that doesn't have the RDP or SSH port open due to
improper network or administrative user configuration.
Restrictions
The following restrictions apply when using Run Command:
Output is limited to last 4096 bytes
The minimum time to run a script is about 20 seconds
Scripts run as System on Windows
One script at a time may run
Scripts that prompt for information (interactive mode) are not supported.
You cannot cancel a running script
The maximum time a script can run is 90 minutes, after which it will time out
Outbound connectivity from the VM is required to return the results of the script.
Run a command
Navigate to a VM in Azure and select Run command under OPERATIONS. You are presented with a list of the
available commands to run on the VM.
Choose a command to run. Some of the commands may have optional or required input parameters. For those
commands the parameters are presented as text fields for you to provide the input values. For each command you
can view the script that is being run by expanding View script. RunPowerShellScript is different from the other
commands as it allows you to provide your own custom script.
NOTE
The built-in commands are not editable.
Once the command is chosen, click Run to run the script. The script runs and when complete, returns the output
and any errors in the output window. The following screenshot shows an example output from running the
RDPSettings command.
Commands
This table shows the list of commands available for Windows VMs. The RunPowerShellScript command can be
used to run any custom script you want.
NAME DESCRIPTION
RunPowerShellScript Executes a PowerShell script
EnableRemotePS Configures the machine to enable remote PowerShell.
EnableAdminAccount Checks if the local Administrator account is disabled, and if so

enables it.
IPConfig Shows detailed information for the IP address, subnet mask

and default gateway for each adapter bound to TCP/IP.
RDPSettings Checks registry settings and domain policy settings. Suggests

policy actions if machine is part of a domain or modifies the
settings to default values.
ResetRDPCert Removes the SSL certificate tied to the RDP listener and
restores the RDP listerner security to default. Use this script if
you see any issues with the certificate.
NAME DESCRIPTION
SetRDPPort Sets the default or user specified port number for Remote
Desktop connections. Enables firewall rule for inbound access
to the port.
PowerShell
The following is an example using the Invoke-AzureRmVMRunCommand cmdlet to run a PowerShell script on an
Azure VM.
Invoke-AzureRmVMRunCommand -ResourceGroupName '<myResourceGroup>' -Name '<myVMName>' -CommandId

'RunPowerShellScript' -ScriptPath '<pathToScript>' -Parameter @{"arg1" = "var1";"arg2" = "var2"}
Limiting access to Run Command

Listing the run commands or showing the details of a command require the
Microsoft.Compute/locations/runCommands/read permission, which the built-in Reader role and higher have.
Running a command requires the Microsoft.Compute/virtualMachines/runCommand/action permission, which the

Contributor role and higher have.
You can use one of the built-in roles or create a custom role to use Run Command.
Next steps
See, Run scripts in your Windows VM to learn about other ways to run scripts and commands remotely in your
VM.
Use the D: drive as a data drive on a Windows VM
If your application needs to use the D drive to store data, follow these instructions to use a different drive letter for
the temporary disk. Never use the temporary disk to store data that you need to keep.
If you resize or Stop (Deallocate) a virtual machine, this may trigger placement of the virtual machine to a new
hypervisor. A planned or unplanned maintenance event may also trigger this placement. In this scenario, the
temporary disk will be reassigned to the first available drive letter. If you have an application that specifically
requires the D: drive, you need to follow these steps to temporarily move the pagefile.sys, attach a new data disk
and assign it the letter D and then move the pagefile.sys back to the temporary drive. Once complete, Azure will
not take back the D: if the VM moves to a different hypervisor.
For more information about how Azure uses the temporary disk, see Understanding the temporary drive on
Microsoft Azure Virtual Machines
Attach the data disk

First, you'll need to attach the data disk to the virtual machine. To do this using the portal, see How to attach a
managed data disk in the Azure portal.
Temporarily move pagefile.sys to C drive

1. Connect to the virtual machine.
2. Right-click the Start menu and select System.
3. In the left-hand menu, select Advanced system settings.
4. In the Performance section, select Settings.
5. Select the Advanced tab.
6. In the Virtual memory section, select Change.
7. Select the C drive and then click System managed size and then click Set.
8. Select the D drive and then click No paging file and then click Set.
9. Click Apply. You will get a warning that the computer needs to be restarted for the changes to take affect.
10. Restart the virtual machine.
Change the drive letters

1. Once the VM restarts, log back on to the VM.
2. Click the Start menu and type diskmgmt.msc and hit Enter. Disk Management will start.
3. Right-click on D, the Temporary Storage drive, and select Change Drive Letter and Paths.
4. Under Drive letter, select a new drive such as T and then click OK.
5. Right-click on the data disk, and select Change Drive Letter and Paths.
6. Under Drive letter, select drive D and then click OK.
Move pagefile.sys back to the temporary storage drive

1. Right-click the Start menu and select System
2. In the left-hand menu, select Advanced system settings.
3. In the Performance section, select Settings.
4. Select the Advanced tab.
5. In the Virtual memory section, select Change.
6. Select the OS drive C and click No paging file and then click Set.
7. Select the temporary storage drive T and then click System managed size and then click Set.
8. Click Apply. You will get a warning that the computer needs to be restarted for the changes to take affect.
9. Restart the virtual machine.
Next steps
You can increase the storage available to your virtual machine by attaching a additional data disk.
Change the availability set for a Windows VM
The following steps describe how to change the availability set of a VM using Azure PowerShell. A VM can only be
added to an availability set when it is created. To change the availability set, you need to delete and then recreate
the virtual machine.
Change the availability set

The following script provides an example of gathering the required information, deleting the original VM and then
recreating it in a new availability set.
# Set variables
$vmName = "myVM"
$newAvailSetName = "myAvailabilitySet"
# Get the details of the VM to be moved to the Availablity Set

$originalVM = Get-AzureRmVM `
-Name $vmName
# Create new availability set if it does not exist

$availSet = Get-AzureRmAvailabilitySet `
-Name $newAvailSetName `
-ErrorAction Ignore
if (-Not $availSet) {
$availSet = New-AzureRmAvailabilitySet `
-Location $originalVM.Location `
-Name $newAvailSetName `
-PlatformUpdateDomainCount 2 `
-Sku Aligned
}
# Remove the original VM

Remove-AzureRmVM -ResourceGroupName $resourceGroup -Name $vmName
# Create the basic configuration for the replacement VM

$newVM = New-AzureRmVMConfig `
-VMName $originalVM.Name `
-VMSize $originalVM.HardwareProfile.VmSize `
-AvailabilitySetId $availSet.Id
Set-AzureRmVMOSDisk `
-VM $newVM -CreateOption Attach `
-ManagedDiskId $originalVM.StorageProfile.OsDisk.ManagedDisk.Id `
-Name $originalVM.StorageProfile.OsDisk.Name `
-Windows
# Add Data Disks

foreach ($disk in $originalVM.StorageProfile.DataDisks) {
Add-AzureRmVMDataDisk -VM $newVM `
-Name $disk.Name `
-ManagedDiskId $disk.ManagedDisk.Id `
-Caching $disk.Caching `
-Lun $disk.Lun `
-DiskSizeInGB $disk.DiskSizeGB `
-CreateOption Attach
}
# Add NIC(s)
foreach ($nic in $originalVM.NetworkProfile.NetworkInterfaces) {
Add-AzureRmVMNetworkInterface `
-VM $newVM `
-Id $nic.Id
}
# Recreate the VM
New-AzureRmVM `
-Location $originalVM.Location `
-VM $newVM `
-DisableBginfoExtension
Next steps
Add additional storage to your VM by adding an additional data disk.
Download the template for a VM
When you create a VM in Azure using the portal or PowerShell, a Resource Manager template is automatically
created for you. You can use this template to quickly duplicate a deployment. The template contains information
about all of the resources in a resource group. For a virtual machine, this means the template contains everything
that is created in support of the VM in that resource group, including the networking resources.
Download the template using the portal

1. Log in to the Azure portal.
2. One the left menu, select Virtual Machines.
3. Select the virtual machine from the list.
4. Select Automation script.
5. Select Download from the menu at the top and save the .zip file to your local computer.
6. Open the .zip file and extract the files to a folder. The .zip file contains:
deploy.ps1
deploy.sh
deployer.rb
DeploymentHelper.cs
parameters.json
template.json
The template.json file is the template.
Download the template using PowerShell

You can also download the .json template file using the Export-AzureRMResourceGroup cmdlet. You can use the
-path parameter to provide the filename and path for the .json file. This example shows how to download the
template for the resource group named myResourceGroup to the C:\users\public\downloads folder on your
local computer.
Export-AzureRmResourceGroup -ResourceGroupName "myResourceGroup" -Path "C:\users\public\downloads"
Next steps
To learn more about deploying resources using templates, see Resource Manager template walkthrough.
Azure Virtual Machine Agent overview
The Microsoft Azure Virtual Machine Agent (VM Agent) is a secure, lightweight process that manages virtual
machine (VM ) interaction with the Azure Fabric Controller. The VM Agent has a primary role in enabling and
executing Azure virtual machine extensions. VM Extensions enable post-deployment configuration of VM, such as
installing and configuring software. VM extensions also enable recovery features such as resetting the
administrative password of a VM. Without the Azure VM Agent, VM extensions cannot be run.
This article details installation, detection, and removal of the Azure Virtual Machine Agent.
Install the VM Agent

Azure Marketplace image
The Azure VM Agent is installed by default on any Windows VM deployed from an Azure Marketplace image.
When you deploy an Azure Marketplace image from the portal, PowerShell, Command Line Interface, or an Azure
Resource Manager template, the Azure VM Agent is also installed.
The Windows Guest Agent Package is broken into two parts:
Provisioning Agent (PA)
Windows Guest Agent (WinGA)
To boot a VM you must have the PA installed on the VM, however the WinGA does not need to be installed. At VM
deploy time, you can select not to install the WinGA. The following example shows how to select the
provisionVmAgent option with an Azure Resource Manager template:
"resources": [{
"name": "[parameters('virtualMachineName')]",
"location": "[parameters('location')]",
"dependsOn": ["[concat('Microsoft.Network/networkInterfaces/', parameters('networkInterfaceName'))]"],
"properties": {
"osProfile": {
"computerName": "[parameters('virtualMachineName')]",
"adminPassword": "[parameters('adminPassword')]",
"windowsConfiguration": {
"provisionVmAgent": "false"
}
If you do not have the Agents installed, you cannot use some Azure services, such as Azure Backup or Azure
Security. These services require an extension to be installed. If you have deployed a VM without the WinGA, you
can install the latest version of the agent later.
Manual installation
The Windows VM agent can be manually installed with a Windows installer package. Manual installation may be
necessary when you create a custom VM image that is deployed to Azure. To manually install the Windows VM
Agent, download the VM Agent installer.
The VM Agent can be installed by double-clicking the Windows installer file. For an automated or unattended
installation of the VM agent, run the following command:
msiexec.exe /i WindowsAzureVmAgent.2.7.1198.778.rd_art_stable.160617-1120.fre /quiet
Detect the VM Agent

PowerShell
The Azure Resource Manager PowerShell module can be used to retrieve information about Azure VMs. To see
information about a VM, such as the provisioning state for the Azure VM Agent, use Get-AzureRmVM:
Get-AzureRmVM
The following condensed example output shows the ProvisionVMAgent property nested inside OSProfile. This
property can be used to determine if the VM agent has been deployed to the VM:
OSProfile :
ComputerName : myVM
AdminUsername : myUserName
WindowsConfiguration :
ProvisionVMAgent : True
EnableAutomaticUpdates : True
The following script can be used to return a concise list of VM names and the state of the VM Agent:
$vms = Get-AzureRmVM
foreach ($vm in $vms) {

$agent = $vm | Select -ExpandProperty OSProfile | Select -ExpandProperty Windowsconfiguration | Select
ProvisionVMAgent
Write-Host $vm.Name $agent.ProvisionVMAgent
}
Manual Detection
When logged in to a Windows VM, Task Manager can be used to examine running processes. To check for the
Azure VM Agent, open Task Manager, click the Details tab, and look for a process name
WindowsAzureGuestAgent.exe. The presence of this process indicates that the VM agent is installed.
Upgrade the VM Agent

The Azure VM Agent for Windows is automatically upgraded. As new VMs are deployed to Azure, they receive the
latest VM agent at VM provision time. Custom VM images should be manually updated to include the new VM
agent at image creation time.
Next steps
For more information about VM extensions, see Azure virtual machine extensions and features overview .
Handling planned maintenance notifications for
Windows virtual machines
Azure periodically performs updates to improve the reliability, performance, and security of the host infrastructure
for virtual machines. Updates are changes like patching the hosting environment or upgrading and
decommissioning hardware. A majority of these updates are performed without any impact to the hosted virtual
machines. However, there are cases where updates do have an impact:
If the maintenance does not require a reboot, Azure uses in-place migration to pause the VM while the host
is updated. These non-rebootful maintenance operations are applied fault domain by fault domain, and
progress is stopped if any warning health signals are received.
If maintenance requires a reboot, you get a notice of when the maintenance is planned. In these cases, you
are given a time window where you can start the maintenance yourself, when it works for you.
Planned maintenance that requires a reboot, is scheduled in waves. Each wave has different scope (regions).
A wave starts with a notification to customers. By default, notification is sent to subscription owner and co-
owners. You can add more recipients and messaging options like email, SMS, and Webhooks, to the
notifications using Azure Activity Log Alerts.
At the time of the notification, a self-service window is made available. During this window, you can find which
of your virtual machines are included in this wave and proactively start maintenance according to your own
scheduling needs.
After the self-service window, a scheduled maintenance window begins. At some point during this window,
Azure schedules and applies the required maintenance to your virtual machine.
The goal in having two windows is to give you enough time to start maintenance and reboot your virtual machine
while knowing when Azure will automatically start maintenance.
You can use the Azure portal, PowerShell, REST API, and CLI to query for the maintenance windows for your VMs
and start self-service maintenance.
Should you start maintenance during the self-service window?

The following guidelines should help you decide whether to use this capability and start maintenance at your own
time.
NOTE
Self-service maintenance might not be available for all of your VMs. To determine if proactive redeploy is available for your
VM, look for the Start now in the maintenance status. Self-service maintenance is currently not available for Cloud Services
(Web/Worker Role) and Service Fabric.
Self-service maintenance is not recommended for deployments using availability sets since these are highly
available setups, where only one update domain is impacted at any given time.
Let Azure trigger the maintenance. For maintenance that requires reboot, be aware that the maintenance will be
done update domain by update domain, that the update domains do not necessarily receive the maintenance
sequentially, and that there is a 30-minute pause between update domains.
If a temporary loss of some of your capacity (1/update domain count) is a concern, it can easily be compensated
for by allocating additional instances during the maintenance period.
For maintenance that does not require reboot, updates are applied at the fault domain level.
Don't use self-service maintenance in the following scenarios:
If you shut down your VMs frequently, either manually, using DevTest Labs, using auto-shutdown, or following
a schedule, it could revert the maintenance status and therefore cause additional downtime.
On short-lived VMs that you know will be deleted before the end of the maintenance wave.
For workloads with a large state stored in the local (ephemeral) disk that is desired to be maintained upon
update.
For cases where you resize your VM often, as it could revert the maintenance status.
If you have adopted scheduled events that enable proactive failover or graceful shutdown of your workload, 15
minutes before start of maintenance shutdown
Use self-service maintenance, if you are planning to run your VM uninterrupted during the scheduled maintenance
phase and none of the counter-indications mentioned above are applicable.
It is best to use self-service maintenance in the following cases:
You need to communicate an exact maintenance window to your management or end-customer.
You need to complete the maintenance by a given date.
You need to control the sequence of maintenance, for example, multi-tier application to guarantee safe recovery.
You need more than 30 minutes of VM recovery time between two update domains (UDs). To control the time
between update domains, you must trigger maintenance on your VMs one update domain (UD ) at a time.
View VMs scheduled for maintenance in the portal

Once a planned maintenance wave is scheduled, you can observe the list of virtual machines that are impacted by
the upcoming maintenance wave.
You can use the Azure portal and look for VMs scheduled for maintenance.
2. In the left navigation, click Virtual Machines.
3. In the Virtual Machines pane, click the Columns button to open the list of available columns.
4. Select and add the following columns:
Maintenance: Shows the maintenance status for the VM. The following are the potential values:
VALUE DESCRIPTION
Start now The VM is in the self-service maintenance window that lets

you initiate the maintenance yourself. See below on how to
start maintenance on your VM.
Scheduled The VM is scheduled for maintenance with no option for

you to initiate maintenance. You can learn of the
maintenance window by selecting the Maintenance -
Scheduled window in this view or by clicking on the VM.
Already updated Your VM is already updated and no further action is

required at this time.
VALUE DESCRIPTION
Retry later You have initiated maintenance with no success. You will
be able to use the self-service maintenance option at a
later time.
Retry now You can retry a previously unsuccessful self-initiated

maintenance.
- Your virtual machine is not part of a planned maintenance

wave.
Maintenance - Self-service window: Shows the time window when you can self-start maintenance on your
VMs.
Maintenance - Scheduled window: Shows the time window when Azure will maintain your VM in order to
complete maintenance.
Notification and alerts in the portal

Azure communicates a schedule for planned maintenance by sending an email to the subscription owner and co-
owners group. You can add additional recipients and channels to this communication by creating Azure activity log
alerts. For more information, see Monitor subscription activity with the Azure Activity Log.
2. In the menu on the left, select Monitor.
3. In the Monitor - Alerts (classic) pane, click + Add activity log alert.
4. Complete the information in the Add activity log alert page and make sure you set the following in Criteria:
Event category: Service Health
Services: Virtual Machine Scale Sets and Virtual Machines
Type: Planned maintenance
To learn more on how to configure Activity Log Alerts, see Create activity log alerts.
Start Maintenance on your VM from the portal

While looking at the VM details, you will be able to see more maintenance-related details.
At the top of the VM details view, a new notification ribbon will be added if your VM is included in a planned
maintenance wave. In addition, a new option is added to start maintenance when possible.
Click on the maintenance notification to see the maintenance page with more details on the planned maintenance.
From there, you will be able to start maintenance on your VM.
Once you start maintenance, your virtual machine will be maintained and the maintenance status will be updated
to reflect the result within few minutes.
If you missed the self-service window, you will still be able to see the window when your VM will be maintained by
Azure.
Check maintenance status using PowerShell

You can also use Azure Powershell to see when VMs are scheduled for maintenance. Planned maintenance
information is available from the Get-AzureRmVM cmdlet when you use the -status parameter.
Maintenance information is returned only if there is maintenance planned. If no maintenance is scheduled that
impacts the VM, the cmdlet does not return any maintenance information.
Get-AzureRmVM -ResourceGroupName rgName -Name vmName -Status
The following properties are returned under MaintenanceRedeployStatus:
VALUE DESCRIPTION
IsCustomerInitiatedMaintenanceAllowed Indicates whether you can start maintenance on the VM at

this time
PreMaintenanceWindowStartTime The beginning of the maintenance self-service window when

you can initiate maintenance on your VM
PreMaintenanceWindowEndTime The end of the maintenance self-service window when you can
initiate maintenance on your VM
MaintenanceWindowStartTime The beginning of the maintenance scheduled in which Azure

initiates maintenance on your VM
MaintenanceWindowEndTime The end of the maintenance scheduled window in which Azure

initiates maintenance on your VM
LastOperationResultCode The result of the last attempt to initiate maintenance on the

VM
You can also get the maintenance status for all VMs in a resource group by using Get-AzureRmVM and not
specifying a VM.
Get-AzureRmVM -ResourceGroupName rgName -Status
The following PowerShell function takes your subscription ID and prints out a list of VMs that are scheduled for
maintenance.
function MaintenanceIterator
{
Select-AzureRmSubscription -SubscriptionId $args[0]
$rgList= Get-AzureRmResourceGroup
for ($rgIdx=0; $rgIdx -lt $rgList.Length ; $rgIdx++)

{
$rg = $rgList[$rgIdx]
$vmList = Get-AzureRMVM -ResourceGroupName $rg.ResourceGroupName
for ($vmIdx=0; $vmIdx -lt $vmList.Length ; $vmIdx++)
{
$vm = $vmList[$vmIdx]
$vmDetails = Get-AzureRMVM -ResourceGroupName $rg.ResourceGroupName -Name $vm.Name -Status
if ($vmDetails.MaintenanceRedeployStatus )
{
Write-Output "VM: $($vmDetails.Name) IsCustomerInitiatedMaintenanceAllowed:
$($vmDetails.MaintenanceRedeployStatus.IsCustomerInitiatedMaintenanceAllowed)
$($vmDetails.MaintenanceRedeployStatus.LastOperationMessage)"
}
}
}
}
Start maintenance on your VM using PowerShell

Using information from the function in the previous section, the following starts maintenance on a VM if
IsCustomerInitiatedMaintenanceAllowed is set to true.
Restart-AzureRmVM -PerformMaintenance -name $vm.Name -ResourceGroupName $rg.ResourceGroupName
Classic deployments
If you still have legacy VMs that were deployed using the classic deployment model, you can use PowerShell to
query for VMs and initiate maintenance.
To get the maintenance status of a VM, type:
Get-AzureVM -ServiceName <Service name> -Name <VM name>
To start maintenance on your classic VM, type:
Restart-AzureVM -InitiateMaintenance -ServiceName <service name> -Name <VM name>
FAQ
Q: Why do you need to reboot my virtual machines now?
A: While the majority of updates and upgrades to the Azure platform do not impact virtual machine's availability,
there are cases where we can't avoid rebooting virtual machines hosted in Azure. We have accumulated several
changes that require us to restart our servers that will result in virtual machines reboot.
Q: If I follow your recommendations for High Availability by using an Availability Set, am I safe?
A: Virtual machines deployed in an availability set or virtual machine scale sets have the notion of Update Domains
(UD ). When performing maintenance, Azure honors the UD constraint and will not reboot virtual machines from
different UD (within the same availability set). Azure also waits for at least 30 minutes before moving to the next
group of virtual machines.
For more information about high availability, see Regions and availability for virtual machines in Azure.
Q: How do I get notified about planned maintenance?
A: A planned maintenance wave starts by setting a schedule to one or more Azure regions. Soon after, an email
notification is sent to the subscription owners (one email per subscription). Additional channels and recipients for
this notification could be configured using Activity Log Alerts. In case you deploy a virtual machine to a region
where planned maintenance is already scheduled, you will not receive the notification but rather need to check the
maintenance state of the VM.
Q: I don't see any indication of planned maintenance in the portal, Powershell, or CLI. What is wrong?
A: Information related to planned maintenance is available during a planned maintenance wave only for the VMs
that are going to be impacted by it. In other words, if you see not data, it could be that the maintenance wave has
already completed (or not started) or that your virtual machine is already hosted in an updated server.
Q: Is there a way to know exactly when my virtual machine will be impacted?
A: When setting the schedule, we define a time window of several days. However, the exact sequencing of servers
(and VMs) within this window is unknown. Customers who would like to know the exact time for their VMs can use
scheduled events and query from within the virtual machine and receive a 15-minute notification before a VM
reboot.
Q: How long will it take you to reboot my virtual machine?
A: Depending on the size of your VM, reboot may take up to several minutes during the self-service maintenance
window. During the Azure initiated reboots in the scheduled maintenance window, the reboot will typically take
about 25 minutes. Note that in case you use Cloud Services (Web/Worker Role), Virtual Machine Scale Sets, or
availability sets, you will be given 30 minutes between each group of VMs (UD ) during the scheduled maintenance
window.
Q: What is the experience in the case of Virtual Machine Scale Sets?
A: Planned maintenance is now available for Virtual Machine Scale Sets. For instructions on how to initiate self-
service maintenance refer planned maintenance for VMSS document.
Q: What is the experience in the case of Cloud Services (Web/Worker Role) and Service Fabric?
A: While these platforms are impacted by planned maintenance, customers using these platforms are considered
safe given that only VMs in a single Upgrade Domain (UD ) will be impacted at any given time. Self-service
maintenance is currently not available for Cloud Services (Web/Worker Role) and Service Fabric.
Q: I don’t see any maintenance information on my VMs. What went wrong?
A: There are several reasons why you’re not seeing any maintenance information on your VMs:
1. You are using a subscription marked as Microsoft internal.
2. Your VMs are not scheduled for maintenance. It could be that the maintenance wave has ended, canceled, or
modified so that your VMs are no longer impacted by it.
3. You don’t have the Maintenance column added to your VM list view. While we have added this column to the
default view, customers who configured to see non-default columns must manually add the Maintenance
column to their VM list view.
Q: My VM is scheduled for maintenance for the second time. Why?
A: There are several use cases where you will see your VM scheduled for maintenance after you have already
completed your maintenance-redeploy:
1. We have canceled the maintenance wave and restarted it with a different payload. It could be that we've
detected faulted payload and we simply need to deploy an additional payload.
2. Your VM was service healed to another node due to a hardware fault
3. You have selected to stop (deallocate) and restart the VM
4. You have auto shutdown turned on for the VM
Next steps
Learn how you can register for maintenance events from within the VM using Scheduled Events.
Guidance for mitigating speculative execution side-
channel vulnerabilities in Azure
Last document update: 14 August 2018 10:00 AM PST.

The disclosure of a new class of CPU vulnerabilities known as speculative execution side-channel attacks has
resulted in questions from customers seeking more clarity.
Microsoft has deployed mitigations across all our cloud services. The infrastructure that runs Azure and isolates
customer workloads from each other is protected. This means that a potential attacker using the same
infrastructure can’t attack your application using these vulnerabilities.
Azure is using memory preserving maintenance whenever possible, to minimize customer impact and eliminate
the need for reboots. Azure will continue utilizing these methods when making systemwide updates to the host and
protect our customers.
More information about how security is integrated into every aspect of Azure is available on the Azure Security
Documentation site.
NOTE
Since this document was first published, multiple variants of this vulnerability class have been disclosed. Microsoft continues
to be heavily invested in protecting our customers and providing guidance. This page will be updated as we continue to
release further fixes.
On August 14, 2018, the industry disclosed a new speculative execution side channel vulnerability known as L1 Terminal Fault
(L1TF) which has been assigned multiple CVEs (CVE-2018-3615, CVE-2018-3620, and CVE-2018-3646). This vulnerability
affects Intel® Core® processors and Intel® Xeon® processors. Microsoft has deployed mitigations across our cloud services
which reinforce the isolation between customers. Please read below for additional guidance to protect against L1TF and
previous vulnerabilities (Spectre Variant 2 CVE-2017-5715 and Meltdown Variant 3 CVE-2017-5754).
Keeping your Operating Systems up-to-date

While an OS update is not required to isolate your applications running on Azure from other Azure customers, it is
always a best practice to keep your software up-to-date. The latest Security Rollups for Windows contain
mitigations for several speculative execution side channel vulnerabilities. Similarly, Linux distributions have
released multiple updates to address these vulnerabilities. Here are our recommended actions to update your
Operating System:
OFFERING RECOMMENDED ACTION
Azure Cloud Services Enable auto update or ensure you are running the newest
Guest OS.
Azure Linux Virtual Machines Install updates from your operating system provider. For more
information, see Linux later in this document.
Azure Windows Virtual Machines Install the latest security rollup.

OFFERING RECOMMENDED ACTION
Other Azure PaaS Services There is no action needed for customers using these services.
Azure automatically keeps your OS versions up-to-date.
Additional guidance if you are running untrusted code

Customers who allow untrusted users to execute arbitrary code may wish to implement some additional security
features inside their Azure Virtual Machines or Cloud Services. These features protect against the intra-process
disclosure vectors that several speculative execution vulnerabilities describe.
Example scenarios where additional security features are recommended:
You allow code that you do not trust to run inside your VM.
For example, you allow one of your customers to upload a binary or script that you then execute within
your application.
You allow users that you do not trust to log into your VM using low privileged accounts.
For example, you allow a low -privileged user to log into one of your VMs using remote desktop or SSH.
You allow untrusted users access to virtual machines implemented via nested virtualization.
For example, you control the Hyper-V host, but allocate the VMs to untrusted users.
Customers who do not implement a scenario involving untrusted code do not need to enable these additional
security features.
Enabling additional security

You can enable additional security features inside your VM or Cloud Service.
Windows
Your target operating system must be up-to-date to enable these additional security features. While numerous
speculative execution side channel mitigations are enabled by default, the additional features described here must
be enabled manually and may cause a performance impact.
Step 1: Contact Azure Support to expose updated firmware (microcode) into your Virtual Machines.
Step 2: Enable Kernel Virtual Address Shadowing (KVAS ) and Branch Target Injection (BTI) OS support. Follow the
instructions in KB4072698 to enable protections via the Session Manager registry keys. A reboot is required.
Step 3: For deployments that are using nested virtualization (D3 and E3 only): These instructions apply inside the
VM you are using as a Hyper-V host.
1. Follow the instructions in KB4072698 to enable protections via the MinVmVersionForCpuBasedMitigations
registry keys.
2. Set the hypervisor scheduler type to Core by following the instructions here.
Step 4: Follow the instructions in KB4072698 to verify protections are enabled using the SpeculationControl
PowerShell module.
NOTE
If you previously downloaded this module, you will need to install the newest version.
All VMs should show:

branch target injection mitigation is enabled: True
kernel VA shadow is enabled: True
L1TFWindowsSupportEnabled: True
Linux
Enabling the set of additional security features inside requires that the target operating system be fully up-to-date.
Some mitigations will be enabled by default. The following section describes the features which are off by default
and/or reliant on hardware support (microcode). Enabling these features may cause a performance impact.
Reference your operating system provider’s documentation for further instructions
Step 1: Contact Azure Support to expose updated firmware (microcode) into your Virtual Machines.
Step 2: Enable Branch Target Injection (BTI) OS support to mitigate CVE -2017-5715 (Spectre Variant 2) by
following your operating system provider’s documentation.
Step 3: Enable Kernel Page Table Isolation (KPTI) to mitigate CVE -2017-5754 (Meltdown Variant 3) by following
your operating system provider’s documentation.
More information is available from your operating system’s provider:
Redhat and CentOS
Suse
Ubuntu
Next steps
To learn more, see Securing Azure customers from CPU vulnerability.
Azure Metadata Service: Scheduled Events for
Windows VMs
Scheduled Events is an Azure Metadata Service that gives your application time to prepare for virtual machine
maintenance. It provides information about upcoming maintenance events (e.g. reboot) so your application can
prepare for them and limit disruption. It is available for all Azure Virtual Machine types including PaaS and IaaS
on both Windows and Linux.
For information about Scheduled Events on Linux, see Scheduled Events for Linux VMs.
NOTE
Scheduled Events is generally available in all Azure Regions. See Version and Region Availability for latest release information.
Why Scheduled Events?

Many applications can benefit from time to prepare for virtual machine maintenance. The time can be used to
perform application specific tasks that improve availability, reliability, and serviceability including:
Checkpoint and restore
Connection draining
Primary replica failover
Removal from load balancer pool
Event logging
Graceful shutdown
Using Scheduled Events your application can discover when maintenance will occur and trigger tasks to limit its
impact.
Scheduled Events provides events in the following use cases:
Platform initiated maintenance (e.g. Host OS Update)
User initiated maintenance (e.g. user restarts or redeploys a VM )
The Basics
Azure Metadata service exposes information about running Virtual Machines using a REST Endpoint accessible
from within the VM. The information is available via a non-routable IP so that it is not exposed outside the VM.
Endpoint Discovery
For VNET enabled VMs, the metadata service is available from a static non-routable IP, 169.254.169.254 . The full
endpoint for the latest version of Scheduled Events is:
http://169.254.169.254/metadata/scheduledevents?api-version=2017-08-01
If the Virtual Machine is not created within a Virtual Network, the default cases for cloud services and classic VMs,
additional logic is required to discover the IP address to use. Refer to this sample to learn how to discover the host
endpoint.
Version and Region Availability
The Scheduled Events Service is versioned. Versions are mandatory and the current version is 2017-08-01 .
VERSION RELEASE TYPE REGIONS RELEASE NOTES
2017-08-01 General Availability All Removed prepended

underscore from resource
names for Iaas VMs
Metadata Header
requirement enforced for all
requests
2017-03-01 Preview All Initial release
NOTE
Previous preview releases of scheduled events supported {latest} as the api-version. This format is no longer supported and
will be deprecated in the future.
Enabling and Disabling Scheduled Events

Scheduled Events is enabled for your service the first time you make a request for events. You should expect a
delayed response in your first call of up to two minutes.
Scheduled Events is disabled for your service if it does not make a request for 24 hours.
User Initiated Maintenance
User initiated virtual machine maintenance via the Azure portal, API, CLI, or PowerShell results in a scheduled
event. This allows you to test the maintenance preparation logic in your application and allows your application to
prepare for user initiated maintenance.
Restarting a virtual machine schedules an event with type Reboot . Redeploying a virtual machine schedules an
event with type Redeploy .
Using the API

Headers
When you query the Metadata Service, you must provide the header Metadata:true to ensure the request was not
unintentionally redirected. The Metadata:true header is required for all scheduled events requests. Failure to
include the header in the request will result in a Bad Request response from the Metadata Service.
Query for events
You can query for Scheduled Events simply by making the following call:
Powershell
curl http://169.254.169.254/metadata/scheduledevents?api-version=2017-08-01 -H @{"Metadata"="true"}
A response contains an array of scheduled events. An empty array means that there are currently no events
scheduled. In the case where there are scheduled events, the response contains an array of events:
{
"DocumentIncarnation": {IncarnationID},
"Events": [
{
"EventId": {eventID},
"EventType": "Reboot" | "Redeploy" | "Freeze",
"ResourceType": "VirtualMachine",
"Resources": [{resourceName}],
"EventStatus": "Scheduled" | "Started",
"NotBefore": {timeInUTC},
}
]
}
Event Properties
PROPERTY DESCRIPTION
EventId Globally unique identifier for this event.
Example:
602d9444-d2cd-49c7-8624-8643e7171297
EventType Impact this event causes.
Values:
Freeze : The Virtual Machine is scheduled to pause
for few seconds. The CPU is suspended, but there is no
impact on memory, open files, or network connections.
Reboot : The Virtual Machine is scheduled for reboot
(non-persistent memory is lost).
Redeploy : The Virtual Machine is scheduled to move
to another node (ephemeral disks are lost).
ResourceType Type of resource this event impacts.
Values:
VirtualMachine
Resources List of resources this event impacts. This is guaranteed to

contain machines from at most one Update Domain, but may
not contain all machines in the UD.
Example:
["FrontEnd_IN_0", "BackEnd_IN_0"]
Event Status Status of this event.
Values:
Scheduled : This event is scheduled to start after the
time specified in the NotBefore property.
Started : This event has started.
No Completed or similar status is ever provided; the event

will no longer be returned when the event is completed.
PROPERTY DESCRIPTION
NotBefore Time after which this event may start.
Example:
Mon, 19 Sep 2016 18:29:47 GMT
Event Scheduling
Each event is scheduled a minimum amount of time in the future based on event type. This time is reflected in an
event's NotBefore property.
EVENTTYPE MINIMUM NOTICE
Freeze 15 minutes
Reboot 15 minutes
Redeploy 10 minutes
Event Scope
Scheduled events are delivered to:
All Virtual Machines in a Cloud Service
All Virtual Machines in an Availability Set
All Virtual Machines in a Scale Set Placement Group.
As a result, you should check the Resources field in the event to identify which VMs are going to be impacted.
Starting an event
Once you have learned of an upcoming event and completed your logic for graceful shutdown, you can approve
the outstanding event by making a POST call to the metadata service with the EventId . This indicates to Azure
that it can shorten the minimum notification time (when possible).
The following is the json expected in the POST request body. The request should contain a list of StartRequests .
Each StartRequest contains the EventId for the event you want to expedite:
{
"StartRequests" : [
{
"EventId": {EventId}
}
]
}
Powershell
curl -H @{"Metadata"="true"} -Method POST -Body '{"StartRequests": [{"EventId": "f020ba2e-3bc0-4c40-a10b-

86575a9eabd5"}]}' -Uri http://169.254.169.254/metadata/scheduledevents?api-version=2017-08-01
NOTE
Acknowledging an event allows the event to proceed for all Resources in the event, not just the virtual machine that
acknowledges the event. You may therefore choose to elect a leader to coordinate the acknowledgement, which may be as
simple as the first machine in the Resources field.
PowerShell sample
The following sample queries the metadata service for scheduled events and approves each outstanding event.
# How to get scheduled events

function Get-ScheduledEvents($uri)
{
$scheduledEvents = Invoke-RestMethod -Headers @{"Metadata"="true"} -URI $uri -Method get
$json = ConvertTo-Json $scheduledEvents
Write-Host "Received following events: `n" $json
return $scheduledEvents
}
# How to approve a scheduled event

function Approve-ScheduledEvent($eventId, $uri)
{
# Create the Scheduled Events Approval Document
$startRequests = [array]@{"EventId" = $eventId}
$scheduledEventsApproval = @{"StartRequests" = $startRequests}
# Convert to JSON string

$approvalString = ConvertTo-Json $scheduledEventsApproval
Write-Host "Approving with the following: `n" $approvalString
# Post approval string to scheduled events endpoint

Invoke-RestMethod -Uri $uri -Headers @{"Metadata"="true"} -Method POST -Body $approvalString
}
function Handle-ScheduledEvents($scheduledEvents)
{
# Add logic for handling events here
}
######### Sample Scheduled Events Interaction #########
# Set up the scheduled events URI for a VNET-enabled VM

$localHostIP = "169.254.169.254"
$scheduledEventURI = 'http://{0}/metadata/scheduledevents?api-version=2017-08-01' -f $localHostIP
# Get events
$scheduledEvents = Get-ScheduledEvents $scheduledEventURI
# Handle events however is best for your service

Handle-ScheduledEvents $scheduledEvents
# Approve events when ready (optional)

foreach($event in $scheduledEvents.Events)
{
Write-Host "Current Event: `n" $event
$entry = Read-Host "`nApprove event? Y/N"
if($entry -eq "Y" -or $entry -eq "y")
{
Approve-ScheduledEvent $event.EventId $scheduledEventURI
}
}
Next steps
Watch a Scheduled Events Demo on Azure Friday.
Review the Scheduled Events code samples in the Azure Instance Metadata Scheduled Events Github
Repository
Read more about the APIs available in the Instance Metadata service.
Learn about planned maintenance for Windows virtual machines in Azure.
Azure Instance Metadata service
The Azure Instance Metadata Service provides information about running virtual machine instances that can be
used to manage and configure your virtual machines. This includes information such as SKU, network
configuration, and upcoming maintenance events. For more information on what type of information is available,
see metadata categories.
Azure's Instance Metadata Service is a REST Endpoint accessible to all IaaS VMs created via the Azure Resource
Manager. The endpoint is available at a well-known non-routable IP address ( 169.254.169.254 ) that can be
accessed only from within the VM.
IMPORTANT
This service is generally available in all Azure Regions. It regularly receives updates to expose new information about
virtual machine instances. This page reflects the up-to-date data categories available.
Service availability
The service is available in generally available Azure regions. Not all API version may be available in all Azure
Regions.
REGIONS AVAILABILITY? SUPPORTED VERSIONS
All Generally Available Global Azure Generally Available 2017-04-02, 2017-08-01, 2017-12-01,
Regions 2018-02-01, 2018-04-02
Azure Government Generally Available 2017-04-02, 2017-08-01, 2017-12-01,

2018-02-01
Azure China Generally Available 2017-04-02, 2017-08-01, 2017-12-01,

2018-02-01
Azure Germany Generally Available 2017-04-02, 2017-08-01, 2017-12-01,

2018-02-01
This table is updated when there are service updates and or new supported versions are available
To try out the Instance Metadata Service, create a VM from Azure Resource Manager or the Azure portal in the
above regions and follow the examples below.
Usage
Versioning
The Instance Metadata Service is versioned. Versions are mandatory and the current version on Global Azure is
2018-04-02 . Current supported versions are (2017 -04 -02, 2017 -08 -01, 2017 -12 -01, 2018 -02 -01, 2018 -04 -02 )
NOTE
Previous preview releases of scheduled events supported {latest} as the api-version. This format is no longer supported and
will be deprecated in the future.
As newer versions are added, older versions can still be accessed for compatibility if your scripts have
dependencies on specific data formats. However, the previous preview version (2017-03-01) may not be available
once the service is generally available.
Using headers
When you query the Instance Metadata Service, you must provide the header Metadata: true to ensure the
request was not unintentionally redirected.
Retrieving metadata
Instance metadata is available for running VMs created/managed using Azure Resource Manager. Access all data
categories for a virtual machine instance using the following request:
curl -H Metadata:true "http://169.254.169.254/metadata/instance?api-version=2017-08-01"
NOTE
All instance metadata queries are case-sensitive.
Data output
By default, the Instance Metadata Service returns data in JSON format ( Content-Type: application/json ).
However, different APIs return data in different formats if requested. The following table is a reference of other
data formats APIs may support.
API DEFAULT DATA FORMAT OTHER FORMATS
/instance json text
/scheduledevents json none
To access a non-default response format, specify the requested format as a querystring parameter in the request.
For example:
curl -H Metadata:true "http://169.254.169.254/metadata/instance?api-version=2017-08-01&format=text"
Security
The Instance Metadata Service endpoint is accessible only from within the running virtual machine instance on a
non-routable IP address. In addition, any request with a X-Forwarded-For header is rejected by the service.
Requests must also contain a Metadata: true header to ensure that the actual request was directly intended and
not a part of unintentional redirection.
Error
If there is a data element not found or a malformed request, the Instance Metadata Service returns standard HTTP
errors. For example:
HTTP STATUS CODE REASON
200 OK
400 Bad Request Missing Metadata: true header
404 Not Found The requested element doesn't exist
405 Method Not Allowed Only GET and POST requests are supported
429 Too Many Requests The API currently supports a maximum of 5 queries per
second
500 Service Error Retry after some time
Examples
NOTE
All API responses are JSON strings. All following example responses are pretty-printed for readability.
Retrieving network information

Request
curl -H Metadata:true "http://169.254.169.254/metadata/instance/network?api-version=2017-08-01"
Response
NOTE
The response is a JSON string. The following example response is pretty-printed for readability.
{
"interface": [
{
"ipv4": {
"ipAddress": [
{
"publicIpAddress": "X.X.X.X"
}
],
"subnet": [
{
"address": "10.1.0.0",
"prefix": "24"
}
]
},
"ipv6": {
"ipAddress": []
},
"macAddress": "000D3AF806EC"
}
]
}
Retrieving public IP address
curl -H Metadata:true
"http://169.254.169.254/metadata/instance/network/interface/0/ipv4/ipAddress/0/publicIpAddress?api-
version=2017-08-01&format=text"
Retrieving all metadata for an instance

Request
curl -H Metadata:true "http://169.254.169.254/metadata/instance?api-version=2017-12-01"
Response
NOTE
{
"compute": {
"name": "avset2",
"offer": "UbuntuServer",
"osType": "Linux",
"placementGroupId": "",
"platformFaultDomain": "1",
"platformUpdateDomain": "1",
"publisher": "Canonical",
"resourceGroupName": "myrg",
"sku": "16.04-LTS",
"subscriptionId": "xxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxx",
"tags": "",
"version": "16.04.201708030",
"vmId": "13f56399-bd52-4150-9748-7190aae1ff21",
"vmScaleSetName": "",
"vmSize": "Standard_D1",
"zone": "1"
},
"network": {
"interface": [
{
"ipv4": {
"ipAddress": [
{
}
],
"subnet": [
{
"address": "10.1.2.0",
"prefix": "24"
}
]
},
"ipv6": {
"ipAddress": []
},
"macAddress": "000D3A36DDED"
}
]
}
}
Retrieving metadata in Windows Virtual Machine

Request
Instance metadata can be retrieved in Windows via the curl program:
curl -H @{'Metadata'='true'} http://169.254.169.254/metadata/instance?api-version=2017-08-01 | select -

ExpandProperty Content
Or through the Invoke-RestMethod PowerShell cmdlet:
Invoke-RestMethod -Headers @{"Metadata"="true"} -URI http://169.254.169.254/metadata/instance?api-

version=2017-08-01 -Method get
Response
NOTE
{
"compute": {
"name": "SQLTest",
"offer": "SQL2016SP1-WS2016",
"osType": "Windows",
"publisher": "MicrosoftSQLServer",
"sku": "Enterprise",
"version": "13.0.400110",
"vmId": "453945c8-3923-4366-b2d3-ea4c80e9b70e",
"vmSize": "Standard_DS2"
},
"network": {
"interface": [
{
"ipv4": {
"ipAddress": [
{
}
],
"subnet": [
{
"address": "10.0.1.0",
"prefix": "24"
}
]
},
"ipv6": {
"ipAddress": [
]
},
"macAddress": "002248020E1E"
}
]
}
}
Instance metadata data categories

The following data categories are available through the Instance Metadata Service:
DATA DESCRIPTION VERSION INTRODUCED
location Azure Region the VM is running in 2017-04-02
name Name of the VM 2017-04-02
offer Offer information for the VM image. 2017-04-02

This value is only present for images
deployed from Azure image gallery.
publisher Publisher of the VM image 2017-04-02
sku Specific SKU for the VM image 2017-04-02
version Version of the VM image 2017-04-02
osType Linux or Windows 2017-04-02
platformUpdateDomain Update domain the VM is running in 2017-04-02
platformFaultDomain Fault domain the VM is running in 2017-04-02
vmId Unique identifier for the VM 2017-04-02
vmSize VM size 2017-04-02
subscriptionId Azure subscription for the Virtual 2017-08-01

Machine
tags Tags for your Virtual Machine 2017-08-01
resourceGroupName Resource group for your Virtual 2017-08-01

Machine
placementGroupId Placement Group of your virtual 2017-08-01

machine scale set
plan Plan for a VM in its a Azure 2017-04-02

Marketplace Image, contains name,
product and publisher
publicKeys Collection of Public 2017-04-02

Keys[https://docs.microsoft.com/rest/api
/compute/virtualmachines/createorupda
te#sshpublickey] assigned to the VM
and paths
vmScaleSetName Virtual Machine ScaleSet Name of your 2017-12-01

virtual machine scale set
zone Availability Zone of your virtual 2017-12-01

machine
ipv4/privateIpAddress Local IPv4 address of the VM 2017-04-02
ipv4/publicIpAddress Public IPv4 address of the VM 2017-04-02
subnet/address Subnet address of the VM 2017-04-02
subnet/prefix Subnet prefix, example 24 2017-04-02
ipv6/ipAddress Local IPv6 address of the VM 2017-04-02

macAddress VM mac address 2017-04-02
scheduledevents See Scheduled Events 2017-08-01
identity (Preview) Managed identities for Azure 2018-02-01

resources. See acquire an access token
Example scenarios for usage

Tracking VM running on Azure
As a service provider, you may require to track the number of VMs running your software or have agents that need
to track uniqueness of the VM. To be able to get a unique ID for a VM, use the vmId field from Instance Metadata
Service.
Request
curl -H Metadata:true "http://169.254.169.254/metadata/instance/compute/vmId?api-version=2017-08-

01&format=text"
Response
5c08b38e-4d57-4c23-ac45-aca61037f084
Placement of containers, data-partitions based fault/update domain

For certain scenarios, placement of different data replicas is of prime importance. For example, HDFS replica
placement or container placement via an orchestrator may you require to know the platformFaultDomain and
platformUpdateDomain the VM is running on. You can also leverage Availability Zones for the instances to make
these decisions. You can query this data directly via the Instance Metadata Service.
Request
curl -H Metadata:true "http://169.254.169.254/metadata/instance/compute/platformFaultDomain?api-version=2017-

08-01&format=text"
Response
Getting more information about the VM during support case

As a service provider, you may get a support call where you would like to know more information about the VM.
Asking the customer to share the compute metadata can provide basic information for the support professional to
know about the kind of VM on Azure.
Request
curl -H Metadata:true "http://169.254.169.254/metadata/instance/compute?api-version=2017-08-01"
Response
NOTE
{
"compute": {
"location": "CentralUS",
"name": "IMDSCanary",
"offer": "RHEL",
"osType": "Linux",
"publisher": "RedHat",
"sku": "7.2",
"version": "7.2.20161026",
"vmId": "5c08b38e-4d57-4c23-ac45-aca61037f084",
"vmSize": "Standard_DS2"
}
}
Getting Azure Environment where the VM is running

Azure has various sovereign clouds like Azure Government. Sometimes you need the Azure Environment to make
some runtime decisions. The following sample will show you how you can achieve this.
Request
$metadataResponse = Invoke-WebRequest "http://169.254.169.254/metadata/instance/compute?api-version=2018-02-

01" -H @{"Metadata"="true"} -UseBasicParsing
$metadata = ConvertFrom-Json ($metadataResponse.Content)
$endpointsResponse = Invoke-WebRequest "https://management.azure.com/metadata/endpoints?api-version=2017-12-

01" -UseBasicParsing
$endpoints = ConvertFrom-Json ($endpointsResponse.Content)
foreach ($cloud in $endpoints.cloudEndpoint.PSObject.Properties) {

$matchingLocation = $cloud.Value.locations | Where-Object {$_ -match $metadata.location}
if ($matchingLocation) {
$cloudName = $cloud.name
break
}
}
$environment = "Unknown"
switch ($cloudName) {
"public" { $environment = "AzureCloud"}
"usGovCloud" { $environment = "AzureUSGovernment"}
"chinaCloud" { $environment = "AzureChinaCloud"}
"germanCloud" { $environment = "AzureGermanCloud"}
}
Write-Host $environment
Examples of calling metadata service using different languages inside the VM

LANGUAGE EXAMPLE
Ruby https://github.com/Microsoft/azureimds/blob/master/IMDSSa
mple.rb
LANGUAGE EXAMPLE
Go https://github.com/Microsoft/azureimds/blob/master/imdssam
ple.go
Python https://github.com/Microsoft/azureimds/blob/master/IMDSSa
mple.py
C++ https://github.com/Microsoft/azureimds/blob/master/IMDSSa
mple-windows.cpp
C# https://github.com/Microsoft/azureimds/blob/master/IMDSSa
mple.cs
JavaScript https://github.com/Microsoft/azureimds/blob/master/IMDSSa
mple.js
PowerShell https://github.com/Microsoft/azureimds/blob/master/IMDSSa
mple.ps1
Bash https://github.com/Microsoft/azureimds/blob/master/IMDSSa
mple.sh
Perl https://github.com/Microsoft/azureimds/blob/master/IMDSSa
mple.pl
Java https://github.com/Microsoft/azureimds/blob/master/imdssam
ple.java
Visual Basic https://github.com/Microsoft/azureimds/blob/master/IMDSSa

mple.vb
Puppet https://github.com/keirans/azuremetadata
FAQ
1. I am getting the error 400 Bad Request, Required metadata header not specified . What does this mean?
The Instance Metadata Service requires the header Metadata: true to be passed in the request. Passing
this header in the REST call allows access to the Instance Metadata Service.
2. Why am I not getting compute information for my VM?
Currently the Instance Metadata Service only supports instances created with Azure Resource Manager.
In the future, support for Cloud Service VMs might be added.
3. I created my Virtual Machine through Azure Resource Manager a while back. Why am I not see compute
metadata information?
For any VMs created after Sep 2016, add a Tag to start seeing compute metadata. For older VMs
(created before Sep 2016), add/remove extensions or data disks to the VM to refresh metadata.
4. I am not seeing all data populated for new version
For any VMs created after Sep 2016, add a Tag to start seeing compute metadata. For older VMs
(created before Sep 2016), add/remove extensions or data disks to the VM to refresh metadata.
5. Why am I getting the error 500 Internal Server Error ?
Retry your request based on exponential back off system. If the issue persists contact Azure support.
6. Where do I share additional questions/comments?
Send your comments on http://feedback.azure.com.
7. Would this work for Virtual Machine Scale Set Instance?
Yes Metadata service is available for Scale Set Instances.
8. How do I get support for the service?
To get support for the service, create a support issue in Azure portal for the VM where you are not able
to get metadata response after long retries
9. I get request timed out for my call to the service?
Metadata calls must be made from the primary IP address assigned to the network card of the VM , in
addition in case you have changed your routes there must be a route for 169.254.0.0/16 address out of
your network card.
10. I updated my tags in Virutal Machine Scale set but they dont appear in the instances unlike VMs?
Currently for ScaleSets tags only show to the VM on a reboot/reimage/or a disk change to the instance.
Next steps
Learn more about Scheduled Events
How to enable nested virtualization in an Azure VM
Nested virtualization is supported in several Azure virtual machine families. This capability provides great flexibility
in supporting scenarios such as development, testing, training, and demonstration environments.
This article steps through enabling Hyper-V on an Azure VM and configuring Internet connectivity to that guest
virtual machine.
Create a nesting capable Azure VM

Create a new Windows Server 2016 Azure VM. For quick refernce, all v3 virtual machines support nested
virtualization. For a complete list of virtual machine sizes that support nesting, check out the Azure Compute Unit
article.
Remember to choose a VM size large enough to support the demands of a guest virtual machine. In this example,
we are using a D3_v3 size Azure VM.
You can view the regional availability of Dv3 or Ev3 series virtual machines here.
NOTE
For detailed instructions on creating a new virtual machine, see Create and Manage Windows VMs with the Azure PowerShell
module
Connect to your Azure VM

Create a remote desktop connection to the virtual machine.
1. Click the Connect button on the virtual machine properties. A Remote Desktop Protocol file (.rdp file) is
created and downloaded.
2. To connect to your VM, open the downloaded RDP file. If prompted, click Connect. On a Mac, you need an
RDP client such as this Remote Desktop Client from the Mac App Store.
3. Enter the user name and password you specified when creating the virtual machine, then click Ok.
4. You may receive a certificate warning during the sign-in process. Click Yes or Continue to proceed with the
connection.
Enable the Hyper-V feature on the Azure VM

You can configure these settings manually or we have provided a PowerShell script to automate the configuration.
Option 1: Use a PowerShell script to configure nested virtualization
A PowerShell script to enable nested virtualization on a Windows Server 2016 host is available on GitHub. The
script checks pre-requisites and then configures nested virtualization on the Azure VM. A restart of the Azure VM
is necessary to complete the configuration. This script may work in other environments but is not guaranteed.
Check out the Azure blog post with a live video demonstration on nested virtualization running on Azure!
https://aka.ms/AzureNVblog.
Option 2: Configure nested virtualization manually
1. On the Azure VM, open PowerShell as an Administrator.
2. Enable the Hyper-V feature and Management Tools.
Install-WindowsFeature -Name Hyper-V -IncludeManagementTools -Restart
WARNING
This command restarts the Azure VM. You will lose your RDP connection during the restart process.
3. After the Azure VM restarts, reconnect to your VM using RDP.
Set up internet connectivity for the guest virtual machine

Create a new virtual network adapter for the guest virtual machine and configure a NAT Gateway to enable
Internet connectivity.
Create a NAT virtual network switch
2. Create an internal switch.
New-VMSwitch -Name "InternalNATSwitch" -SwitchType Internal
3. View the properties of the switch and note the ifIndex for the new adapter.
Get-NetAdapter
NOTE
Take note of the "ifIndex" for the virtual switch you just created.
4. Create an IP address for the NAT Gateway.

In order to configure the gateway, you need some information about your network:
IPAddress - The NAT Gateway IP specifies the IPv4 or IPv6 address to use as the default gateway address for
the virtual network subnet. The generic form is a.b.c.1 (for example, "192.168.0.1"). While the final position
doesn’t have to be .1, it usually is (based on prefix length). Typically you should use an RFC 1918 private
network address space.
PrefixLength - The subnet prefix length defines the local subnet size (subnet mask). The subnet prefix length will
be an integer value between 0 and 32. 0 would map the entire internet, 32 would only allow one mapped IP.
Common values range from 24 to 12 depending on how many IPs need to be attached to the NAT. A common
PrefixLength is 24 -- this is a subnet mask of 255.255.255.0.
InterfaceIndex - ifIndex is the interface index of the virtual switch created in the previous step.
New-NetIPAddress -IPAddress 192.168.0.1 -PrefixLength 24 -InterfaceIndex 13
Create the NAT network

In order to configure the gateway, you will need to provide information about the network and NAT Gateway:
Name - This is the name of the NAT network.
InternalIPInterfaceAddressPrefix - The NAT subnet prefix describes both the NAT Gateway IP prefix from above
as well as the NAT Subnet Prefix Length from above. The generic form will be a.b.c.0/NAT Subnet Prefix Length.
In PowerShell, create a new NAT network.
New-NetNat -Name "InternalNat" -InternalIPInterfaceAddressPrefix 192.168.0.0/24
Create the guest virtual machine

1. Open Hyper-V Manager and create a new virtual machine. Configure the virtual machine to use the new
Internal network you created.
2. Install an operating system on the guest virtual machine.
NOTE
You need installation media for an operating system to install on the VM. In this case we are using Windows 10
Enterprise.
Assign an IP address to the guest virtual machine

You can assign an IP address to the guest virtual machine either by manually setting a static IP address on the
guest virtual machine or configuring DHCP on the Azure VM to assign the IP address dynamically.
Option 1: Configure DHCP to dynamically assign an IP address to the guest virtual machine
Follow the steps below to configure DHCP on the host virtual machine for dynamic address assignment.
Install DCHP Server on the Azure VM
1. Open Server Manager. On the Dashboard, click Add roles and features. The Add Roles and Features
Wizard appears.
2. In wizard, click Next until the Server Roles page.
3. Click to select the DHCP Server checkbox, click Add Features, and then click Next until you complete the
wizard.
4. Click Install.
Configure a new DHCP scope
1. Open DHCP Manager.
2. In the navigation pane, expand the server name, right-click IPv4, and click New Scope. The New Scope
Wizard appears, click Next.
3. Enter a Name and Description for the scope and click Next.
4. Define an IP Range for your DCHP Server (for example, 192.168.0.100 to 192.168.0.200).
5. Click Next until the Default Gateway page. Enter the IP Address you created earlier (for example,
192.168.0.1) as the Default Gateway.
6. Click Next until the wizard completes, leaving all default values, then click Finish.
Option 2: Manually set a static IP address on the guest virtual machine
If you did not configure DHCP to dynamically assign an IP address to the guest virtual machine, follow these steps
to set a static IP address.
2. Right-click the guest virtual machine and click Connect.
3. Log on to the guest virtual machine.
4. On the guest virtual machine, open the Network and Sharing Center.
5. Configure the network adapter for an address within the range of the NAT network you created in the
previous section.
In this example you will use an address in the 192.168.0.0/24 range.
Test connectivity in guest virtual machine

In the guest virtual machine, open your browser and navigate to a web page.
Prepay for Virtual Machines with Azure Reserved VM
Instances
Prepay for virtual machines and save money with Azure Reserved Virtual Machine (VM ) Instances. For more
information, see Azure Reserved VM Instances offering.
You can buy a Reserved VM Instance in the Azure portal. To buy an instance:
You must be in an Owner role for at least one Enterprise or Pay-As-You-Go subscription.
For Enterprise subscriptions, reservation purchases must be enabled in the EA portal.
For the Cloud Solution Provider (CSP ) program, only the admin agents or sales agents can buy reservations.
Determine the right VM size before you buy

The Meter Sub-category and Product fields in the usage data don't distinguish between VM sizes that use
premium storage from VMs that don't. If you use these fields to determine the VM size to use for the reservation,
you may buy the wrong size and won't get the reservation discount you expect. Use one of the following methods
to determine the right VM size when you buy the reservation:
Refer to the AdditionalInfo field in your usage file or usage API to determine the correct VM size. Don't use the
values from Meter Sub-category or Product fields. These fields don't differentiate between S and Non-S
versions of a VM.
Get accurate VM size information using Powershell, Azure Resource Manager, or from the VM details in the
Azure portal.
Reserved VM Instances are available for most VM sizes with some exceptions:
Reservation discount doesn't apply for the following VMs:
Classic VMs and Cloud services
Constrained vCPU sizes
VM Series: A-series, Av2-series, or G -series
VMs in Preview: Any VM -series or size that is in preview
Clouds: Reservations aren't available for purchase in Germany or China regions.
Insufficient quota: A reservation that is scoped to a single subscription must have vCPU quota available in the
subscription for the new RI. For example, if the target subscription has a quota limit of 10 vCPUs for D -Series,
then you can't buy a reservation for 11 Standard_D1 instances. The quota check for reservations includes the
VMs already deployed in the subscription. For example, if the subscription has a quota of 10 vCPUs for D -
Series and has two standard_D1 instances deployed, then you can buy a reservation for 10 standard_D1
instances in this subscription.
Capacity restrictions: In rare circumstances, Azure limits the purchase of new reservations for subset of VM
sizes, due to low capacity in a region.
Buy a Reserved VM Instance

2. Select All services > Reservations.
3. Select Add to purchase a new reservation.
4. Fill in the required fields. Running VM instances that match the attributes you select qualify to get the
reservation discount. The actual number of your VM instances that get the discount depend on the scope
and quantity selected.
FIELD DESCRIPTION
Name The name of this reservation.
Subscription The subscription used to pay for the reservation. The

payment method on the subscription is charged the
upfront costs for the reservation. The subscription type
must be an enterprise agreement (offer number: MS-AZR-
0017P) or Pay-As-You-Go (offer number: MS-AZR-
0003P). For an enterprise subscription, the charges are
deducted from the enrollment's monetary commitment
balance or charged as overage. For Pay-As-You-Go
subscription, the charges are billed to the credit card or
invoice payment method on the subscription.
Scope The reservation’s scope can cover one subscription or

multiple subscriptions (shared scope). If you select:
Single subscription - The reservation discount is
applied to VMs in this subscription.
Shared - The reservation discount is applied to
VMs running in any subscriptions within your
billing context. For enterprise customers, the
shared scope is the enrollment and includes all
subscriptions (except dev/test subscriptions) within
the enrollment. For Pay-As-You-Go customers, the
shared scope is all Pay-As-You-Go subscriptions
created by the account administrator.
Region The Azure region that’s covered by the reservation.
VM Size The size of the VM instances.
Optimize for VM instance size flexibility applies the reservation discount

to other VMs in the same VM size group. Capacity priority
prioritizes data center capacity for your deployments. This
offers additional confidence in your ability to launch the
VM instances when you need them. Capacity priority is
only available when the reservation scope is single
subscription.
Term One year or three years.
Quantity The number of instances being purchased within the

reservation. The quantity is the number of running VM
instances that can get the billing discount. For example, if
you are running 10 Standard_D2 VMs in the East US, then
you would specify quantity as 10 to maximize the benefit
for all running machines.
5. You can view the cost of the reservation when you select Calculate cost.
6. Select Purchase.
7. Select View this Reservation to see the status of your purchase.
Cancellations and exchanges

If you need to cancel your reservation, there may be a 12% early termination fee. Refunds are based on the lowest
price of either your purchase price or the current price of the reservation. Refunds are limited to $50,000 per year.
The refund you receive is the remaining pro-rated balance minus the 12% early termination fee. To request a
cancellation, go to the reservation in the Azure portal and select Refund to create a support request.
If you need to change your Reserved VM Instances reservation to another region, VM size group, or term, you can
exchange it for another reservation that's of equal or greater value. The term start date for the new reservation
doesn't carry over from the exchanged reservation. The 1 or 3 year term starts from when you create the new
reservation. To request an exchange, go to the reservation in the Azure portal, and select Exchange to create a
support request.
Next steps
The reservation discount is applied automatically to the number of running virtual machines that match the
reservation scope and attributes. You can update the scope of the reservation through Azure portal, PowerShell,
CLI, or through the API.
To learn how to manage a reservation, see Manage Azure Reservations.
To learn more about Azure Reservations, see the following articles:
What are Azure Reservations?
Manage Reservations in Azure
Understand how the reservation discount is applied
Understand reservation usage for your Pay-As-You-Go subscription
Understand reservation usage for your Enterprise enrollment
Windows software costs not included with reservations
Azure Reservations in Partner Center Cloud Solution Provider (CSP ) program
Need help? Contact support

If you still have further questions, contact support to get your issue resolved quickly.
What are Azure Reservations?
Azure Reservations helps you save money by pre-paying for one-year or three-years of virtual machine, SQL
Database compute capacity, Azure Cosmos DB throughput, or other Azure resources. Pre-paying allows you to get
a discount on the resources you use. Reservations can significantly reduce your virtual machine, SQL database
compute, Azure Cosmos DB, or other resource costs up to 72% on pay-as-you-go prices. Reservations provide a
billing discount and don't affect the runtime state of your resources.
You can buy a reservation in the Azure portal. For more information, see the following topics:
Prepay for Virtual Machines with Azure Reserved VM Instances
Prepay for SQL Database compute resources with Azure SQL Database reserved capacity
Prepay for Azure Cosmos DB resources with Azure Cosmos DB reserved capacity
Why should I buy a reservation?

If you have virtual machines, Azure Cosmos DB or SQL databases that run for long periods of time, purchasing a
reservation gives you the most cost-effective option. For example, if you continuously run four instances of a
service without a reservation, you are charged at pay-as-you-go rates. If you purchase a reservation for those
resources, you immediately get the reservation discount. The resources are no longer charged at the pay-as-you-
go rates.
What charges does a reservation cover?

Reserved Virtual Machine Instance: A reservation only covers the virtual machine compute costs. It doesn't
cover additional software, networking, or storage charges.
SQL Database reserved vCore: Only the compute costs are included with a reservation. The license is billed
separately.
Azure Cosmos DB reserved capacity: A reservation covers throughput provisioned for your resources, it does
not cover the storage and networking charges.
For Windows virtual machines and SQL Database, you can cover the licensing costs with Azure Hybrid Benefit.
Who’s eligible to purchase a reservation?

Azure customers with these subscriptions types can buy a reservation:
Enterprise agreement subscription offer type (MS -AZR -0017P ).
Pay-As-You-Go subscription offer type (MS -AZR -003P ). You must have the “Owner” role on the subscription to
buy a reservation.
Cloud Solution Provider (CSP ) partners can use Azure portal or Partner Center to purchase Azure
Reservations.
A reservation discount only applies to resources associated with Enterprise, Pay-As-You-Go, or CSP subscription
types.
How is a reservation billed?

The reservation is charged to the payment method tied to the subscription. If you have an Enterprise subscription,
the reservation cost is deducted from your monetary commitment balance. If your monetary commitment balance
doesn’t cover the cost of the reservation, you’re billed the overage. If you have a Pay-As-You-Go subscription, the
credit card you have on your account is billed immediately. If you’re billed by invoice, you see the charges on your
next invoice.
How is the reservation discount applied?

The reservation discount applies to the resource usage that matches the attributes you select when you buy the
reservation. The attributes include the scope where the matching VMs, SQL databases, Azure Cosmos DB, or other
resources run. For example, if you want a reservation discount for four Standard D2 virtual machines in the West
US region, select the subscription where the VMs are running. If the virtual machines are running in different
subscriptions within your enrollment/account, then select the scope as shared. Shared scope allows the reservation
discount to be applied across subscriptions. You can change the scope after you buy a reservation. For more
information, see Manage Azure Reservations.
A reservation discount only applies to resources associated with Enterprise, Pay-As-You-Go, or CSP subscription
types. Resources that run in a subscription with other offer types don't receive the reservation discount. For
enterprise enrollments, enterprise Dev/Test subscriptions aren’t eligible for the reservation benefits.
To better understand how Reservations affects your billing, see the following topics:
Understand Azure Reserved VM Instances discount
Understand Azure reservation discount
Understand Azure Cosmos DB reservation discount
Understand Azure reservation discount and usage for SUSE
What happens when the reservation term expires?

At the end of the reservation term, the billing discount expires, and the virtual machine, SQL database, Azure
Cosmos DB, or other resource is billed at the pay-as-you go price. Azure Reservations don't auto-renew. To
continue getting the billing discount, you must buy a new reservation for eligible services and software.
Discount applies to different sizes with instance size flexibility

When you buy a reservation, the discount can apply to other instances with attributes that are within the same size
group. The flexibility of the discount coverage depends on the type of reservation and the attributes you pick when
you buy the reservation.
Reserved VM Instances: When you buy the reservation, if you select Optimized for: instance size flexibility,
the discount coverage depends on the VM size you select. The reservation can apply to the virtual machines
(VMs) sizes in the same size series group. For more information, see Virtual machine size flexibility with
Reserved VM Instances.
SUSE Linux Enterprise software plan: The discount coverage depends on the vCPUs of the VMs where you run
the SUSE software. For more information, see Understand how the SUSE Linux Enterprise software plan
discount is applied.
SQL Database reserved capacity: The discount coverage depends on the performance tier you pick. For more
information, see Understand how an Azure reservation discount is applied.
Azure Cosmos DB reserved capacity: The discount coverage depends on the provisioned throughput. For more
information, see Understand how an Azure Cosmos DB reservation discount is applied.
Next steps
Start saving on your virtual machines by purchasing a Reserved VM Instance, SQL Database reserved capacity, or
Azure Cosmos DB reserved capacity.
To learn more about Azure Reservations, see the following articles:
Manage Azure Reservations
Understand reservation usage for your Pay-As-You-Go subscription
Understand reservation usage for your Enterprise enrollment
Windows software costs not included with reservations
Azure Reservations in Partner Center Cloud Solution Provider (CSP ) program
Need help? Contact support

If you still have further questions, contact support to get your issue resolved quickly.
Virtual machine size flexibility with Reserved VM
Instances
With a reserved virtual machine instance that's optimized for instance size flexibility, the reservation you buy can
apply to the virtual machines (VMs) sizes in the same size series group. For example, if you buy a reservation for a
VM size that's listed in the DSv2-series table, like Standard_DS5_v2, the reservation discount can apply to the other
four sizes that are listed in that same table:
Standard_DS1_v2
Standard_DS2_v2
Standard_DS3_v2
Standard_DS4_v2
But that reservation discount doesn't apply to VMs sizes that are listed in different tables, like what's in the DSv2-
series high memory table: Standard_DS11_v2, Standard_DS12_v2, and so on.
Within the size series group, the number of VMs the reservation discount applies to depends on the VM size you
pick when you buy a reservation. It also depends on the sizes of the VMs that you have running. The ratio column
that's listed in the following tables compares the relative footprint for each VM size in that group. Use the ratio
value to calculate how the reservation discount applies to the VMs you have running.
Examples
The following examples use the sizes and ratios in the DSv2-series table.
You buy a reserved VM instance with the size Standard_DS4_v2 where the ratio or relative footprint compared to
the other sizes in that series is 8.
Scenario 1: Run eight Standard_DS1_v2 sized VMs with a ratio of 1. Your reservation discount applies to all
eight of those VMs.
Scenario 2: Run two Standard_DS2_v2 sized VMs with a ratio of 2 each. Also run a Standard_DS3_v2 sized VM
with a ratio of 4. The total footprint is 2+2+4=8. So your reservation discount applies to all three of those VMs.
Scenario 3: Run one Standard_DS5_v2 with a ratio of 16. Your reservation discount applies to half that VM's
compute cost.
The following sections show what sizes are in the same size series group when you buy a reserved VM instance
optimized for instance size flexibility.
B-series
SIZE RATIO
Standard_B1s 1
Standard_B2s 4
For more information, see B -series burstable virtual machine sizes.

B-series high memory
SIZE RATIO
Standard_B1ms 1
Standard_B2ms 4
Standard_B4ms 8
Standard_B8ms 16
For more information, see B -series burstable virtual machine sizes.
D-series
SIZE RATIO
Standard_D1 1
Standard_D2 2
Standard_D3 4
Standard_D4 8
For more information, see Previous generations of virtual machine sizes.
D-series high memory

SIZE RATIO
Standard_D11 1
Standard_D12 2
Standard_D13 4
Standard_D14 8
Ds-series
SIZE RATIO
Standard_DS1 1
Standard_DS2 2
Standard_DS3 4
SIZE RATIO
Standard_DS4 8
Ds-series high memory

SIZE RATIO
Standard_DS11 1
Standard_DS12 2
Standard_DS13 4
Standard_DS14 8
DSv2-series
SIZE RATIO
Standard_DS1_v2 1
Standard_DS2_v2 2
Standard_DS3_v2 4
Standard_DS4_v2 8
Standard_DS5_v2 16
For more information, see General purpose virtual machine sizes.
DSv2-series high memory

SIZE RATIO
Standard_DS11_v2 1
Standard_DS12_v2 2
Standard_DS13_v2 4
Standard_DS14_v2 8
Standard_DS15_v2 10
For more information, see Memory optimized virtual machine sizes.

DSv3-series
SIZE RATIO
Standard_D2s_v3 1
Standard_D4s_v3 2
Standard_D8s_v3 4
Standard_D16s_v3 8
Standard_D32s_v3 16
Standard_D64s_v3 32
Dv2-series
SIZE RATIO
Standard_D1_v2 1
Standard_D2_v2 2
Standard_D3_v2 4
Standard_D4_v2 8
Standard_D5_v2 16
Dv2-series high memory

SIZE RATIO
Standard_D11_v2 1
Standard_D12_v2 2
Standard_D13_v2 4
Standard_D14_v2 8
Standard_D15_v2 10
Dv3-series
SIZE RATIO
Standard_D2_v3 1
Standard_D4_v3 2
Standard_D8_v3 4
Standard_D16_v3 8
Standard_D32_v3 16
Standard_D64_v3 32
ESv3-series
SIZE RATIO
Standard_E2s_v3 1
Standard_E4s_v3 2
Standard_E8s_v3 4
Standard_E16s_v3 8
Standard_E32s_v3 16
Standard_E64s_v3 32
Ev3-series
SIZE RATIO
Standard_E2_v3 1
Standard_E4_v3 2
Standard_E8_v3 4
Standard_E16_v3 8
Standard_E32_v3 16
Standard_E64_v3 32

F-series
SIZE RATIO
Standard_F1 1
Standard_F2 2
Standard_F4 4
Standard_F8 8
Standard_F16 16
For more information, see Compute optimized virtual machine sizes.
FS-series
SIZE RATIO
Standard_F1s 1
Standard_F2s 2
Standard_F4s 4
Standard_F8s 8
Standard_F16s 16
Fsv2-series
SIZE RATIO
Standard_F2s_v2 1
Standard_F4s_v2 2
Standard_F8s_v2 4
Standard_F16s_v2 8
Standard_F32s_v2 16
Standard_F64s_v2 32
Standard_F72s_v2 36

H-series
SIZE RATIO
Standard_H8 1
Standard_H16 2
For more information, see High performance compute VM sizes.
H-series high memory

SIZE RATIO
Standard_H8m 1
Standard_H16m 2
For more information, see High performance compute VM sizes.
Ls-series
SIZE RATIO
Standard_L4s 1
Standard_L8s 2
Standard_L16s 4
Standard_L32s 8
For more information, see Storage optimized virtual machine sizes.
M-series
SIZE RATIO
Standard_M64s 1
Standard_M128s 2
M-series fractional
SIZE RATIO
Standard_M16s 1
Standard_M32s 2
M-series fractional high memory

SIZE RATIO
Standard_M8ms 1
Standard_M16ms 2
Standard_M32ms 4
M-series fractional large

SIZE RATIO
Standard_M32ls 1
Standard_M64ls 2
M-series high memory

SIZE RATIO
Standard_M64ms 1
Standard_M128ms 2
NC-series
SIZE RATIO
Standard_NC6 1
Standard_NC12 2
Standard_NC24 4
For more information, see GPU optimized virtual machine sizes.
NCv2-series
SIZE RATIO
Standard_NC6s_v2 1
SIZE RATIO
Standard_NC12s_v2 2
Standard_NC24s_v2 4
NCv3-series
SIZE RATIO
Standard_NC6s_v3 1
Standard_NC12s_v3 2
Standard_NC24s_v3 4
ND-series
SIZE RATIO
Standard_ND6s 1
Standard_ND12s 2
Standard_ND24s 4
NV-series
SIZE RATIO
Standard_NV6 1
Standard_NV12 2
Standard_NV24 4

Shared Image Gallery overview
Shared Image Gallery is a service that helps you build structure and organization around your custom VM images.
Shared Image Gallery provides three main value propositions
Simple management
Scale your customer images
Share your images - share your images to different users, service principals, or AD groups within your
organization as well as different regions using the multi-region replication
A managed image is a copy of either a full VM (including any attached data disks) or just the OS disk, depending
on how you create the image. When you create a VM from the image, the copy of the VHDs in the image are used
to create the disks for the new VM. The managed image remains in storage and can be used over and over again to
create new VMs.
If you have a large number of managed images that you need to maintain and would like to make them available
throughout your company, you can use a shared image gallery as a repository that makes it easy to update and
share your images. The charges for using a shared image gallery are just the costs for the storage used by the
images, plus any network egress costs for replicating images from the source region to the published regions.
The Shared Image Gallery feature has multiple resource types:
RESOURCE DESCRIPTION
Managed image This is a baseline image that can be used alone or used to
create multiple shared image versions in an image gallery.
Image gallery Like the Azure Marketplace, an image gallery is a repository

for managing and sharing images, but you control who has
access.
Gallery image Images are defined within a gallery and carry information
about the image and requirements for using it internally. This
includes whether the image is Windows or Linux, release
notes, and minimum and maximum memory requirements.
This type of image is a resource within the Resource Manager
deployment model, but it isn't used directly for creating VMs.
It is a definition of a type of image.
Shared image version An image version is what you use to create a VM when
using a gallery. You can have multiple versions of an image as
needed for your environment. Like a managed image, when
you use an image version to create a VM, the image version
is used to create new disks for the VM. Image versions can be
used multiple times.
Regional Support
Regional support for shared image galleries is limited, but will expand over time. For preview, here are the lists of
where you can create galleries and regions where can replicate any gallery:
CREATE GALLERY IN REPLICATE VERSION TO
West Central US South Central US
East US 2 East US
South Central US East US 2
Southeast Asia West US
West Europe West US 2
Central US
North Central US
Canada Central
Canada East
North Europe
West Europe
South India
Southeast Asia
Scaling
Shared Image Gallery allows you to specify the number of replicas you want Azure to keep for the images. This
helps in multi-VM deployment scenarios as the VM deployments can be spread to different replicas reducing the
chance of instance creation process being throttled due to overloading of a single replica.
Replication
Shared Image Gallery also allows you to replicate your images to other Azure regions automatically. Each Shared
image version can be replicated to different regions depending on what makes sense for your organization. One
example is to always replicate the latest image in multi-regions while all older versions are only available in 1
region. This can help save on storage costs for Shared image versions. The regions a Shared image version is
replicated to can be updated after creation time. The time it takes to replicate to different regions depends on the
amount of data being copied and the number of regions the version is replicate to. This can take few hours in some
cases. While the replication is happening, you can view the status of replication per region. Once the image
replication is complete in a region, you can then deploy a VM or VMSS using that version in the region.
Access
As the shared image gallery, shared image and shared image version are all resources, they can be shared using
the built-in native Azure RBAC controls. Using RBAC you can share these resources to other users, service
principals, and groups in your organization. The scope of sharing these resources is within the same AD tenant.
Once a user has access to the Shared image version, they can deploy a VM or a virtual machine scale set in any of
the subscriptions they have access to within the same AD tenant as the Shared image version. Here is the sharing
matrix that helps understand what the user gets access to:
SHARED WITH USER SHARED IMAGE GALLERY SHARED IMAGE SHARED IMAGE VERSION
Shared Image Gallery Yes Yes Yes
Shared Image No Yes Yes
Shared image version No No Yes
Billing
There is no extra charge for using the Shared Image Gallery service. You will be charged for the following
resources:
Storage costs of storing the Shared image versions. This depends on the number of replicas of the version and
the number of regions the version is replicated to.
Network egress charges for replication from source region of the version to the replicated regions.
Frequently asked questions

Q. How do I sign up for the Shared Image Gallery Public Preview?
A. In order to sign up for the Shared Image Gallery public preview, you need to register for the feature by running
the following commands from each of the subscriptions in which you intend to create a shared image gallery,
image definition, or image version resources, and also where you intend to deploy Virtual Machines using the
image versions.
CLI:
az feature register --namespace Microsoft.Compute --name GalleryPreview

az provider register -n Microsoft.Compute
PowerShell:
Register-AzureRmProviderFeature -FeatureName GalleryPreview -ProviderNamespace Microsoft.Compute

Register-AzureRmResourceProvider -ProviderNamespace Microsoft.Compute
Q. How can I list all the Shared Image Gallery resources across subscriptions?
A. In order to list all the Shared Image Gallery resources across subscriptions that you have access to on the Azure
portal, follow the steps below:
2. Go to All Resources.
3. Select all the subscriptions under which you’d like to list all the resources.
4. Look for resources of type Private gallery.
To see the image definitions and image versions, you should also select Show hidden types.
To list all the shared image gallery resources across subscriptions that you have permissions to, use the
following command in the Azure CLI:
az account list -otsv --query "[].id" | xargs -n 1 az sig list --subscription
Q. How do I share my images across subscriptions?

A. You can share images across subscriptions using Role Based Access Control (RBAC ). Any user that has read
permissions to an image version, even across subscriptions, will be able to deploy a Virtual Machine using the
image version.
Q. Can I move my existing image to the shared image gallery?
A. Yes. There are 3 scenarios based on the types of images you may have.
Scenario 1: If you have a managed image, then you can create an image definition and image version from it.
Scenario 2: If you have an unmanaged generalized image, you can create a managed image from it, and then
create an image definition and image version from it.
Scenario 3: If you have a VHD in your local file system, then you need to upload the VHD, create a managed
image, then you can create and image definition and image version from it.
If the VHD is of a Windows VM, see Upload a generalized VHD.
If the VHD is for a Linux VM, see Upload a VHD
Q. Can I create an image version from a specialized disk?
A. No, we do not currently support specialized disks as images. If you have a specialized disk, you need to create a
VM from the VHD by attaching the specializeddisk to a new VM. Once you have a running VM, you need to follow
the instructions to create a managed image from the Windows VM or Linux VM. Once you have a generalized
managed image, you can start the process to create a shared image description and image version.
Q. Can I create a shared image gallery, image definition, and image version through the Azure portal?
A. No, currently we do not support the creation of any of the Shared Image Gallery resources through Azure
portal. However, we do support the creation of the Shared Image Gallery resources through CLI, Templates, and
SDKs. PowerShell will also be release soon.
Q. Once created, can I update the image definition or the image version? What kind of details can I modify?
A. The details that can be updated on each of the resources are mentioned below:
Shared image gallery:
Description
image definition:
Recommended vCPUs
Memory
Description
End of life date
Image version:
Regional replica count
Target regions
Exclusion from latest
End of life date
Q. Once created, can I move the Shared Image Gallery resource to a different subscription?
A. No, you cannot move the shared image gallery resource to a different subscription. However, you will be able to
replicate the image versions in the gallery to other regions as required.
Q. Can I replicate my image versions across clouds – Azure China 21Vianet, Azure Germany and Azure
Government Cloud?
A. No, you cannot replicate image versions across clouds.
Q. Can I replicate my image versions across subscriptions?
A. No, you may replicate the image versions across regions in a subscription and use it in other subscriptions
through RBAC.
Q. Can I share image versions across AD tenants?
A. No, currently shared image gallery does not support the sharing of image versions across AD tenants. However,
you may use the Private Offers feature on Azure Marketplace to achieve this.
Q. How long does it take to replicate image versions across the target regions?
A. The image version replication time is entirely dependent on the size of the image and the number of regions it is
being replicated to. However, as a best practice, it is recommended that you keep the image small, and the source
and target regions close for best results. You can check the status of the replication using the -ReplicationStatus
flag.
Q. How many shared image galleries can I create in a subscription?
A. The default quota is
10 shared image galleries, per subscription, per region
200 image definitions, per subscription, per region
2000 image versions, per subscription, per region
Q. What is the difference between source region and target region?
A. Source region is the region in which your image version will be created, and target regions are the regions in
which a copy of your image version will be stored. For each image version, you can only have one source region.
Also, make sure that you pass the source region location as one of the target regions when you create an image
version.
Q. How do I specify the source region while creating the image version?
A. While creating an image version, you can use the --location tag in CLI and the -Location tag in PowerShell to
specify the source region. Please ensure the managed image that you are using as the base image to create the
image version is in the same location as the location in which you intend to create the image version. Also, make
sure that you pass the source region location as one of the target regions when you create an image version.
Q. How do I specify the number of image version replicas to be created in each region?
A. There are two ways you can specify the number of image version replicas to be created in each region:
1. The regional replica count which specifies the number of replicas you want to create per region.
2. The common replica count which is the default per region count in case regional replica count is not specified.
To specify the regional replica count, pass the location along with the number of replicas you want to create in that
region like this: “South Central US=2”.
If regional replica count is not specified with each location, then the default number of replicas will be the common
replica count that you specified.
To specify the common replica count in CLI, use the --replica-count argument in the az sig image-version create
command.
Q. Can I create the shared image gallery in a different location than the one where I want to create the image
definition and image version?
A. Yes, this is possible. But as a best practice, we encourage you to keep the resource group, shared image gallery,
image definition and image version in the same location.
Q. What are the charges for using the Shared Image Gallery?
A. There are no charges for using the Shared Image Gallery service, except the storage charges for storing the
image versions and network egress charges for replicating the image versions from source region to target
regions.
Q. What API version should I use to create Shared Image Gallery, Image Definition, Image Version, and
VM/VMSS out of the Image Version?
A. For VM and virtual machine scale set deployments using an image version, we recommend you use API version
2018-04-01 or higher. To work with shared image galleries, image definitions, and image versions, we recommend
you use API version 2018-06-01.
Next steps
Learn how to deploy shared images using Azure CLI.
Preview: Create a shared image gallery with the
Azure CLI
The Shared Image Gallery greatly simplifies custom image sharing across your organization. Custom images are
like marketplace images, but you create them yourself. Custom images can be used to bootstrap configurations
such as preloading applications, application configurations, and other OS configurations. The Shared Image
Gallery lets you share your custom VM images with others in your organization, within or across regions, within an
AAD tenant. Choose which images you want to share, which regions you want to make them available in, and who
you want to share them with. You can create multiple galleries so that you can logically group shared images. The
gallery is a top-level resource that provides full role-based access control (RBAC ). Images can be versioned, and
you can choose to replicate each image version to a different set of Azure regions. The gallery only works with
Managed Images.
In this article, you create your own gallery and custom images of an Azure virtual machine. You learn how to:
Create a shared image gallery
Create a shared image definition
Create a shared image version
Create a VM from a shared image
Delete a resources
Open Azure Cloud Shell

Azure Cloud Shell is a free, interactive shell that you can use to run the steps in this article. Common Azure tools
are preinstalled and configured in Cloud Shell for you to use with your account. Just select the Copy button to
copy the code, paste it in Cloud Shell, and then press Enter to run it. There are a few ways to open Cloud Shell:
Select Try It in the upper-right corner of a code block.
Select the Cloud Shell button on the menu in the upper-right

corner of the Azure portal.
If you choose to install and use the CLI locally, this article requires that you are running the Azure CLI version
2.0.46 or later. Run az --version to find the version. If you need to install or upgrade, see Install Azure CLI 2.0.
Before you begin

To complete the example in this article, you must have an existing managed image of a generalized VM. For more
information, see Tutorial: Create a custom image of an Azure VM with the Azure CLI 2.0.
Preview: Register the feature

Shared Image Galleries is in preview, but you need to register the feature before you can use it. To register the
Shared Image Galleries feature:
az feature register --namespace Microsoft.Compute --name GalleryPreview

az provider register -n Microsoft.Compute
It might take a few minutes to register the feature. You can check the progress using:
az provider show -n Microsoft.Compute
Create an image gallery

An image gallery is the primary resource used for enabling image sharing. Gallery names must be unique within
your subscription. Create an image gallery using az sig create. The following example creates a gallery named
myGallery in myGalleryRG.
az group create --name myGalleryRG --location WestCentralUS

az sig create -g myGalleryRG --gallery-name myGallery
Create an image definition

Create an initial image definition in the gallery using az sig image-definition create.
az sig image-definition create \

-g myGalleryRG \
--gallery-name myGallery \
--gallery-image-definition myImageDefinition \
--publisher myPublisher \
--offer myOffer \
--sku 16.04-LTS \
--os-type Linux
Create an image version

Create versions of the image as needed using az image gallery create-image-version. You will need to pass in the
ID of the managed image to use as a baseline for creating the image version. You can use az image list to get
information about images that are in a resource group. In this example, the version of our image is 1.0.0 and we
are going to create 5 total replicas in the West Central US, South Central US and East US 2* regions.
az sig image-version create \

-g myGalleryRG \
--gallery-name myGallery \
--gallery-image-definition myImageDefinition \
--gallery-image-version 1.0.0 \
--target-regions "WestCentralUS" "SouthCentralUS=1" "EastUS2=1" \
--replica-count 5 \
--managed-image "/subscriptions/<subscription
ID>/resourceGroups/myResourceGroup/providers/Microsoft.Compute/images/myImage"
Create a VM
Create a VM from the image version using az vm create.
az vm create\
-g myGalleryRG \
-n myVM \
--image "/subscriptions/<subscription-
ID>/resourceGroups/myGalleryRG/providers/Microsoft.Compute/galleries/myGallery/images/myImageDefinition/versio
ns/1.0.0" \
--generate-ssh-keys
Using RBAC to share images

You can share images across subscriptions using Role Based Access Control (RBAC ). Any user that has read
permissions to an image version, even across subscriptions, will be able to deploy a Virtual Machine using the
image version.
For more information about how to share resources using RBAC, see Manage access using RBAC and Azure CLI.
List information
Get the location, status and other information about the available image galleries using az sig list.
az sig list -o table
List the image definitions in a gallery, including information about OS type and status, using az sig image-
definition list.
az sig image-definition list -g myGalleryRG -r myGallery -o table
List the shared image versions in a gallery, using az sig image-version list.
az sig image-version list -g myGalleryRG -r myGallery -i myGalleryImage -o table
Get the ID of an image version using az sig image-version show.
az sig image-version show \

-g myGalleryRG \
-r myGallery \
-i myGalleryImage \
--gallery-image-version-name 1.0.0 \
--query "id"
Next steps
For more information about Shared Image Galleries, see the Overview.
Troubleshooting shared image galleries
If you run into issues while performing any operations on shared image galleries, image definitions, and image
versions, run the failing command again in debug mode. Debug mode is activated by passing the -debug switch
with CLI and the -Debug switch with PowerShell. Once you’ve located the error, follow this document to
troubleshoot the errors.
Unable to create a shared image gallery

Possible causes:
The gallery name is invalid.
Allowed characters for Gallery name are uppercase or lowercase letters, digits, dots, and periods. The gallery name
cannot have dashes in it. Change the gallery name and try again.
The gallery name is not unique within your subscription.
Pick another gallery name and try again.
Unable to create an image definition

Possible causes:
image definition name is invalid.
Allowed characters for image definition are uppercase or lowercase letters, digits, dots, dashes and periods. Change
the image definition name and try again.
The mandatory properties for creating an image definition are not populated.
The properties such as name, publisher, offer, sku, and OS type are mandatory. Verify if all the properties are being
passed.
Make sure that the OSType, either Linux or Windows, of the image definition is the same as the source managed
image that you are using to create the image version.
Unable to create an image version

Possible causes:
Image version name is invalid.
Allowed characters for image version are numbers and periods. Numbers must be within the range of a 32-bit
integer. Format: MajorVersion.MinorVersion.Patch. Change the image version name and try again.
Source managed image from which the image version is being created is not found.
Check if the source image exists and is in the same region as the image version.
The managed image isn't done being provisioned.
Make sure the provisioning state of the source managed image is Succeeded.
The source region is not supported yet.
Use the table below to see if the intended source region is supported:
CREATE GALLERY IN OR “SOURCE REGION” REPLICATE RERSION TO OR “TARGET REGION”
West Central US All Azure Public Cloud regions
South Central US
East US 2
Southeast Asia
West Europe
The target region list must include the source region of the image version. Make sure you have included the source
region in the list of target regions where you want Azure to replicate your image version to.
Replication to all the target regions not completed.
Use the --expand ReplicationStatus flag to check if the replication to all the specified target regions has been
completed. If not, wait up to 6 hours for the job to complete. If it fails, run the command again to create and
replicate the image version. If there are a lot of target regions the image version is being replicated to, consider
doing the replication in phases.
Unable to create a VM or a scale set

Possible causes:
The user trying to create a VM or virtual machine scale set doesn’t have the read access to the image version.
Contact the subscription owner and ask them to give read access to the image version or the parent resources (like
the shared image gallery or image definition) through Role Based Access Control (RBAC ).
The image version is not found.
Verify that the region you are trying to create a VM or virtual machine scale in is included in the list of target
regions of the image version. If the region is already in the list of target regions, then verify if the replication job has
been completed. You can use the -ReplicationStatus flag to check if the replication to all the specified target
regions has been completed.
The VM or virtual machine scale set creation takes a long time.
Verify that the OSType of the image version that you are trying to create the VM or virtual machine scale set from
has the same OSType of the source managed image that you used to create the image version.
Unable to share resources

The sharing of shared image gallery, image definition, and image version resources across subscriptions is enabled
using Role Based Access Control (RBAC ).
Replication is slow
Use the --expand ReplicationStatus flag to check if the replication to all the specified target regions has been
completed. If not, wait for up to 6 hours for the job to complete. If it fails, trigger the command again to create and
replicate the image version. If there are a lot of target regions the image version is being replicated to, consider
doing the replication in phases.
Azure limits and quotas

Azure limits and quotas apply to all shared image gallery, image definition, and image version resources. Make
sure you are within the limits for your subscriptions.
Next steps
Learn more about shared image galleries.
Find Windows VM images in the Azure Marketplace
with Azure PowerShell
This article describes how to use Azure PowerShell to find VM images in the Azure Marketplace. You can then
specify a Marketplace image when you create a VM programmatically with PowerShell, Resource Manager
templates, or other tools.
You can also browse available images and offers using the Azure Marketplace storefront, the Azure portal, or the
Azure CLI.
Make sure that you've installed and configured the latest Azure PowerShell module.
Terminology
A Marketplace image in Azure has the following attributes:
Publisher: The organization that created the image. Examples: Canonical, MicrosoftWindowsServer
Offer: The name of a group of related images created by a publisher. Examples: Ubuntu Server, WindowsServer
SKU: An instance of an offer, such as a major release of a distribution. Examples: 16.04-LTS, 2016-Datacenter
Version: The version number of an image SKU.
To identify a Marketplace image when you deploy a VM programmatically, supply these values individually as
parameters. Some tools accept an image URN, which combines these values, separated by the colon (:) character:
Publisher:Offer:Sku:Version. In a URN, you can replace the version number with "latest", which selects the latest
version of the image.
If the image publisher provides additional license and purchase terms, then you must accept those terms and
enable programmatic deployment. You'll also need to supply purchase plan parameters when deploying a VM
programmatically. See Deploy an image with Marketplace terms.
Table of commonly used Windows images

PUBLISHER OFFER SKU
MicrosoftWindowsServer WindowsServer 2016-Datacenter
MicrosoftWindowsServer WindowsServer 2016-Datacenter-Server-Core
MicrosoftWindowsServer WindowsServer 2016-Datacenter-with-Containers
MicrosoftWindowsServer WindowsServer 2012-R2-Datacenter
MicrosoftWindowsServer WindowsServer 2012-Datacenter
MicrosoftWindowsServer WindowsServer 2008-R2-SP1
MicrosoftDynamicsNAV DynamicsNAV 2017

PUBLISHER OFFER SKU
MicrosoftSharePoint MicrosoftSharePointServer 2016
MicrosoftSQLServer SQL2017-WS2016 Enterprise
MicrosoftRServer RServer-WS2016 Enterprise
Navigate the images

One way to find an image in a location is to run the Get-AzureRMVMImagePublisher, Get-
AzureRMVMImageOffer, and Get-AzureRMVMImageSku cmdlets in order:
1. List the image publishers.
2. For a given publisher, list their offers.
3. For a given offer, list their SKUs.
Then, for a selected SKU, run Get-AzureRMVMImage to list the versions to deploy.
1. List the publishers:
$locName="<Azure location, such as West US>"

Get-AzureRMVMImagePublisher -Location $locName | Select PublisherName
2. Fill in your chosen publisher name and list the offers:
$pubName="<publisher>"
Get-AzureRMVMImageOffer -Location $locName -Publisher $pubName | Select Offer
3. Fill in your chosen offer name and list the SKUs:
$offerName="<offer>"
Get-AzureRMVMImageSku -Location $locName -Publisher $pubName -Offer $offerName | Select Skus
4. Fill in your chosen SKU name and get the image version:
$skuName="<SKU>"
Get-AzureRMVMImage -Location $locName -Publisher $pubName -Offer $offerName -Sku $skuName | Select
Version
From the output of the Get-AzureRMVMImage command, you can select a version image to deploy a new virtual
machine.
The following example shows the full sequence of commands and their outputs:
$locName="West US"
Get-AzureRMVMImagePublisher -Location $locName | Select PublisherName
Partial output:
PublisherName
-------------
...
a10networks
aiscaler-cache-control-ddos-and-url-rewriting-
alertlogic
AlertLogic.Extension
Barracuda.Azure.ConnectivityAgent
barracudanetworks
basho
boxless
bssw
Canonical
...
For the MicrosoftWindowsServer publisher:
$pubName="MicrosoftWindowsServer"
Get-AzureRMVMImageOffer -Location $locName -Publisher $pubName | Select Offer
Output:
Offer
-----
Windows-HUB
WindowsServer
WindowsServerSemiAnnual
For the WindowsServer offer:
$offerName="WindowsServer"
Get-AzureRMVMImageSku -Location $locName -Publisher $pubName -Offer $offerName | Select Skus
Output:
Skus
----
2008-R2-SP1
2008-R2-SP1-smalldisk
2012-Datacenter
2012-Datacenter-smalldisk
2012-R2-Datacenter
2012-R2-Datacenter-smalldisk
2016-Datacenter
2016-Datacenter-Server-Core
2016-Datacenter-Server-Core-smalldisk
2016-Datacenter-smalldisk
2016-Datacenter-with-Containers
2016-Datacenter-with-RDSH
Then, for the 2016 -Datacenter SKU:
$skuName="2016-Datacenter"
Get-AzureRMVMImage -Location $locName -Publisher $pubName -Offer $offerName -Sku $skuName | Select Version
Now you can combine the selected publisher, offer, SKU, and version into a URN (values separated by :). Pass this
URN with the --image parameter when you create a VM with the New -AzureRmVM cmdlet. You can optionally
replace the version number in the URN with "latest" to get the latest version of the image.
If you deploy a VM with a Resource Manager template, then you'll set the image parameters individually in the
imageReference properties. See the template reference.
Deploy an image with Marketplace terms

Some VM images in the Azure Marketplace have additional license and purchase terms that you must accept
before you can deploy them programmatically.
To deploy a VM from such an image, you'll need to both accept the image's terms and enable programmatic
deployment. You'll only need to do this once per subscription. Afterward, each time you deploy a VM
programmatically from the image you'll also need to specify purchase plan parameters.
The following sections show how to:
Find out whether a Marketplace image has additional license terms
Accept the terms programmatically
Provide purchase plan parameters when you deploy a VM programmatically
View plan properties
To view an image's purchase plan information, run the Get-AzureRMVMImage cmdlet. If the PurchasePlan property in
the output is not null , the image has terms you need to accept before programmatic deployment.
For example, the Windows Server 2016 Datacenter image doesn't have additional terms, so the PurchasePlan
information is null :
$version = "2016.127.20170406"
Get-AzureRMVMImage -Location $locName -Publisher $pubName -Offer $offerName -Skus $skuName -Version $version
Output:
Id : /Subscriptions/xxxxxxxx-xxxx-xxxx-xxxx-
xxxxxxxxxxxx/Providers/Microsoft.Compute/Locations/westus/Publishers/MicrosoftWindowsServer/ArtifactTypes/VMIm
age/Offers/WindowsServer/Skus/2016-Datacenter/
Versions/2016.127.20170406
Location : westus
PublisherName : MicrosoftWindowsServer
Offer : WindowsServer
Skus : 2016-Datacenter
Version : 2016.127.20170406
FilterExpression :
Name : 2016.127.20170406
OSDiskImage : {
"operatingSystem": "Windows"
}
PurchasePlan : null
DataDiskImages : []
The example below shows a similar command for the Data Science Virtual Machine - Windows 2016 image, which
has the following PurchasePlan properties: name , product , and publisher . Some images also have a
promotion code property. To deploy this image, see the following sections to accept the terms and to enable
programmatic deployment.
Get-AzureRMVMImage -Location "westus" -Publisher "microsoft-ads" -Offer "windows-data-science-vm" -Skus

"windows2016" -Version "0.2.02"
Output:
Id : /Subscriptions/xxxxxxxx-xxxx-xxxx-xxxx-
xxxxxxxxxxxx/Providers/Microsoft.Compute/Locations/westus/Publishers/microsoft-ads/ArtifactTypes/VMIma
ge/Offers/windows-data-science-vm/Skus/windows2016/Versions/0.2.02
Location : westus
PublisherName : microsoft-ads
Offer : windows-data-science-vm
Skus : windows2016
Version : 0.2.02
FilterExpression :
Name : 0.2.02
OSDiskImage : {
"operatingSystem": "Windows"
}
PurchasePlan : {
"publisher": "microsoft-ads",
"name": "windows2016",
"product": "windows-data-science-vm"
}
DataDiskImages : []
Accept the terms

To view the license terms, use the Get-AzureRmMarketplaceterms cmdlet and pass in the purchase plan
parameters. The output provides a link to the terms for the Marketplace image and shows whether you previously
accepted the terms. Be sure to use all lowercase letters in the parameter values.
Get-AzureRmMarketplaceterms -Publisher "microsoft-ads" -Product "windows-data-science-vm" -Name "windows2016"
Output:
Publisher : microsoft-ads
Product : windows-data-science-vm
Plan : windows2016
LicenseTextLink :
https://storelegalterms.blob.core.windows.net/legalterms/3E5ED_legalterms_MICROSOFT%253a2DADS%253a24WINDOWS%25
3a2DDATA%253a2DSCIENCE%253a2DV
M%253a24WINDOWS2016%253a24OC5SKMQOXSED66BBSNTF4XRCS4XLOHP7QMPV54DQU7JCBZWYFP35IDPOWTUKXUC7ZAG7W6ZMDD6NHWNKUIVS
YBZUTZ245F44SU5AD7Q.txt
PrivacyPolicyLink : https://www.microsoft.com/EN-US/privacystatement/OnlineServices/Default.aspx
Signature :
2UMWH6PHSAIM4U22HXPXW25AL2NHUJ7Y7GRV27EBL6SUIDURGMYG6IIDO3P47FFIBBDFHZHSQTR7PNK6VIIRYJRQ3WXSE6BTNUNENXA
Accepted : False
Signdate : 2/23/2018 7:43:00 PM
Use the Set-AzureRmMarketplaceterms cmdlet to accept or reject the terms. You only need to accept terms once
per subscription for the image. Be sure to use all lowercase letters in the parameter values.
$agreementTerms=Get-AzureRmMarketplaceterms -Publisher "microsoft-ads" -Product "windows-data-science-vm" -

Name "windows2016"
Set-AzureRmMarketplaceTerms -Publisher "microsoft-ads" -Product "windows-data-science-vm" -Name "windows2016"

-Terms $agreementTerms -Accept
Output:
Publisher : microsoft-ads
Product : windows-data-science-vm
Plan : windows2016
LicenseTextLink :
https://storelegalterms.blob.core.windows.net/legalterms/3E5ED_legalterms_MICROSOFT%253a2DADS%253a24WINDOWS%25
3a2DDATA%253a2DSCIENCE%253a2DV
M%253a24WINDOWS2016%253a24OC5SKMQOXSED66BBSNTF4XRCS4XLOHP7QMPV54DQU7JCBZWYFP35IDPOWTUKXUC7ZAG7W6ZMDD6NHWNKUIVS
YBZUTZ245F44SU5AD7Q.txt
PrivacyPolicyLink : https://www.microsoft.com/EN-US/privacystatement/OnlineServices/Default.aspx
Signature :
XXXXXXK3MNJ5SROEG2BYDA2YGECU33GXTD3UFPLPC4BAVKAUL3PDYL3KBKBLG4ZCDJZVNSA7KJWTGMDSYDD6KRLV3LV274DLBXXXXXX
Accepted : True
Signdate : 2/23/2018 7:49:31 PM
Deploy using purchase plan parameters

After accepting the terms for an image, you can deploy a VM in that subscription. As shown in the following
snippet, use the Set-AzureRmVMPlan cmdlet to set the Marketplace plan information for the VM object. For a
complete script to create network settings for the VM and complete the deployment, see the PowerShell script
examples.
...
$vmConfig = New-AzureRmVMConfig -VMName "myVM" -VMSize Standard_D1
# Set the Marketplace plan information
$publisherName = "microsoft-ads"
$productName = "windows-data-science-vm"
$planName = "windows2016"
$vmConfig = Set-AzureRmVMPlan -VM $vmConfig -Publisher $publisherName -Product $productName -Name $planName
$cred=Get-Credential
$vmConfig = Set-AzureRmVMOperatingSystem -Windows -VM $vmConfig -ComputerName "myVM" -Credential $cred
# Set the Marketplace image
$offerName = "windows-data-science-vm"
$skuName = "windows2016"
$version = "0.2.02"
$vmConfig = Set-AzureRmVMSourceImage -VM $vmConfig -PublisherName $publisherName -Offer $offerName -Skus

$skuName -Version $version
...
You'll then pass the VM configuration along with network configuration objects to the New-AzureRmVM cmdlet.
Next steps
To create a virtual machine quickly with the New-AzureRmVM cmdlet by using basic image information, see Create a
Windows virtual machine with PowerShell.
See a PowerShell script example to create a fully configured virtual machine.
Prepare a Windows VHD or VHDX to upload to
Azure
Before you upload a Windows virtual machines (VM ) from on-premises to Microsoft Azure, you must prepare the
virtual hard disk (VHD or VHDX). Azure supports only generation 1 VMs that are in the VHD file format and
have a fixed sized disk. The maximum size allowed for the VHD is 1,023 GB. You can convert a generation 1 VM
from the VHDX file system to VHD and from a dynamically expanding disk to fixed-sized. But you can't change a
VM's generation. For more information, see Should I create a generation 1 or 2 VM in Hyper-V.
For more information about the support policy for Azure VM, see Microsoft server software support for
Microsoft Azure VMs.
NOTE
The instructions in this article apply to the 64-bit version of Windows Server 2008 R2 and later Windows server operating
system. For information about running 32-bit version of operating system in Azure, see Support for 32-bit operating
systems in Azure virtual machines.
Convert the virtual disk to VHD and fixed size disk

If you need to convert your virtual disk to the required format for Azure, use one of the methods in this section.
Back up the VM before you run the virtual disk conversion process and make sure that the Windows VHD works
correctly on the local server. Resolve any errors within the VM itself before you try to convert or upload it to
Azure.
After you convert the disk, create a VM that uses the converted disk. Start and sign in to the VM to finish
preparing the VM for upload.
Convert disk using Hyper-V Manager
1. Open Hyper-V Manager and select your local computer on the left. In the menu above the computer list, click
Action > Edit Disk.
2. On the Locate Virtual Hard Disk screen, locate and select your virtual disk.
3. On the Choose Action screen, and then select Convert and Next.
4. If you need to convert from VHDX, select VHD and then click Next.
5. If you need to convert from a dynamically expanding disk, select Fixed size and then click Next.
6. Locate and select a path to save the new VHD file to.
7. Click Finish.
NOTE
The commands in this article must be run on an elevated PowerShell session.
Convert disk by using PowerShell

You can convert a virtual disk by using the Convert-VHD command in Windows PowerShell. Select Run as
administrator when you start PowerShell.
The following example command converts from VHDX to VHD, and from a dynamically expanding disk to fixed-
size:
Convert-VHD –Path c:\test\MY-VM.vhdx –DestinationPath c:\test\MY-NEW-VM.vhd -VHDType Fixed
In this command, replace the value for "-Path" with the path to the virtual hard disk that you want to convert and
the value for "-DestinationPath" with the new path and name of the converted disk.
Convert from VMware VMDK disk format
If you have a Windows VM image in the VMDK file format, convert it to a VHD by using the Microsoft VM
Converter. For more information, see the blog article How to Convert a VMware VMDK to Hyper-V VHD.
Set Windows configurations for Azure

On the VM that you plan to upload to Azure, run all commands in the following steps from an elevated Command
Prompt window:
1. Remove any static persistent route on the routing table:
To view the route table, run route print at the command prompt.
Check the Persistence Routes sections. If there is a persistent route, use the route delete command to
remove it.
2. Remove the WinHTTP proxy:
netsh winhttp reset proxy
3. Set the disk SAN policy to Onlineall:
diskpart
In the open Command Prompt window, type the following commands:
san policy=onlineall
exit
4. Set Coordinated Universal Time (UTC ) time for Windows and the startup type of the Windows Time
(w32time) service to Automatically:
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\TimeZoneInformation' -name

"RealTimeIsUniversal" -Value 1 -Type DWord -force
Set-Service -Name w32time -StartupType Automatic
5. Set the power profile to the High Performance:
powercfg /setactive SCHEME_MIN
6. Make sure that the environmental variables TEMP and TMP are set to their default values:
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\Session Manager\Environment' -name
"TEMP" -Value "%SystemRoot%\TEMP" -Type ExpandString -force
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\Session Manager\Environment' -name "TMP"

-Value "%SystemRoot%\TEMP" -Type ExpandString -force
Check the Windows services

Make sure that each of the following Windows services is set to the Windows default values. These are the
minimum numbers of services that must be set up to make sure that the VM has connectivity. To reset the startup
settings, run the following commands:
Set-Service -Name bfe -StartupType Automatic

Set-Service -Name dhcp -StartupType Automatic
Set-Service -Name dnscache -StartupType Automatic
Set-Service -Name IKEEXT -StartupType Automatic
Set-Service -Name iphlpsvc -StartupType Automatic
Set-Service -Name netlogon -StartupType Manual
Set-Service -Name netman -StartupType Manual
Set-Service -Name nsi -StartupType Automatic
Set-Service -Name termService -StartupType Manual
Set-Service -Name MpsSvc -StartupType Automatic
Set-Service -Name RemoteRegistry -StartupType Automatic
Update Remote Desktop registry settings

Make sure that the following settings are configured correctly for remote desktop connection:
NOTE
You may receive an error message when you run the Set-ItemProperty -Path
'HKLM:\SOFTWARE\Policies\Microsoft\Windows NT\Terminal Services -name <object name> -value <value> in
these steps. The error message can be safely ignored. It means only that the domain is not pushing that configuration
through a Group Policy object.
1. Remote Desktop Protocol (RDP ) is enabled:
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\Terminal Server' -name

"fDenyTSConnections" -Value 0 -Type DWord -force
Set-ItemProperty -Path 'HKLM:\SOFTWARE\Policies\Microsoft\Windows NT\Terminal Services' -name

"fDenyTSConnections" -Value 0 -Type DWord -force
2. The RDP port is set up correctly (Default port 3389):
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\Terminal Server\Winstations\RDP-Tcp' -

name "PortNumber" -Value 3389 -Type DWord -force
When you deploy a VM, the default rules are created against port 3389. If you want to change the port
number, do that after the VM is deployed in Azure.
3. The listener is listening in every network interface:
name "LanAdapter" -Value 0 -Type DWord -force
4. Configure the Network Level Authentication mode for the RDP connections:
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\Terminal Server\WinStations\RDP-Tcp' -

name "UserAuthentication" -Value 1 -Type DWord -force

name "SecurityLayer" -Value 1 -Type DWord -force

name "fAllowSecProtocolNegotiation" -Value 1 -Type DWord -force
5. Set the keep-alive value:

"KeepAliveEnable" -Value 1 -Type DWord -force
"KeepAliveInterval" -Value 1 -Type DWord -force
name "KeepAliveTimeout" -Value 1 -Type DWord -force
6. Reconnect:

"fDisableAutoReconnect" -Value 0 -Type DWord -force
name "fInheritReconnectSame" -Value 1 -Type DWord -force
name "fReconnectSame" -Value 0 -Type DWord -force
7. Limit the number of concurrent connections:

name "MaxInstanceCount" -Value 4294967295 -Type DWord -force
8. If there are any self-signed certificates tied to the RDP listener, remove them:
Remove-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\Terminal Server\WinStations\RDP-Tcp'

-name "SSLCertificateSHA1Hash" -force
This is to make sure that you can connect at the beginning when you deploy the VM. You can also review
this on a later stage after the VM is deployed in Azure if needed.
9. If the VM will be part of a Domain, check all the following settings to make sure that the former settings
are not reverted. The policies that must be checked are the following:
GOAL POLICY VALUE

GOAL POLICY VALUE
RDP is enabled Computer Allow users to connect remotely by

Configuration\Policies\Windows using Remote Desktop
Settings\Administrative
Templates\Components\Remote
Desktop Services\Remote Desktop
Session Host\Connections
NLA group policy Settings\Administrative Require user Authentication for

Templates\Components\Remote remote connections by using
Desktop Services\Remote Desktop Network Level Authentication
Session Host\Security
Keep Alive settings Computer Configure keep-alive connection

Configuration\Policies\Windows interval
Templates\Windows
Components\Remote Desktop
Services\Remote Desktop Session
Host\Connections
Reconnect settings Computer Automatic reconnection

Configuration\Policies\Windows
Templates\Windows
Host\Connections
Limit the number of connections Computer Limit number of connections

settings Configuration\Policies\Windows
Templates\Windows
Host\Connections
Configure Windows Firewall rules

1. Turn on Windows Firewall on the three profiles (Domain, Standard, and Public):
Set-NetFirewallProfile -Profile Domain,Public,Private -Enabled True
2. Run the following command in PowerShell to allow WinRM through the three firewall profiles (Domain,
Private, and Public) and enable the PowerShell Remote service:
Enable-PSRemoting -force
Set-NetFirewallRule -DisplayName "Windows Remote Management (HTTP-In)" -Enabled True
3. Enable the following firewall rules to allow the RDP traffic:
Set-NetFirewallRule -DisplayGroup "Remote Desktop" -Enabled True
4. Enable the File and Printer Sharing rule so that the VM can respond to a ping command inside the Virtual
Network:
Set-NetFirewallRule -DisplayName "File and Printer Sharing (Echo Request - ICMPv4-In)" -Enabled True
5. If the VM will be part of a Domain, check the following settings to make sure that the former settings are
not reverted. The AD policies that must be checked are the following:
GOAL POLICY VALUE
Enable the Windows Firewall profiles Computer Protect all network connections
Templates\Network\Network
Connection\Windows
Firewall\Domain Profile\Windows
Firewall
Enable RDP Computer Allow inbound Remote Desktop

Configuration\Policies\Windows exceptions
Connection\Windows
Firewall
Computer Allow inbound Remote Desktop

Configuration\Policies\Windows exceptions
Connection\Windows
Firewall\Standard Profile\Windows
Firewall
Enable ICMP-V4 Computer Allow ICMP exceptions

Connection\Windows
Firewall
Computer Allow ICMP exceptions

Connection\Windows
Firewall\Standard Profile\Windows
Firewall
Verify VM is healthy, secure, and accessible with RDP

1. To make sure the disk is healthy and consistent, run a check disk operation at the next VM restart:
Chkdsk /f
Make sure that the report shows a clean and healthy disk.
2. Set the Boot Configuration Data (BCD ) settings.
NOTE
Make sure you run these commands on an elevated PowerShell window.
cmd
bcdedit /set {bootmgr} integrityservices enable

bcdedit /set {default} device partition=C:
bcdedit /set {default} integrityservices enable
bcdedit /set {default} recoveryenabled Off
bcdedit /set {default} osdevice partition=C:
bcdedit /set {default} bootstatuspolicy IgnoreAllFailures
#Enable Serial Console Feature

bcdedit /set {bootmgr} displaybootmenu yes
bcdedit /set {bootmgr} timeout 5
bcdedit /set {bootmgr} bootems yes
bcdedit /ems {current} ON
bcdedit /emssettings EMSPORT:1 EMSBAUDRATE:115200
exit
3. The Dump log can be helpful in troubleshooting Windows crash issues. Enable the Dump log collection:
# Setup the Guest OS to collect a kernel dump on an OS crash event

Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\CrashControl' -name CrashDumpEnabled -
Type DWord -force -Value 2
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\CrashControl' -name DumpFile -Type
ExpandString -force -Value "%SystemRoot%\MEMORY.DMP"
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\CrashControl' -name NMICrashDump -Type
DWord -force -Value 1
#Setup the Guest OS to collect user mode dumps on a service crash event
$key = 'HKLM:\SOFTWARE\Microsoft\Windows\Windows Error Reporting\LocalDumps'
if ((Test-Path -Path $key) -eq $false) {(New-Item -Path 'HKLM:\SOFTWARE\Microsoft\Windows\Windows Error
Reporting' -Name LocalDumps)}
New-ItemProperty -Path $key -name DumpFolder -Type ExpandString -force -Value "c:\CrashDumps"
New-ItemProperty -Path $key -name CrashCount -Type DWord -force -Value 10
New-ItemProperty -Path $key -name DumpType -Type DWord -force -Value 2
Set-Service -Name WerSvc -StartupType Manual
4. Verify that the Windows Management Instrumentations repository is consistent. To perform this, run the
following command:
winmgmt /verifyrepository
If the repository is corrupted, see WMI: Repository Corruption, or Not.

5. Make sure that any other application is not using the port 3389. This port is used for the RDP service in
Azure. You can run netstat -anob to see which ports are in used on the VM:
netstat -anob
6. If the Windows VHD that you want to upload is a domain controller, then follow these steps:
a. Follow these extra steps to prepare the disk.
b. Make sure that you know the DSRM password in case you have to start the VM in DSRM at some
point. You may want to refer to this link to set the DSRM password.
7. Make sure that the Built-in Administrator account and password are known to you. You may want to reset
the current local administrator password and make sure that you can use this account to sign in to
Windows through the RDP connection. This access permission is controlled by the "Allow log on through
Remote Desktop Services" Group Policy object. You can view this object in the Local Group Policy Editor
under:
Computer Configuration\Windows Settings\Security Settings\Local Policies\User Rights Assignment
8. Check the following AD polices to make sure that you are not blocking your RDP access through RDP nor
from the network:
Computer Configuration\Windows Settings\Security Settings\Local Policies\User Rights
Assignment\Deny access to this computer from the network
Assignment\Deny log on through Remote Desktop Services
9. Check the following AD policy to make sure that you are not removing any of the following the required
access accounts:
Assignment\Access this compute from the network
The following groups should be listed on this policy:
Administrators
Backup Operators
Everyone
Users
10. Restart the VM to make sure that Windows is still healthy can be reached by using the RDP connection. At
this point, you may want to create a VM in your local Hyper-V to make sure the VM is starting completely
and then test whether it is RDP reachable.
11. Remove any extra Transport Driver Interface filters, such as software that analyzes TCP packets or extra
firewalls. You can also review this on a later stage after the VM is deployed in Azure if needed.
12. Uninstall any other third-party software and driver that is related to physical components or any other
virtualization technology.
Install Windows Updates
The ideal configuration is to have the patch level of the machine at the latest. If this is not possible, make
sure that the following updates are installed:
WINDOWS
10 WINDOWS
WINDOWS VERSION WINDOWS 10
7 1607WIND 10 1709 1803WIND
SP1,WIND WINDOWS WINDOWS OWS WINDOWS OWS
OWS 8,WINDO 8.1,WIND SERVER WINDOWS SERVER SERVER
SERVER WS OWS 2016 10 2016 2016
COMPONE 2008 R2 SERVER SERVER VERSION VERSION VERSION VERSION
NT BINARY SP1 2012 2012 R2 1607 1703 1709 1803
WINDOWS
10 WINDOWS
7 1607WIND 10 1709 1803WIND
SERVER WS OWS 2016 10 2016 2016
NT BINARY SP1 2012 2012 R2 1607 1703 1709 1803
Storage disk.sys 6.1.7601. 6.2.9200. 6.3.9600. - - - -

23403 - 17638 / 18203 -
KB31255 6.2.9200. KB31370
74 21757 - 61
KB31370
61
storport.s 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

ys 23403 - 17188 / 18573 - 3.1358 - 3.332
KB31255 6.2.9200. KB40227 KB40227
74 21306 - 26 15
KB30184
89
ntfs.sys 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

23403 - 17623 / 18654 - 3.1198 - 3.447
KB31255 6.2.9200. KB40227 KB40227
74 21743 - 26 15
KB31212
55
Iologmsg. 6.1.7601. 6.2.9200. - - - - -

dll 23403 - 16384 -
KB31255 KB29953
74 87
Classpnp. 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 - - -

sys 23403 - 17061 / 18334 - 3.953 -
KB31255 6.2.9200. KB31726 KB40227
74 21180 - 14 15
KB29953
87
Volsnap.s 6.1.7601. 6.2.9200. 6.3.9600. - 10.0.1506 - -

ys 23403 - 17047 / 18265 - 3.0
KB31255 6.2.9200. KB31453
74 21165 - 84
KB29753
31
partmgr.s 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

ys 23403 - 16681 - 17401 - 3.953 - 3.0
KB31255 KB28771 KB30008 KB40227
74 14 50 15
volmgr.sy 10.0.1506 - -
s 3.0
WINDOWS
10 WINDOWS
7 1607WIND 10 1709 1803WIND
SERVER WS OWS 2016 10 2016 2016
NT BINARY SP1 2012 2012 R2 1607 1703 1709 1803
Volmgrx.s 6.1.7601. - - - 10.0.1506 - -

ys 23403 - 3.0
KB31255
74
Msiscsi.sy 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

s 23403 - 21006 - 18624 - 3.1066 - 3.447
KB31255 KB29551 KB40227 KB40227
74 63 26 15
Msdsm.sy 6.1.7601. 6.2.9200. 6.3.9600. - - - -

s 23403 - 21474 - 18592 -
KB31255 KB30461 KB40227
74 01 26
Mpio.sys 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 - - -

23403 - 21190 - 18616 - 3.1198 -
KB31255 KB30461 KB40227 KB40227
74 01 26 15
vmstorfl.s 6.3.9600. 6.3.9600. 6.3.9600. 10.0.1439 10.0.1506 10.0.1629 -

ys 18907 - 18080 - 18907 - 3.2007 - 3.850 - 9.371 -
KB40726 KB30631 KB40726 KB43454 KB43454 KB43454
50 09 50 18 19 20
Fveapi.dll 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 - - -

23311 - 20930 - 18294 - 3.576 -
KB31255 KB29302 KB31726 KB40227
74 44 14 15
Fveapibas 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 - - -

e.dll 23403 - 20930 - 17415 - 3.206 -
KB31255 KB29302 KB31726 KB40227
74 44 14 15
Network netvsc.sys - - - 10.0.1439 10.0.1506 - -

3.1198 - 3.250 -
KB40227 KB40200
15 01
mrxsmb1 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

0.sys 23816 - 22108 - 18603 - 3.479 - 3.483
KB40227 KB40227 KB40227 KB40227
22 24 26 15
mrxsmb2 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

0.sys 23816 - 21548 - 18586 - 3.953 - 3.483
KB40227 KB40227 KB40227 KB40227
22 24 26 15
WINDOWS
10 WINDOWS
7 1607WIND 10 1709 1803WIND
SERVER WS OWS 2016 10 2016 2016
NT BINARY SP1 2012 2012 R2 1607 1703 1709 1803
mrxsmb.s 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

ys 23816 - 22074 - 18586 - 3.953 - 3.0
KB40227 KB40227 KB40227 KB40227
22 24 26 15
tcpip.sys 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

23761 - 22070 - 18478 - 3.1358 - 3.447
KB40227 KB40227 KB40227 KB40227
22 24 26 15
http.sys 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

23403 - 17285 - 18574 - 3.251 - 3.483
KB31255 KB30425 KB40227 KB40227
74 53 26 15
vmswitch. 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

sys 23727 - 22117 - 18654 - 3.1358 - 3.138
KB40227 KB40227 KB40227 KB40227
19 24 26 15
Core ntoskrnl.e 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

xe 23807 - 22170 - 18696 - 3.1358 - 3.483
KB40227 KB40227 KB40227 KB40227
19 18 26 15
Remote rdpcorets. 6.2.9200. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

Desktop dll 21506 - 22104 - 18619 - 3.1198 - 3.0
Services KB40227 KB40227 KB40227 KB40227
19 24 26 15
termsrv.dl 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 10.0.1506 - -

l 23403 - 17048 - 17415 - 3.0 - 3.0
KB31255 KB29735 KB30008 KB40227
74 01 50 15
termdd.sy 6.1.7601. - - - - - -
s 23403 -
KB31255
74
win32k.sy 6.1.7601. 6.2.9200. 6.3.9600. 10.0.1439 - - -

s 23807 - 22168 - 18698 - 3.594 -
KB40227 KB40227 KB40227 KB40227
19 18 26 15
rdpdd.dll 6.1.7601. - - - - - -
23403 -
KB31255
74
WINDOWS
10 WINDOWS
7 1607WIND 10 1709 1803WIND
SERVER WS OWS 2016 10 2016 2016
NT BINARY SP1 2012 2012 R2 1607 1703 1709 1803
rdpwd.sys 6.1.7601. - - - - - -
23403 -
KB31255
74
Security MS17- KB40122 KB40122 KB40122 KB40126 KB40126 - -

010 12 13 13 06 06
KB40122 KB40131 KB40131 - -

16 98 98
KB40122 KB40122 KB40122 KB40134 KB40134 - -

15 14 16 29 29
KB40122 KB40134 KB40134 - -

17 29 29
CVE- KB41037 KB41037 KB41037 KB41037 KB41037 KB41037 KB41037

2018- 18 30 25 23 31 27 21
0886
KB41037 KB41037 KB41037

12 26 15
When to use sysprep

Sysprep is a process that you could run into a windows installation that will reset the installation of the system
and will provide an “out of the box experience” by removing all personal data and resetting several components.
You typically do this if you want to create a template from which you can deploy several other VMs that have a
specific configuration. This is called a generalized image.
If, instead, you want only to create one VM from one disk, you don’t have to use sysprep. In this situation, you can
just create the VM from what is known as a specialized image.
For more information about how to create a VM from a specialized disk, see:
Create a VM from a specialized disk
Create a VM from a specialized VHD disk
If you want to create a generalized image, you need to run sysprep. For more information about Sysprep, see How
to Use Sysprep: An Introduction.
Not every role or application that’s installed on a Windows-based computer supports this generalization. So
before you run this procedure, refer to the following article to make sure that the role of that computer is
supported by sysprep. For more information, Sysprep Support for Server Roles.
Steps to generalize a VHD
NOTE
After you run sysprep.exe as specified in the following steps, turn off the VM, and do not turn it back on until you create an
image from it in Azure.
1. Sign in to the Windows VM.

2. Run Command Prompt as an administrator.
3. Change the directory to: %windir%\system32\sysprep, and then run sysprep.exe.
4. In the System Preparation Tool dialog box, select Enter System Out-of-Box Experience (OOBE ), and
make sure that the Generalize check box is selected.
5. In Shutdown Options, select Shutdown.

6. Click OK.
7. When Sysprep completes, shut down the VM. Do not use Restart to shut down the VM.
8. Now the VHD is ready to be uploaded. For more information about how to create a VM from a generalized
disk, see Upload a generalized VHD and use it to create a new VMs in Azure.
Complete recommended configurations

The following settings do not affect VHD uploading. However, we strongly recommend that you configured them.
Install the Azure VMs Agent. Then you can enable VM extensions. The VM extensions implement most of
the critical functionality that you might want to use with your VMs such as resetting passwords,
configuring RDP, and so on. For more information, see:
VM Agent and Extensions – Part 1
VM Agent and Extensions – Part 2
After the VM is created in Azure, we recommend that you put the pagefile on the ”Temporal drive” volume
to improve performance. You can set up this as follows:
Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\Session Manager\Memory Management' -name

"PagingFiles" -Value "D:\pagefile" -Type MultiString -force
If there’s any data disk that is attached to the VM, the Temporal drive volume's drive letter is typically "D."
This designation could be different, depending on the number of available drives and the settings that you
make.
Next steps
Upload a Windows VM image to Azure for Resource Manager deployments
Troubleshoot Azure Windows virtual machine activation problems
Create a managed image of a generalized VM in
Azure
A managed image resource can be created from a generalized virtual machine (VM ) that is stored as either a
managed disk or an unmanaged disk in a storage account. The image can then be used to create multiple VMs.
Generalize the Windows VM using Sysprep

Sysprep removes all your personal account and security information, and then prepares the machine to be used as
an image. For information about Sysprep, see Sysprep overview.
Make sure the server roles running on the machine are supported by Sysprep. For more information, see Sysprep
support for server roles.
IMPORTANT
After you have run Sysprep on a VM, that VM is considered generalized and cannot be restarted. The process of
generalizing a VM is not reversible. If you need to keep the original VM functioning, you should create a copy of the VM and
generalize its copy.
If you plan to run Sysprep before uploading your virtual hard disk (VHD) to Azure for the first time, make sure you have
prepared your VM.
To generalize your Windows VM, follow these steps:

1. Sign in to your Windows VM.
2. Open a Command Prompt window as an administrator. Change the directory to
%windir%\system32\sysprep, and then run sysprep.exe .
3. In the System Preparation Tool dialog box, select Enter System Out-of-Box Experience (OOBE ) and
select the Generalize check box.
4. For Shutdown Options, select Shutdown.
5. Select OK.
6. When Sysprep completes, it shuts down the VM. Do not restart the VM.
Create a managed image in the portal
2. In the menu on the left, select Virtual machines and then select the VM from the list.
3. In the Virtual machine page for the VM, on the upper menu, select Capture.
The Create image page appears.
4. For Name, either accept the pre-populated name or enter a name that you would like to use for the image.
5. For Resource group, either select Create new and enter a name, or select Use existing and select a
resource group to use from the drop-down list.
6. If you want to delete the source VM after the image has been created, select Automatically delete this
virtual machine after creating the image.
7. If you want the ability to use the image in any availability zone, select On for Zone resiliency.
8. Select Create to create the image.
9. After the image is created, you can find it as an Image resource in the list of resources in the resource
group.
Create an image of a VM using Powershell

Creating an image directly from the VM ensures that the image includes all of the disks associated with the VM,
including the OS disk and any data disks. This example shows how to create a managed image from a VM that
uses managed disks.
Before you begin, make sure that you have the latest version of the AzureRM.Compute PowerShell module, which
must be version 5.7.0 or later. To find the version, run Get-Module -ListAvailable AzureRM.Compute in PowerShell. If
you need to upgrade, see Install Azure PowerShell on Windows with PowerShellGet. If you are running
PowerShell locally, run Connect-AzureRmAccount to create a connection with Azure.
NOTE
If you would like to store your image in zone-redundant storage, you need to create it in a region that supports availability
zones and include the -ZoneResilient parameter in the image configuration ( New-AzureRmImageConfig command).
To create a VM image, follow these steps:

1. Create some variables.
$vmName = "myVM"
$location = "EastUS"
$imageName = "myImage"
2. Make sure the VM has been deallocated.
Stop-AzureRmVM -ResourceGroupName $rgName -Name $vmName -Force
3. Set the status of the virtual machine to Generalized.

Set-AzureRmVm -ResourceGroupName $rgName -Name $vmName -Generalized
4. Get the virtual machine.
$vm = Get-AzureRmVM -Name $vmName -ResourceGroupName $rgName
5. Create the image configuration.
$image = New-AzureRmImageConfig -Location $location -SourceVirtualMachineId $vm.Id
6. Create the image.
New-AzureRmImage -Image $image -ImageName $imageName -ResourceGroupName $rgName
Create an image from a managed disk using PowerShell

If you want to create an image of only the OS disk, specify the managed disk ID as the OS disk:
$vmName = "myVM"
$snapshotName = "mySnapshot"
2. Get the VM.
$vm = Get-AzureRmVm -Name $vmName -ResourceGroupName $rgName
3. Get the ID of the managed disk.
$diskID = $vm.StorageProfile.OsDisk.ManagedDisk.Id
$imageConfig = New-AzureRmImageConfig -Location $location

$imageConfig = Set-AzureRmImageOsDisk -Image $imageConfig -OsState Generalized -OsType Windows -
ManagedDiskId $diskID
New-AzureRmImage -ImageName $imageName -ResourceGroupName $rgName -Image $imageConfig
Create an image from a snapshot using Powershell

You can create a managed image from a snapshot of a generalized VM by following these steps:
$snapshotName = "mySnapshot"
2. Get the snapshot.
$snapshot = Get-AzureRmSnapshot -ResourceGroupName $rgName -SnapshotName $snapshotName

$imageConfig = Set-AzureRmImageOsDisk -Image $imageConfig -OsState Generalized -OsType Windows -
SnapshotId $snapshot.Id
New-AzureRmImage -ImageName $imageName -ResourceGroupName $rgName -Image $imageConfig
Create an image from a VHD in a storage account

Create a managed image from a generalized OS VHD in a storage account. You need the URI of the VHD in the
storage account, which is in the following format:
https://mystorageaccount.blob.core.windows.net/vhdcontainer/vhdfilename.vhd. In this example, the VHD is in
mystorageaccount, in a container named vhdcontainer, and the VHD filename is vhdfilename.vhd.
$vmName = "myVM"
$osVhdUri = "https://mystorageaccount.blob.core.windows.net/vhdcontainer/vhdfilename.vhd"
2. Stop/deallocate the VM.
3. Mark the VM as generalized.
Set-AzureRmVm -ResourceGroupName $rgName -Name $vmName -Generalized
4. Create the image by using your generalized OS VHD.

$imageConfig = Set-AzureRmImageOsDisk -Image $imageConfig -OsType Windows -OsState Generalized -BlobUri
$osVhdUri
$image = New-AzureRmImage -ImageName $imageName -ResourceGroupName $rgName -Image $imageConfig
Next steps
Create a VM from a managed image.
Create a VM from a managed image
You can create multiple virtual machines (VMs) from an Azure managed VM image using the Azure portal or
PowerShell. A managed VM image contains the information necessary to create a VM, including the OS and data
disks. The virtual hard disks (VHDs) that make up the image, including both the OS disks and any data disks, are
stored as managed disks.
Before creating a new VM, you'll need to create a managed VM image to use as the source image.
Use the portal

2. On the left menu, select All resources. You can sort the resources by Type to easily find your images.
3. Select the image you want to use from the list. The image Overview page opens.
4. Select Create VM from the menu.
5. Enter the virtual machine information. The user name and password entered here will be used to log in to the
virtual machine. When complete, select OK. You can create the new VM in an existing resource group, or
choose Create new to create a new resource group to store the VM.
6. Select a size for the VM. To see more sizes, select View all or change the Supported disk type filter.
7. Under Settings, make changes as necessary and select OK.
8. On the summary page, you should see your image name listed as a Private image. Select Ok to start the
virtual machine deployment.
Use PowerShell
You can use PowerShell to create a VM from an image by using the simplified parameter set for the New -
AzureRmVm cmdlet. The image needs to be in the same resource group where you'll create the VM.
This example requires the AzureRM module version 5.6.0 or later. Run Get-Module -ListAvailable AzureRM to find
the version. If you need to upgrade, see Install Azure PowerShell module.
The simplified parameter set for New -AzureRmVm only requires that you provide a name, resource group, and
image name to create a VM from an image. New -AzureRmVm will use the value of the -Name parameter as the
name of all of the resources that it creates automatically. In this example, we provide more detailed names for each
of the resources but let the cmdlet create them automatically. You can also create resources beforehand, such as
the virtual network, and pass the resource name into the cmdlet. New -AzureRmVm will use the existing resources
if it can find them by their name.
The following example creates a VM named myVMFromImage, in the myResourceGroup resource group, from
the image named myImage.
New-AzureRmVm `
-Name "myVMfromImage" `
-ImageName "myImage" `
-VirtualNetworkName "myImageVnet" `
-SubnetName "myImageSubnet" `
-SecurityGroupName "myImageNSG" `
-PublicIpAddressName "myImagePIP" `
-OpenPorts 3389
Next steps
Create and manage Windows VMs with the Azure PowerShell module
How to use Packer to create Windows virtual
machine images in Azure
Each virtual machine (VM ) in Azure is created from an image that defines the Windows distribution and OS
version. Images can include pre-installed applications and configurations. The Azure Marketplace provides many
first and third-party images for most common OS' and application environments, or you can create your own
custom images tailored to your needs. This article details how to use the open-source tool Packer to define and
build custom images in Azure.
Create Azure resource group

During the build process, Packer creates temporary Azure resources as it builds the source VM. To capture that
source VM for use as an image, you must define a resource group. The output from the Packer build process is
stored in this resource group.
Create a resource group with New -AzureRmResourceGroup. The following example creates a resource group
named myResourceGroup in the eastus location:
New-AzureRmResourceGroup -Name $rgName -Location $location
Create Azure credentials

Packer authenticates with Azure using a service principal. An Azure service principal is a security identity that you
can use with apps, services, and automation tools like Packer. You control and define the permissions as to what
operations the service principal can perform in Azure.
Create a service principal with New -AzureRmADServicePrincipal and assign permissions for the service principal
to create and manage resources with New -AzureRmRoleAssignment. Replace <password> in the example with
your own password.
$sp = New-AzureRmADServicePrincipal -DisplayName "AzurePacker" `

-Password (ConvertTo-SecureString "<password>" -AsPlainText -Force)
Sleep 20
New-AzureRmRoleAssignment -RoleDefinitionName Contributor -ServicePrincipalName $sp.ApplicationId
To authenticate to Azure, you also need to obtain your Azure tenant and subscription IDs with Get-
AzureRmSubscription:
$sub = Get-AzureRmSubscription
$sub.TenantId[0]
$sub.SubscriptionId[0]
You use these two IDs in the next step.
Define Packer template

To build images, you create a template as a JSON file. In the template, you define builders and provisioners that
carry out the actual build process. Packer has a builder for Azure that allows you to define Azure resources, such as
the service principal credentials created in the preceding step.
Create a file named windows.json and paste the following content. Enter your own values for the following:
PARAMETER WHERE TO OBTAIN
client_id View service principal ID with $sp.applicationId
client_secret Password you specified in $securePassword
tenant_id Output from $sub.TenantId command
subscription_id Output from $sub.SubscriptionId command
object_id View service principal object ID with $sp.Id
managed_image_resource_group_name Name of resource group you created in the first step
managed_image_name Name for the managed disk image that is created

{
"builders": [{
"type": "azure-arm",
"client_id": "0831b578-8ab6-40b9-a581-9a880a94aab1",
"client_secret": "P@ssw0rd!",
"tenant_id": "72f988bf-86f1-41af-91ab-2d7cd011db47",
"subscription_id": "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxx",
"object_id": "a7dfb070-0d5b-47ac-b9a5-cf214fff0ae2",
"managed_image_resource_group_name": "myResourceGroup",
"managed_image_name": "myPackerImage",
"os_type": "Windows",
"image_publisher": "MicrosoftWindowsServer",
"image_offer": "WindowsServer",
"image_sku": "2016-Datacenter",
"communicator": "winrm",
"winrm_use_ssl": true,
"winrm_insecure": true,
"winrm_timeout": "3m",
"winrm_username": "packer",
"azure_tags": {
"dept": "Engineering",
"task": "Image deployment"
},
"location": "East US",

"vm_size": "Standard_DS2_v2"
}],
"provisioners": [{
"type": "powershell",
"inline": [
"Add-WindowsFeature Web-Server",
"& $env:SystemRoot\\System32\\Sysprep\\Sysprep.exe /oobe /generalize /quiet /quit",
"while($true) { $imageState = Get-ItemProperty
HKLM:\\SOFTWARE\\Microsoft\\Windows\\CurrentVersion\\Setup\\State | Select ImageState;
if($imageState.ImageState -ne 'IMAGE_STATE_GENERALIZE_RESEAL_TO_OOBE') { Write-Output $imageState.ImageState;
Start-Sleep -s 10 } else { break } }"
]
}]
}
This template builds a Windows Server 2016 VM, installs IIS, then generalizes the VM with Sysprep. The IIS install
shows how you can use the PowerShell provisioner to run additional commands. The final Packer image then
includes the required software install and configuration.
Build Packer image

If you don't already have Packer installed on your local machine, follow the Packer installation instructions.
Build the image by specifying your Packer template file as follows:
./packer build windows.json
An example of the output from the preceding commands is as follows:
azure-arm output will be in this color.
==> azure-arm: Running builder ...

azure-arm: Creating Azure Resource Manager (ARM) client ...
azure-arm: Creating Azure Resource Manager (ARM) client ...
==> azure-arm: Creating resource group ...
==> azure-arm: -> ResourceGroupName : ‘packer-Resource-Group-pq0mthtbtt’
==> azure-arm: -> Location : ‘East US’
==> azure-arm: -> Tags :
==> azure-arm: ->> task : Image deployment
==> azure-arm: ->> dept : Engineering
==> azure-arm: Validating deployment template ...
==> azure-arm: -> DeploymentName : ‘pkrdppq0mthtbtt’
==> azure-arm: Deploying deployment template ...
==> azure-arm: Getting the certificate’s URL ...
==> azure-arm: -> Key Vault Name : ‘pkrkvpq0mthtbtt’
==> azure-arm: -> Key Vault Secret Name : ‘packerKeyVaultSecret’
==> azure-arm: -> Certificate URL :
‘https://pkrkvpq0mthtbtt.vault.azure.net/secrets/packerKeyVaultSecret/8c7bd823e4fa44e1abb747636128adbb'
==> azure-arm: Setting the certificate’s URL ...
==> azure-arm: Validating deployment template ...
==> azure-arm: Deploying deployment template ...
==> azure-arm: Getting the VM’s IP address ...
==> azure-arm: -> PublicIPAddressName : ‘packerPublicIP’
==> azure-arm: -> NicName : ‘packerNic’
==> azure-arm: -> Network Connection : ‘PublicEndpoint’
==> azure-arm: -> IP Address : ‘40.76.55.35’
==> azure-arm: Waiting for WinRM to become available...
==> azure-arm: Connected to WinRM!
==> azure-arm: Provisioning with Powershell...
==> azure-arm: Provisioning with shell script: /var/folders/h1/ymh5bdx15wgdn5hvgj1wc0zh0000gn/T/packer-
powershell-provisioner902510110
azure-arm: #< CLIXML
azure-arm:
azure-arm: Success Restart Needed Exit Code Feature Result
azure-arm: ------- -------------- --------- --------------
azure-arm: True No Success {Common HTTP Features, Default Document, D...
azure-arm: <Objs Version=“1.1.0.1” xmlns=“http://schemas.microsoft.com/powershell/2004/04"><Obj
S=“progress” RefId=“0"><TN RefId=“0”><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T>
</TN><MS><I64 N=“SourceId”>1</I64><PR N=“Record”><AV>Preparing modules for first use.</AV><AI>0</AI><Nil />
<PI>-1</PI><PC>-1</PC><T>Completed</T><SR>-1</SR><SD> </SD></PR></MS></Obj></Objs>
==> azure-arm: Querying the machine’s properties ...
==> azure-arm: -> ComputeName : ‘pkrvmpq0mthtbtt’
==> azure-arm: -> Managed OS Disk : ‘/subscriptions/guid/resourceGroups/packer-Resource-Group-
pq0mthtbtt/providers/Microsoft.Compute/disks/osdisk’
==> azure-arm: Powering off machine ...
==> azure-arm: -> ComputeName : ‘pkrvmpq0mthtbtt’
==> azure-arm: Capturing image ...
==> azure-arm: -> Compute ResourceGroupName : ‘packer-Resource-Group-pq0mthtbtt’
==> azure-arm: -> Compute Name : ‘pkrvmpq0mthtbtt’
==> azure-arm: -> Compute Location : ‘East US’
==> azure-arm: -> Image ResourceGroupName : ‘myResourceGroup’
==> azure-arm: -> Image Name : ‘myPackerImage’
==> azure-arm: -> Image Location : ‘eastus’
==> azure-arm: Deleting resource group ...
==> azure-arm: Deleting the temporary OS disk ...
==> azure-arm: -> OS Disk : skipping, managed disk was used...
Build ‘azure-arm’ finished.
==> Builds finished. The artifacts of successful builds are:

--> azure-arm: Azure.ResourceManagement.VMImage:
ManagedImageResourceGroupName: myResourceGroup
ManagedImageResourceGroupName: myResourceGroup
ManagedImageName: myPackerImage
ManagedImageLocation: eastus
It takes a few minutes for Packer to build the VM, run the provisioners, and clean up the deployment.
Create a VM from the Packer image

You can now create a VM from your Image with New -AzureRmVM. The supporting network resources are created
if they do not already exist. When prompted, enter an administrative username and password to be created on the
VM. The following example creates a VM named myVM from myPackerImage:
New-AzureRmVm `
-Name "myVM" `
-OpenPorts 80 `
-Image "myPackerImage"
If you wish to create VMs in a different resource group or region than your Packer image, specify the image ID
rather than image name. You can obtain the image ID with Get-AzureRmImage.
It takes a few minutes to create the VM from your Packer image.
Test VM and webserver

Obtain the public IP address of your VM with Get-AzureRmPublicIPAddress. The following example obtains the IP
address for myPublicIP created earlier:
-Name "myPublicIPAddress" | select "IpAddress"
To see your VM, that includes the IIS install from the Packer provisioner, in action, enter the public IP address in to
a web browser.
Next steps
In this example, you used Packer to create a VM image with IIS already installed. You can use this VM image
alongside existing deployment workflows, such as to deploy your app to VMs created from the Image with Azure
DevOps Services, Ansible, Chef, or Puppet.
For additional example Packer templates for other Windows distros, see this GitHub repo.
Use Windows client in Azure for dev/test scenarios
You can use Windows 7, Windows 8, or Windows 10 Enterprise (x64) in Azure for dev/test scenarios provided you
have an appropriate Visual Studio (formerly MSDN ) subscription. This article outlines the eligibility requirements
for running Windows 7, Windows 8.1, Windows 10 Enterprise in Azure and use of the following Azure Gallery
images.
NOTE
For Windows 10 Pro and Windows 10 Pro N image in Azure Gallery, please refer to How to deploy Windows 10 on Azure
with Multitenant Hosting Rights
Subscription eligibility
Active Visual Studio subscribers (people who have acquired a Visual Studio subscription license) can use Windows
client for development and testing purposes. Windows client can be used on your own hardware and Azure virtual
machines running in any type of Azure subscription. Windows client may not be deployed to or used on Azure for
normal production use, or used by people who are not active Visual Studio subscribers.
For your convenience, certain Windows 10 images are available from the Azure Gallery within eligible dev/test
offers. Visual Studio subscribers within any type of offer can also adequately prepare and create a 64-bit Windows
7, Windows 8, or Windows 10 image and then upload to Azure. The use remains limited to dev/test by active
Visual Studio subscribers.
Eligible offers
The following table details the offer IDs that are eligible to deploy Windows 10 through the Azure Gallery. The
Windows 10 images are only visible to the following offers. Visual Studio subscribers who need to run Windows
client in a different offer type require you to adequately prepare and create a 64-bit Windows 7, Windows 8, or
Windows 10 image and then upload to Azure.
OFFER NAME OFFER NUMBER AVAILABLE CLIENT IMAGES
Pay-As-You-Go Dev/Test 0023P Windows 10
Visual Studio Enterprise (MPN) 0029P Windows 10

subscribers
OFFER NAME OFFER NUMBER AVAILABLE CLIENT IMAGES
Visual Studio Professional subscribers 0059P Windows 10
Visual Studio Test Professional 0060P Windows 10

subscribers
Visual Studio Premium with MSDN 0061P Windows 10

(benefit)
Visual Studio Enterprise subscribers 0063P Windows 10
Visual Studio Enterprise (BizSpark) 0064P Windows 10

subscribers
Enterprise Dev/Test 0148P Windows 10
Check your Azure subscription

If you do not know your offer ID, you can obtain it through the Azure portal in one of these two ways:
On the Subscriptions window:
Or, click Billing and then click your subscription ID. The offer ID appears in the Billing window.
You can also view the offer ID from the 'Subscriptions' tab of the Azure Account portal:
Next steps
You can now deploy your VMs using PowerShell, Resource Manager templates, or Visual Studio.
Download a Windows VHD from Azure
In this article, you learn how to download a Windows virtual hard disk (VHD ) file from Azure using the Azure
portal.
Stop the VM
A VHD can’t be downloaded from Azure if it's attached to a running VM. You need to stop the VM to download a
VHD. If you want to use a VHD as an image to create other VMs with new disks, you use Sysprep to generalize the
operating system contained in the file and then stop the VM. To use the VHD as a disk for a new instance of an
existing VM or data disk, you only need to stop and deallocate the VM.
To use the VHD as an image to create other VMs, complete these steps:
1. If you haven't already done so, sign in to the Azure portal.
3. On the VM, open the Command Prompt window as an administrator.
4. Change the directory to %windir%\system32\sysprep and run sysprep.exe.
5. In the System Preparation Tool dialog box, select Enter System Out-of-Box Experience (OOBE ), and make
sure that Generalize is selected.
6. In Shutdown Options, select Shutdown, and then click OK.
To use the VHD as a disk for a new instance of an existing VM or data disk, complete these steps:
1. On the Hub menu in the Azure portal, click Virtual Machines.
2. Select the VM from the list.
3. On the blade for the VM, click Stop.
Generate SAS URL

To download the VHD file, you need to generate a shared access signature (SAS ) URL. When the URL is generated,
an expiration time is assigned to the URL.
1. On the menu of the blade for the VM, click Disks.
2. Select the operating system disk for the VM, and then click Export.
3. Set the expiration time of the URL to 36000.
4. Click Generate URL.
NOTE
The expiration time is increased from the default to provide enough time to download the large VHD file for a Windows
Server operating system. You can expect a VHD file that contains the Windows Server operating system to take several hours
to download depending on your connection. If you are downloading a VHD for a data disk, the default time is sufficient.
Download VHD
1. Under the URL that was generated, click Download the VHD file.
2. You may need to click Save in the browser to start the download. The default name for the VHD file is abcd.
Next steps
Learn how to upload a VHD file to Azure.
Create managed disks from unmanaged disks in a storage account.
Manage Azure disks with PowerShell.
2 minutes to read
Manage the availability of Windows virtual machines
in Azure
Learn ways to set up and manage multiple virtual machines to ensure high availability for your Windows
application in Azure. You can also manage the availability of Linux virtual machines.
For instructions on creating and using availability sets when using the classic deployment model, see How to
Configure an Availability Set.
Understand VM Reboots - maintenance vs. downtime

There are three scenarios that can lead to virtual machine in Azure being impacted: unplanned hardware
maintenance, unexpected downtime, and planned maintenance.
Unplanned Hardware Maintenance Event occurs when the Azure platform predicts that the hardware
or any platform component associated to a physical machine, is about to fail. When the platform predicts a
failure, it will issue an unplanned hardware maintenance event to reduce the impact to the virtual machines
hosted on that hardware. Azure uses Live Migration technology to migrate the Virtual Machines from the
failing hardware to a healthy physical machine. Live Migration is a VM preserving operation that only
pauses the Virtual Machine for a short time. Memory, open files, and network connections are maintained,
but performance might be reduced before and/or after the event. In cases where Live Migration cannot be
used, the VM will experience Unexpected Downtime, as described below.
An Unexpected Downtime is when the hardware or the physical infrastructure for the virtual machine
fails unexpectedly. This can include local network failures, local disk failures, or other rack level failures.
When detected, the Azure platform automatically migrates (heals) your virtual machine to a healthy
physical machine in the same datacenter. During the healing procedure, virtual machines experience
downtime (reboot) and in some cases loss of the temporary drive. The attached OS and data disks are
always preserved.
Virtual machines can also experience downtime in the unlikely event of an outage or disaster that affects an
entire datacenter, or even an entire region. For these scenarios, Azure provides protection options including
availability zones and paired regions.
Planned Maintenance events are periodic updates made by Microsoft to the underlying Azure platform
to improve overall reliability, performance, and security of the platform infrastructure that your virtual
machines run on. Most of these updates are performed without any impact upon your Virtual Machines or
Cloud Services (see VM Preserving Maintenance). While the Azure platform attempts to use VM
Preserving Maintenance in all possible occasions, there are rare instances when these updates require a
reboot of your virtual machine to apply the required updates to the underlying infrastructure. In this case,
you can perform Azure Planned Maintenance with Maintenance-Redeploy operation by initiating the
maintenance for their VMs in the suitable time window. For more information, see Planned Maintenance
for Virtual Machines.
To reduce the impact of downtime due to one or more of these events, we recommend the following high
availability best practices for your virtual machines:
Configure multiple virtual machines in an availability set for redundancy
Use managed disks for VMs in an availability set
Use Scheduled Events to proactively response to VM impacting events
Configure each application tier into separate availability sets
Combine a Load Balancer with availability sets
Use availability zones to protect from datacenter level failures
Configure multiple virtual machines in an availability set for

redundancy
To provide redundancy to your application, we recommend that you group two or more virtual machines in an
availability set. This configuration within a datacenter ensures that during either a planned or unplanned
maintenance event, at least one virtual machine is available and meets the 99.95% Azure SLA. For more
information, see the SLA for Virtual Machines.
IMPORTANT
Avoid leaving a single instance virtual machine in an availability set by itself. VMs in this configuration do not qualify for a
SLA guarantee and face downtime during Azure planned maintenance events, except when a single VM is using Azure
Premium Storage. For single VMs using premium storage, the Azure SLA applies.
Each virtual machine in your availability set is assigned an update domain and a fault domain by the
underlying Azure platform. For a given availability set, five non-user-configurable update domains are assigned by
default (Resource Manager deployments can then be increased to provide up to 20 update domains) to indicate
groups of virtual machines and underlying physical hardware that can be rebooted at the same time. When more
than five virtual machines are configured within a single availability set, the sixth virtual machine is placed into the
same update domain as the first virtual machine, the seventh in the same update domain as the second virtual
machine, and so on. The order of update domains being rebooted may not proceed sequentially during planned
maintenance, but only one update domain is rebooted at a time. A rebooted update domain is given 30 minutes to
recover before maintenance is initiated on a different update domain.
Fault domains define the group of virtual machines that share a common power source and network switch. By
default, the virtual machines configured within your availability set are separated across up to three fault domains
for Resource Manager deployments (two fault domains for Classic). While placing your virtual machines into an
availability set does not protect your application from operating system or application-specific failures, it does
limit the impact of potential physical hardware failures, network outages, or power interruptions.
Use managed disks for VMs in an availability set
If you are currently using VMs with unmanaged disks, we highly recommend you convert VMs in Availability Set
to use Managed Disks.
Managed disks provide better reliability for Availability Sets by ensuring that the disks of VMs in an Availability
Set are sufficiently isolated from each other to avoid single points of failure. It does this by automatically placing
the disks in different storage fault domains (storage clusters) and aligning them with the VM fault domain. If a
storage fault domain fails due to hardware or software failure, only the VM instance with disks on the storage fault
domain fails.
IMPORTANT
The number of fault domains for managed availability sets varies by region - either two or three per region. The following
table shows the number per region
Number of Fault Domains per region

REGION MAX # OF FAULT DOMAINS
East US 3
East US 2 3
West US 3
West US 2 2
Central US 3
North Central US 3
South Central US 3
West Central US 2
Canada Central 3
Canada East 2
North Europe 3
West Europe 3
UK South 2
UK West 2
East Asia 2
South East Asia 2
Japan East 2
Japan West 2
South India 2
Central India 2
West India 2
Korea Central 2
Korea South 2
Australia East 2
REGION MAX # OF FAULT DOMAINS
Australia Southeast 2
Brazil South 2
US Gov Virginia 2
US Gov Texas 2
US Gov Arizona 2
US DoD Central 2
US DoD East 2
If you plan to use VMs with unmanaged disks, follow below best practices for Storage accounts where virtual hard
disks (VHDs) of VMs are stored as page blobs.
1. Keep all disks (OS and data) associated with a VM in the same storage account
2. Review the limits on the number of unmanaged disks in a Storage account before adding more VHDs
to a storage account
3. Use separate storage account for each VM in an Availability Set. Do not share Storage accounts with
multiple VMs in the same Availability Set. It is acceptable for VMs across different Availability Sets to share
storage accounts if above best practices are followed
Configure each application tier into separate availability sets

If your virtual machines are all nearly identical and serve the same purpose for your application, we recommend
that you configure an availability set for each tier of your application. If you place two different tiers in the same
availability set, all virtual machines in the same application tier can be rebooted at once. By configuring at least
two virtual machines in an availability set for each tier, you guarantee that at least one virtual machine in each tier
is available.
For example, you could put all the virtual machines in the front end of your application running IIS, Apache, Nginx
in a single availability set. Make sure that only front-end virtual machines are placed in the same availability set.
Similarly, make sure that only data-tier virtual machines are placed in their own availability set, like your replicated
SQL Server virtual machines, or your MySQL virtual machines.
Combine a load balancer with availability sets

Combine the Azure Load Balancer with an availability set to get the most application resiliency. The Azure Load
Balancer distributes traffic between multiple virtual machines. For our Standard tier virtual machines, the Azure
Load Balancer is included. Not all virtual machine tiers include the Azure Load Balancer. For more information
about load balancing your virtual machines, see Load Balancing virtual machines.
If the load balancer is not configured to balance traffic across multiple virtual machines, then any planned
maintenance event affects the only traffic-serving virtual machine, causing an outage to your application tier.
Placing multiple virtual machines of the same tier under the same load balancer and availability set enables traffic
to be continuously served by at least one instance.
Use availability zones to protect from datacenter level failures

Availability zones, an alternative to availability sets, expand the level of control you have to maintain the
availability of the applications and data on your VMs. An Availability Zone is a physically separate zone within an
Azure region. There are three Availability Zones per supported Azure region. Each Availability Zone has a distinct
power source, network, and cooling, and is logically separate from the other Availability Zones within the Azure
region. By architecting your solutions to use replicated VMs in zones, you can protect your apps and data from the
loss of a datacenter. If one zone is compromised, then replicated apps and data are instantly available in another
zone.
Learn more about deploying a Windows or Linux VM in an Availability Zone.
Next steps
To learn more about load balancing your virtual machines, see Load Balancing virtual machines.
Vertically scale Windows VMs with Azure Automation
Vertical scaling is the process of increasing or decreasing the resources of a machine in response to the workload.
In Azure this can be accomplished by changing the size of the Virtual Machine. This can help in the following
scenarios
If the Virtual Machine is not being used frequently, you can resize it down to a smaller size to reduce your
monthly costs
If the Virtual Machine is seeing a peak load, it can be resized to a larger size to increase its capacity
The outline for the steps to accomplish this is as below
1. Setup Azure Automation to access your Virtual Machines
2. Import the Azure Automation Vertical Scale runbooks into your subscription
3. Add a webhook to your runbook
4. Add an alert to your Virtual Machine
NOTE
Because of the size of the first Virtual Machine, the sizes it can be scaled to, may be limited due to the availability of the other
sizes in the cluster current Virtual Machine is deployed in. In the published automation runbooks used in this article we take
care of this case and only scale within the below VM size pairs. This means that a Standard_D1v2 Virtual Machine will not
suddenly be scaled up to Standard_G5 or scaled down to Basic_A0.
VM SIZES SCALING PAIR
Basic_A0 Basic_A4
Standard_A0 Standard_A4
Standard_D1 Standard_D4
Standard_D11 Standard_D14
Standard_DS1 Standard_DS4
Standard_DS11 Standard_DS14
Standard_D1v2 Standard_D5v2
Standard_D11v2 Standard_D14v2
Standard_G1 Standard_G5
Standard_GS1 Standard_GS5
Setup Azure Automation to access your Virtual Machines
The first thing you need to do is create an Azure Automation account that will host the runbooks used to scale a
Virtual Machine. Recently the Automation service introduced the "Run As account" feature which makes setting up
the Service Principal for automatically running the runbooks on the user's behalf very easy. You can read more
about this in the article below:
Authenticate Runbooks with Azure Run As account
Import the Azure Automation Vertical Scale runbooks into your

subscription
The runbooks that are needed for Vertically Scaling your Virtual Machine are already published in the Azure
Automation Runbook Gallery. You will need to import them into your subscription. You can learn how to import
runbooks by reading the following article.
Runbook and module galleries for Azure Automation
The runbooks that need to be imported are shown in the image below
Add a webhook to your runbook

Once you've imported the runbooks you'll need to add a webhook to the runbook so it can be triggered by an alert
from a Virtual Machine. The details of creating a webhook for your Runbook can be read here
Azure Automation webhooks
Make sure you copy the webhook before closing the webhook dialog as you will need this in the next section.
Add an alert to your Virtual Machine

1. Select Virtual Machine settings
2. Select "Alert rules"
3. Select "Add alert"
4. Select a metric to fire the alert on
5. Select a condition, which when fulfilled will cause the alert to fire
6. Select a threshold for the condition in Step 5. to be fulfilled
7. Select a period over which the monitoring service will check for the condition and threshold in Steps 5 & 6
8. Paste in the webhook you copied from the previous section.
Create a Windows virtual machine in an availability
zone with PowerShell
This article details using Azure PowerShell to create an Azure virtual machine running Windows Server 2016 in an
Azure availability zone. An availability zone is a physically separate zone in an Azure region. Use availability zones
to protect your apps and data from an unlikely failure or loss of an entire datacenter.
To use an availability zone, create your virtual machine in a supported Azure region.
Make sure that you have installed the latest Azure PowerShell module. If you need to install or upgrade, see Install
Azure PowerShell module.
Log in to Azure
Log in to your Azure subscription with the Connect-AzureRmAccount command and follow the on-screen directions.
Check VM SKU availability

The availability of VM sizes, or SKUs, may vary by region and zone. To help you plan for the use of Availability
Zones, you can list the available VM SKUs by Azure region and zone. This ability makes sure that you choose an
appropriate VM size, and obtain the desired resiliency across zones. For more information on the different VM
types and sizes, see VM Sizes overview.
You can view the available VM SKUs with the Get-AzureRmComputeResourceSku command. The following
example lists available VM SKUs in the eastus2 region:
Get-AzureRmComputeResourceSku | where {$_.Locations.Contains("eastus2")};
The output is similar to the following condensed example, which shows the Availability Zones in which each VM
size is available:
ResourceType Name Location Zones [...]

------------ ---- -------- -----
virtualMachines Standard_DS1_v2 eastus2 {1, 2, 3}
virtualMachines Standard_DS2_v2 eastus2 {1, 2, 3}
[...]
virtualMachines Standard_F1s eastus2 {1, 2, 3}
virtualMachines Standard_F2s eastus2 {1, 2, 3}
[...]
virtualMachines Standard_D2s_v3 eastus2 {1, 2, 3}
virtualMachines Standard_D4s_v3 eastus2 {1, 2, 3}
[...]
virtualMachines Standard_E2_v3 eastus2 {1, 2, 3}
virtualMachines Standard_E4_v3 eastus2 {1, 2, 3}

Create an Azure resource group with New -AzureRmResourceGroup. A resource group is a logical container into
which Azure resources are deployed and managed. In this example, a resource group named myResourceGroup is
created in the eastus2 region.
New-AzureRmResourceGroup -Name myResourceGroup -Location EastUS2
Create networking resources

Create a virtual network, subnet, and a public IP address
These resources are used to provide network connectivity to the virtual machine and connect it to the internet.
Create the IP address in an availability zone, 2 in this example. In a later step, you create the VM in the same zone
used to create the IP address.
# Create a subnet configuration

$subnetConfig = New-AzureRmVirtualNetworkSubnetConfig -Name mySubnet -AddressPrefix 192.168.1.0/24
# Create a virtual network

$vnet = New-AzureRmVirtualNetwork -ResourceGroupName myResourceGroup -Location eastus2 `
-Name myVNet -AddressPrefix 192.168.0.0/16 -Subnet $subnetConfig
# Create a public IP address in an availability zone and specify a DNS name

$pip = New-AzureRmPublicIpAddress -ResourceGroupName myResourceGroup -Location eastus2 -Zone 2 `
-AllocationMethod Static -IdleTimeoutInMinutes 4 -Name "mypublicdns$(Get-Random)"
Create a network security group and a network security group rule

The network security group secures the virtual machine using inbound and outbound rules. In this case, an
inbound rule is created for port 3389, which allows incoming remote desktop connections. We also want to create
an inbound rule for port 80, which allows incoming web traffic.
# Create an inbound network security group rule for port 3389

$nsgRuleRDP = New-AzureRmNetworkSecurityRuleConfig -Name myNetworkSecurityGroupRuleRDP -Protocol Tcp `
-Direction Inbound -Priority 1000 -SourceAddressPrefix * -SourcePortRange * -DestinationAddressPrefix * `
-DestinationPortRange 3389 -Access Allow
# Create an inbound network security group rule for port 80

$nsgRuleWeb = New-AzureRmNetworkSecurityRuleConfig -Name myNetworkSecurityGroupRuleWWW -Protocol Tcp `
-Direction Inbound -Priority 1001 -SourceAddressPrefix * -SourcePortRange * -DestinationAddressPrefix * `
-DestinationPortRange 80 -Access Allow
# Create a network security group

$nsg = New-AzureRmNetworkSecurityGroup -ResourceGroupName myResourceGroup -Location eastus2 `
-Name myNetworkSecurityGroup -SecurityRules $nsgRuleRDP,$nsgRuleWeb
Create a network card for the virtual machine

Create a network card with New -AzureRmNetworkInterface for the virtual machine. The network card connects
the virtual machine to a subnet, network security group, and public IP address.
# Create a virtual network card and associate with public IP address and NSG
$nic = New-AzureRmNetworkInterface -Name myNic -ResourceGroupName myResourceGroup -Location eastus2 `
-SubnetId $vnet.Subnets[0].Id -PublicIpAddressId $pip.Id -NetworkSecurityGroupId $nsg.Id

Create a virtual machine configuration. This configuration includes the settings that are used when deploying the
virtual machine such as a virtual machine image, size, and authentication configuration. The Standard_DS1_v2 size
in this example is supported in availability zones. This configuration also specifies the availability zone you set
when creating the IP address. When running this step, you are prompted for credentials. The values that you enter
are configured as the user name and password for the virtual machine.
# Define a credential object

# Create a virtual machine configuration

$vmConfig = New-AzureRmVMConfig -VMName myVM -VMSize Standard_DS1_v2 -Zone 2 | `
Set-AzureRmVMOperatingSystem -Windows -ComputerName myVM -Credential $cred | `
Set-AzureRmVMSourceImage -PublisherName MicrosoftWindowsServer -Offer WindowsServer `
-Skus 2016-Datacenter -Version latest | Add-AzureRmVMNetworkInterface -Id $nic.Id
New-AzureRmVM -ResourceGroupName myResourceGroup -Location eastus2 -VM $vmConfig
Confirm zone for managed disk

You created the VM's IP address resource in the same availability zone as the VM. The managed disk resource for
the VM is created in the same availability zone. You can verify this with Get-AzureRmDisk:
Get-AzureRmDisk -ResourceGroupName myResourceGroup
The output shows that the managed disk is in the same availability zone as the VM:
ResourceGroupName : myResourceGroup
AccountType : PremiumLRS
OwnerId : /subscriptions/xxxxxxxx-xxxx-xxxx-xxxx-
xxxxxxxxxxxx/resourceGroups/myResourceGroup/providers/Microsoft.
Compute/virtualMachines/myVM
ManagedBy : /subscriptions/xxxxxxxx-xxxx-xxxx-xxxx-
xxxxxxxxxxxx//resourceGroups/myResourceGroup/providers/Microsoft.
Compute/virtualMachines/myVM
Sku : Microsoft.Azure.Management.Compute.Models.DiskSku
Zones : {2}
TimeCreated : 9/7/2017 6:57:26 PM
OsType : Windows
CreationData : Microsoft.Azure.Management.Compute.Models.CreationData
DiskSizeGB : 127
EncryptionSettings :
ProvisioningState : Succeeded
Id : /subscriptions/xxxxxxxx-xxxx-xxxx-xxxx-
xxxxxxxxxxxx/resourceGroups/myResourceGroup/providers/Microsoft.
Compute/disks/myVM_OsDisk_1_bd921920bb0a4650becfc2d830000000
Name : myVM_OsDisk_1_bd921920bb0a4650becfc2d830000000
Type : Microsoft.Compute/disks
Location : eastus2
Tags : {}
Next steps
In this article, you learned how to create a VM in an availability zone. Learn more about regions and availability for
Azure VMs.
Create a Windows virtual machine in an availability
zone with the Azure portal
This article steps through using the Azure portal to create a virtual machine in an Azure availability zone. An
availability zone is a physically separate zone in an Azure region. Use availability zones to protect your apps and
data from an unlikely failure or loss of an entire datacenter.
To use an availability zone, create your virtual machine in a supported Azure region.
Log in to Azure
Log in to the Azure portal at https://portal.azure.com.

1. Click Create a resource in the upper left-hand corner of the Azure portal.
2. Select Compute, and then select Windows Server 2016 Datacenter.
3. Enter the virtual machine information. The user name and password entered here is used to log in to the
virtual machine. The password must be at least 12 characters long and meet the defined complexity
requirements. Choose a Location such as East US 2 that supports availability zones. When complete, click
OK.
4. Choose a size for the VM. Select a recommended size, or filter based on features. Confirm the size is
available in the zone you want to use.
5. Under Settings > High availability, select one of the numbered zones from the Availability zone
dropdown, keep the remaining defaults, and click OK.
6. On the summary page, click Create to start the virtual machine deployment.
7. The VM will be pinned to the Azure portal dashboard. Once the deployment has completed, the VM
summary automatically opens.
Confirm zone for managed disk and IP address

When the VM is deployed in an availability zone, a managed disk for the VM is created in the same availability
zone. By default, a public IP address is also created in that zone.
You can confirm the zone settings for these resources in the portal.
1. Click Resource groups and then the name of the resource group for the VM, such as myResourceGroup.
2. Click the name of the Disk resource. The Overview page includes details about the location and availability
zone of the resource.
3. Click the name of the Public IP address resource. The Overview page includes details about the location
and availability zone of the resource.
Next steps
In this article, you learned how to create a VM in an availability zone. Learn more about regions and availability for
Azure VMs.
Automating Azure virtual machine deployment with
Chef
NOTE
Chef is a great tool for delivering automation and desired state configurations.
With the latest cloud API release, Chef provides seamless integration with Azure, giving you the ability to provision
and deploy configuration states through a single command.
In this article, you set up your Chef environment to provision Azure virtual machines and walk through creating a
policy or “CookBook” and then deploying this cookbook to an Azure virtual machine.
Let’s begin!
Chef basics
Before you begin, review the basic concepts of Chef.
The following diagram depicts the high-level Chef architecture.
Chef has three main architectural components: Chef Server, Chef Client (node), and Chef Workstation.
The Chef Server is the management point and there are two options for the Chef Server: a hosted solution or an
on-premises solution. We will be using a hosted solution for this tutorial.
The Chef Client (node) is the agent that sits on the servers you are managing.
The Chef Workstation is the admin workstation where we create policies and execute management commands. We
run the knife command from the Chef Workstation to manage the infrastructure.
There is also the concept of “Cookbooks” and “Recipes”. These are effectively the policies we define and apply to
the servers.
Preparing the workstation

First, lets prep the workstation. I’m using a standard Windows workstation. We need to create a directory to store
the config files and cookbooks.
First create a directory called C:\chef.
Then create a second directory called c:\chef\cookbooks.
We now need to download the Azure settings file so Chef can communicate with the Azure subscription.
Download your publish settings using the PowerShell Azure Get-AzurePublishSettingsFile command.
Save the publish settings file in C:\chef.
Creating a managed Chef account
Sign up for a hosted Chef account here.
During the signup process, you will be asked to create a new organization.
Once your organization is created, download the starter kit.
NOTE
If you receive a prompt warning you that your keys will be reset, it’s okay to proceed as we have no existing infrastructure
configured as yet.
This starter kit zip file contains your organization config files and keys.
Configuring the Chef workstation

Extract the content of the chef-starter.zip to C:\chef.
Copy all files under chef-starter\chef-repo.chef to your c:\chef directory.
Your directory should now look something like the following example.
You should now have four files including the Azure publishing file in the root of c:\chef.
The PEM files contain your organization and admin private keys for communication while the knife.rb file contains
your knife configuration. We will need to edit the knife.rb file.
Open the file in your editor of choice and modify the “cookbook_path” by removing the /../ from the path so it
appears as:
cookbook_path ["#{current_dir}/cookbooks"]
Also add the following line reflecting the name of your Azure publish settings file.
knife[:azure_publish_settings_file] = "yourfilename.publishsettings"
Your knife.rb file should now look similar to the following example.
These lines will ensure that Knife references the cookbooks directory under c:\chef\cookbooks, and also uses our
Azure Publish Settings file during Azure operations.
Installing the Chef Development Kit

Next, download and install the ChefDK (Chef Development Kit) to set up your Chef Workstation.
Install in the default location of c:\opscode. This install will take around 10 minutes.
Confirm your PATH variable contains entries for
C:\opscode\chefdk\bin;C:\opscode\chefdk\embedded\bin;c:\users\yourusername.chefdk\gem\ruby\2.0.0\bin
If they are not there, make sure you add these paths!
NOTE
The order of the path is important! If your opscode paths are not in the correct order you will have issues.
Reboot your workstation before you continue.

Next, we will install the Knife Azure extension. This provides Knife with the “Azure Plugin”.
Run the following command.
chef gem install knife-azure ––pre
NOTE
The –pre argument ensures you are receiving the latest RC version of the Knife Azure Plugin which provides access to the
latest set of APIs.
It’s likely that a number of dependencies will also be installed at the same time.
To ensure everything is configured correctly, run the following command.
knife azure image list
If everything is configured correctly, you will see a list of available Azure images scroll through.
Congratulations. The workstation is set up!
Creating a Cookbook
A Cookbook is used by Chef to define a set of commands that you wish to execute on your managed client.
Creating a Cookbook is straightforward and we use the chef generate cookbook command to generate the
Cookbook template. I will be calling my Cookbook web server as I would like a policy that automatically deploys
IIS.
Under your C:\Chef directory run the following command.
chef generate cookbook webserver
This will generate a set of files under the directory C:\Chef\cookbooks\webserver. We now need to define the set
of commands we would like the Chef client to execute on the managed virtual machine.
The commands are stored in the file default.rb. In this file, I’ll be defining a set of commands that installs IIS, starts
IIS and copies a template file to the wwwroot folder.
Modify the C:\chef\cookbooks\webserver\recipes\default.rb file and add the following lines.
powershell_script 'Install IIS' do

action :run
code 'add-windowsfeature Web-Server'
end
service 'w3svc' do
action [ :enable, :start ]
end
template 'c:\inetpub\wwwroot\Default.htm' do
source 'Default.htm.erb'
rights :read, 'Everyone'
end
Save the file once you are done.

Creating a template
As we mentioned previously, we need to generate a template file which will be used as the default.html page.
Run the following command to generate the template.
chef generate template webserver Default.htm
Now navigate to the C:\chef\cookbooks\webserver\templates\default\Default.htm.erb file. Edit the file by adding

some simple “Hello World” HTML code, and then save the file.
Upload the Cookbook to the Chef Server

In this step, we are taking a copy of the Cookbook that we have created on the local machine and uploading it to
the Chef Hosted Server. Once uploaded, the Cookbook will appear under the Policy tab.
knife cookbook upload webserver
Deploy a virtual machine with Knife Azure

We will now deploy an Azure virtual machine and apply the “Webserver” Cookbook which will install the IIS web
service and default web page.
In order to do this, use the knife azure server create command.
Am example of the command appears next.
knife azure server create --azure-dns-name 'diegotest01' --azure-vm-name 'testserver01' --azure-vm-size

'Small' --azure-storage-account 'portalvhdsxxxx' --bootstrap-protocol 'cloud-api' --azure-source-image
'a699494373c04fc0bc8f2bb1389d6106__Windows-Server-2012-Datacenter-201411.01-en.us-127GB.vhd' --azure-service-
location 'Southeast Asia' --winrm-user azureuser --winrm-password 'myPassword123' --tcp-endpoints 80,3389 --r
'recipe[webserver]'
The parameters are self-explanatory. Substitute your particular variables and run.
NOTE
Through the command line, I’m also automating my endpoint network filter rules by using the –tcp-endpoints parameter. I’ve
opened up ports 80 and 3389 to provide access to my web page and RDP session.
Once you run the command, go to the Azure portal and you will see your machine begin to provision.
The command prompt appears next.

Once the deployment is complete, we should be able to connect to the web service over port 80 as we had opened
the port when we provisioned the virtual machine with the Knife Azure command. As this virtual machine is the
only virtual machine in my cloud service, I’ll connect it with the cloud service url.
As you can see, I got creative with my HTML code.

Don’t forget we can also connect through an RDP session from the Azure portal via port 3389.
I hope this has been helpful! Go and start your infrastructure as code journey with Azure today!
Publish an ASP.NET Web App to an Azure VM from
Visual Studio
This document describes how to publish an ASP.NET web application to an Azure virtual machine (VM ) using the
Microsoft Azure Virtual Machines publishing feature in Visual Studio 2017.
Prerequisites
In order to use Visual Studio to publish an ASP.NET project to an Azure VM, the VM must be correctly set up.
Machine must be configured to run an ASP.NET web application and have WebDeploy installed.
The VM must have a DNS name configured. For more information, see Create a fully qualified domain
name in the Azure portal for a Windows VM.
Publish your ASP.NET web app to the Azure VM using Visual Studio
The following section describes how to publish an existing ASP.NET web application to an Azure virtual machine.
1. Open your web app solution in Visual Studio 2017.
2. Right-click the project in Solution Explorer and choose Publish...
3. Use the arrow on the right of the page to scroll through the publishing options until you find Microsoft
Azure Virtual Machines.
4. Select the Microsoft Azure Virtual Machines icon and select Publish.
5. Choose the appropriate account (with Azure subscription connected to your virtual machine).
If you're signed in to Visual Studio, the account list is populated with all your authenticated accounts.
If you are not signed in, or if the account you need is not listed, choose "Add an account..." and follow the
prompts to log in.
6. Select the appropriate VM from the list of Existing Virtual Machines.
NOTE
Populating this list can take some time.
7. Click OK to begin publishing.
8. When prompted for credentials, supply the username and password of a user account on the target VM that
is configured with publishing rights (typically the admin username and password used when creating the
VM ).
9. Accept the security certificate.
10. Watch the Output window to check the progress of the publish operation.
11. If publishing is successful, a browser launches to open the URL of the newly published site.
Success!
You have now successfully published your web app to an Azure virtual machine.
Publish Page Options

After completing the publish wizard, the Publish page is opened in the document well with the new publishing
profile selected.
Re -publish
To publish updates to your web application, select the Publish button on the Publish page.
If prompted, enter username and password.
Publishing begins immediately.
Modify publish profile settings

To view and modify the publish profile settings, select Settings....
Your settings should look something like this:

Save User name and Password
To avoid providing authentication information every time you publish, you can populate the User name and
Password fields and select the Save password box.
Use the Validate Connection button to confirm that you have entered the right information.
Deploy to clean web server
If you want to ensure that the web server has a clean copy of the web application after each upload (and that
no other files are left hanging around from a previous deployment), you can check the Remove additional
files at destination checkbox in the Settings tab.
Warning: Publishing with this setting deletes all files that exist on the web server (wwwroot directory). Be
sure you know the state of the machine before publishing with this option enabled.
Next steps
Set up CI/CD for automated deployment to Azure VM
To set up a continuous delivery pipeline with Azure Pipelines, see Deploy to a Windows Virtual Machine.
What is SQL Server on Azure Virtual Machines?
(Windows)
SQL Server on Azure virtual machines enables you to use full versions of SQL Server in the Cloud without having
to manage any on-premises hardware. SQL Server VMs also simplify licensing costs when you pay as you go.
Azure virtual machines run in many different geographic regions around the world. They also offer a variety of
machine sizes. The virtual machine image gallery allows you to create a SQL Server VM with the right version,
edition, and operating system. This makes virtual machines a good option for a many different SQL Server
workloads.
Automated updates
SQL Server Azure VMs can use Automated Patching to schedule a maintenance window for installing important
windows and SQL Server updates automatically.
Automated backups
SQL Server Azure VMs can take advantage of Automated Backup, which regularly creates backups of your
database to blob storage. You can also manually use this technique. For more information, see Use Azure Storage
for SQL Server Backup and Restore.
High availability
If you require high availability, consider configuring SQL Server Availability Groups. This involves multiple SQL
Server Azure VMs in a virtual network. You can configure your high availability solution manually, or you can use
templates in the Azure portal for automatic configuration. For an overview of all high availability options, see High
Availability and Disaster Recovery for SQL Server in Azure Virtual Machines.
Performance
Azure virtual machines offer different machine sizes to meet various workload demands. SQL VMs also provide
automated storage configuration, which is optimized for your performance requirements. For more information
about configuring storage for SQL VMs, see Storage configuration for SQL Server VMs. To fine-tune performance,
see the Performance best practices for SQL Server in Azure Virtual Machines.
Get started with SQL VMs

To get started, choose a SQL Server virtual machine image with your required version, edition, and operating
system. The following sections provide direct links to the Azure portal for the SQL Server virtual machine gallery
images.
TIP
For more information about how to understand pricing for SQL images, see Pricing guidance for SQL Server Azure VMs.
Pay as you go
The following table provides a matrix of pay-as-you-go SQL Server images.
VERSION OPERATING SYSTEM EDITION
SQL Server 2017 Windows Server 2016 Enterprise, Standard, Web, Express,
Developer
SQL Server 2016 SP1 Windows Server 2016 Enterprise, Standard, Web, Express,
Developer
SQL Server 2014 SP2 Windows Server 2012 R2 Enterprise, Standard, Web, Express
SQL Server 2012 SP4 Windows Server 2012 R2 Enterprise, Standard, Web, Express
SQL Server 2008 R2 SP3 Windows Server 2008 R2 Enterprise, Standard, Web, Express
To see the available Linux SQL Server virtual machine images, see Overview of SQL Server on Azure Virtual
Machines (Linux).
Bring your own license
You can also bring your own license (BYOL ). In this scenario, you only pay for the VM without any additional
charges for SQL Server licensing. Bringing your own license can save you money over time for continuous
production workloads. For requirements to use this option, see Pricing guidance for SQL Server Azure VMs.
VERSION OPERATING SYSTEM EDITION
SQL Server 2017 Windows Server 2016 Enterprise BYOL, Standard BYOL
SQL Server 2016 SP1 Windows Server 2016 Enterprise BYOL, Standard BYOL
SQL Server 2014 SP2 Windows Server 2012 R2 Enterprise BYOL, Standard BYOL
SQL Server 2012 SP4 Windows Server 2012 R2 Enterprise BYOL, Standard BYOL
In the portal, these image names are prefixed with {BYOL }.

Connect to the VM
After creating your SQL Server VM, connect to it from applications or tools, such as SQL Server Management
Studio (SSMS ). For instructions, see Connect to a SQL Server Virtual Machine on Azure.
Migrate your data
If you have an existing database, you'll want to move that to the newly provisioned SQL VM. For a list of migration
options and guidance, see Migrating a Database to SQL Server on an Azure VM.
SQL VM image refresh policy

Azure only maintains one virtual machine image for each supported operating system, version, and edition
combination. This means that over time images are refreshed, and older images are removed. For more
information, see the Images section of the SQL Server VMs FAQ.
Customer experience improvement program (CEIP)

The Customer Experience Improvement Program (CEIP ) is enabled by default. This periodically sends reports to
Microsoft to help improve SQL Server. There is no management task required with CEIP unless you want to
disable it after provisioning. You can customize or disable the CEIP by connecting to the VM with remote desktop.
Then run the SQL Server Error and Usage Reporting utility. Follow the instructions to disable reporting. For
more information about data collection, see the SQL Server Privacy Statement.
Related products and services

Windows Virtual Machines
Virtual Machines overview
Storage
Introduction to Microsoft Azure Storage
Networking
Virtual Network overview
IP addresses in Azure
Create a Fully Qualified Domain Name in the Azure portal
SQL
SQL Server documentation
Azure SQL Database comparison
Next steps
Get started with SQL Server on Azure virtual machines:
Create a SQL Server VM in the Azure portal
Get answers to commonly asked questions about SQL VMs:
SQL Server on Azure Virtual Machines FAQ
Install and configure MongoDB on a Windows VM in
Azure
MongoDB is a popular open-source, high-performance NoSQL database. This article guides you through installing
and configuring MongoDB on a Windows Server 2016 virtual machine (VM ) in Azure. You can also install
MongoDB on a Linux VM in Azure.
Prerequisites
Before you install and configure MongoDB, you need to create a VM and, ideally, add a data disk to it. See the
following articles to create a VM and add a data disk:
Create a Windows Server VM using the Azure portal or Azure PowerShell.
Attach a data disk to a Windows Server VM using the Azure portal or Azure PowerShell.
To begin installing and configuring MongoDB, log on to your Windows Server VM by using Remote Desktop.
Install MongoDB
IMPORTANT
MongoDB security features, such as authentication and IP address binding, are not enabled by default. Security features
should be enabled before deploying MongoDB to a production environment. For more information, see MongoDB Security
and Authentication.
1. After you've connected to your VM using Remote Desktop, open Internet Explorer from the taskbar.
2. Select Use recommended security, privacy, and compatibility settings when Internet Explorer first opens,
and click OK.
3. Internet Explorer enhanced security configuration is enabled by default. Add the MongoDB website to the
list of allowed sites:
Select the Tools icon in the upper-right corner.
In Internet Options, select the Security tab, and then select the Trusted Sites icon.
Click the Sites button. Add https://*.mongodb.com to the list of trusted sites, and then close the
dialog box.
4. Browse to the MongoDB - Downloads page (http://www.mongodb.com/downloads).
5. If needed, select the Community Server edition and then select the latest current stable release
forWindows Server 2008 R2 64 -bit and later. To download the installer, click DOWNLOAD (msi).
Run the installer after the download is complete.

6. Read and accept the license agreement. When you're prompted, select Complete install.
7. If desired, you can choose to also install Compass, a graphical interface for MongoDB.
8. On the final screen, click Install.
Configure the VM and MongoDB

1. The path variables are not updated by the MongoDB installer. Without the MongoDB bin location in your
path variable, you need to specify the full path each time you use a MongoDB executable. To add the
location to your path variable:
Right-click the Start menu, and select System.
Click Advanced system settings, and then click Environment Variables.
Under System variables, select Path, and then click Edit.
Add the path to your MongoDB bin folder. MongoDB is typically installed in C:\Program
Files\MongoDB. Verify the installation path on your VM. The following example adds the default
MongoDB install location to the PATH variable:
;C:\Program Files\MongoDB\Server\3.6\bin
NOTE
Be sure to add the leading semicolon ( ; ) to indicate that you are adding a location to your PATH variable.
2. Create MongoDB data and log directories on your data disk. From the Start menu, select Command
Prompt. The following examples create the directories on drive F:
mkdir F:\MongoData
mkdir F:\MongoLogs
3. Start a MongoDB instance with the following command, adjusting the path to your data and log directories
accordingly:
mongod --dbpath F:\MongoData\ --logpath F:\MongoLogs\mongolog.log
It may take several minutes for MongoDB to allocate the journal files and start listening for connections. All
log messages are directed to the F:\MongoLogs\mongolog.log file as mongod.exe server starts and allocates
journal files.
NOTE
The command prompt stays focused on this task while your MongoDB instance is running. Leave the command
prompt window open to continue running MongoDB. Or, install MongoDB as service, as detailed in the next step.
4. For a more robust MongoDB experience, install the mongod.exe as a service. Creating a service means you
don't need to leave a command prompt running each time you want to use MongoDB. Create the service as
follows, adjusting the path to your data and log directories accordingly:
mongod --dbpath F:\MongoData\ --logpath F:\MongoLogs\mongolog.log --logappend --install
The preceding command creates a service named MongoDB, with a description of "Mongo DB". The
following parameters are also specified:
The --dbpath option specifies the location of the data directory.
The --logpath option must be used to specify a log file, because the running service does not have a
command window to display output.
The --logappend option specifies that a restart of the service causes output to append to the existing log
file.
To start the MongoDB service, run the following command:
net start MongoDB
For more information about creating the MongoDB service, see Configure a Windows Service for
MongoDB.
Test the MongoDB instance

With MongoDB running as a single instance or installed as a service, you can now start creating and using your
databases. To start the MongoDB administrative shell, open another command prompt window from the Start
menu, and enter the following command:
mongo
You can list the databases with the db command. Insert some data as follows:
db.foo.insert( { a : 1 } )
Search for data as follows:
db.foo.find()
The output is similar to the following example:
{ "_id" : "ObjectId("57f6a86cee873a6232d74842"), "a" : 1 }
Exit the mongo console as follows:
exit
Configure firewall and Network Security Group rules

Now that MongoDB is installed and running, open a port in Windows Firewall so you can remotely connect to
MongoDB. To create a new inbound rule to allow TCP port 27017, open an administrative PowerShell prompt and
enter the following command:
New-NetFirewallRule `
-DisplayName "Allow MongoDB" `
-Protocol TCP `
-LocalPort 27017 `
-Action Allow
You can also create the rule by using the Windows Firewall with Advanced Security graphical management
tool. Create a new inbound rule to allow TCP port 27017.
If needed, create a Network Security Group rule to allow access to MongoDB from outside of the existing Azure
virtual network subnet. You can create the Network Security Group rules by using the Azure portal or Azure
PowerShell. As with the Windows Firewall rules, allow TCP port 27017 to the virtual network interface of your
MongoDB VM.
NOTE
TCP port 27017 is the default port used by MongoDB. You can change this port by using the --port parameter when
starting mongod.exe manually or from a service. If you change the port, make sure to update the Windows Firewall and
Network Security Group rules in the preceding steps.
Next steps
In this tutorial, you learned how to install and configure MongoDB on your Windows VM. You can now access
MongoDB on your Windows VM, by following the advanced topics in the MongoDB documentation.
Using Azure for hosting and running SAP workload
scenarios
By choosing Microsoft Azure as your SAP ready cloud partner, you are able to reliably run your mission critical
SAP workloads and scenarios on a scalable, compliant, and enterprise-proven platform. Get the scalability,
flexibility, and cost savings of Azure. With the expanded partnership between Microsoft and SAP, you can run SAP
applications across dev/test and production scenarios in Azure - and be fully supported. From SAP NetWeaver to
SAP S4/HANA, SAP BI, Linux to Windows, SAP HANA to SQL, we have you covered.
Besides hosting SAP NetWeaver scenarios with the different DBMS on Azure, you can host different other SAP
workload scenarios, like SAP BI on Azure. Documentation regarding SAP NetWeaver deployments on Azure
native Virtual Machines can be found in the section "SAP NetWeaver on Azure Virtual Machines."
The uniqueness of Azure for SAP HANA is a unique offer that sets Azure apart from competition. In order to
enable hosting more memory and CPU resource demanding SAP scenarios involving SAP HANA, Azure offers the
usage of customer dedicated bare-metal hardware for the purpose of running SAP HANA deployments that
require up to 20 TB (60 TB scale-out) of memory for S/4HANA or other SAP HANA workload. This unique Azure
solution of SAP HANA on Azure (Large Instances) allows you to run SAP HANA on the dedicated bare-metal
hardware with the SAP application layer or workload middle-ware layer hosted in native Azure Virtual Machines.
This solution is documented in several documents in the section "SAP HANA on Azure (Large Instances)."
Hosting SAP workload scenarios in Azure also can create requirements of Identity integration and Single-Sign-On
using Azure Activity Directory to different SAP components and SAP SaaS or PaaS offers. A list of such integration
and Single-Sign-On scenarios with Azure Active Directory (AAD ) and SAP entities is described and documented in
the section "AAD SAP Identity Integration and Single-Sign-On."
Latest Changes
Documentation around SAP HANA Dynamic Tiering for Azure VMs
SAP HANA infrastructure configurations and operations on Azure
Documentation around SAP HANA Scale-out on Azure VM M128s got added to:
SAP HANA availability within one Azure region
SAP HANA on SAP HANA on Azure (Large Instances)

A series of documentation leads you through SAP HANA on Azure (Large Instances) or in short HANA Large
Instances. The documents cover the listed areas of HANA Large Instances:
Overview of SAP HANA on Azure (Large Instances)
Architecture of SAP HANA on Azure (Large Instances)
Infrastructure and Connectivity to SAP HANA on Azure (Large Instances)
Install SAP HANA on SAP HANA on Azure (Large Instances)
High Availability and Disaster Recovery of SAP HANA on Azure (Large Instances)
Troubleshooting and Monitoring of SAP HANA on Azure (Large Instances)
Next steps:
Read Overview and Architecture of SAP HANA on Azure (Large Instances)
SAP HANA on Azure Virtual Machines

This section of the documentation covers different aspects of SAP HANA. As a prerequisite, you should be familiar
with the principal services of Azure that provide elementary services of Azure IaaS, so mostly knowledge of Azure
compute, storage, and networking. A lot of those topics are handled in the SAP NetWeaver related Azure Planning
Guide.
The documentation specific for HANA on Azure consists of this list of articles and their sub-articles:
Quickstart: Manual installation of single-instance SAP HANA on Azure VMs
Deploy SAP S/4HANA or BW/4HANA on Azure
SAP HANA high availability for Azure virtual machines
SAP HANA availability within one Azure region
SAP HANA availability across Azure regions
High availability of SAP HANA on Azure virtual machines
Backup guide for SAP HANA on Azure Virtual Machines
SAP HANA Azure Backup on file level
SAP HANA backup based on storage snapshots
SAP NetWeaver deployed on Azure Virtual Machines

In this section you find planning and deployment documentation for SAP NetWeaver and Business One on Azure.
The documentation in this chapter is focused mostly around the basics and the usage of non-HANA databases with
SAP workload on Azure. Whereas the documents and articles for HA are foundation for HANA high availability in
Azure as well. the articles list like:
SAP Business One on Azure Virtual Machines
Deploy SAP IDES EHP7 SP3 for SAP ERP 6.0 on Azure
Running SAP NetWeaver on Microsoft Azure SUSE Linux VMs
Azure Virtual Machines planning and implementation for SAP NetWeaver
Azure Virtual Machines deployment for SAP NetWeaver
Protect a multi-tier SAP NetWeaver application deployment by using Site Recovery
SAP LaMa connector for Azure
Regarding non-HANA databases under SAP workload on azure the documents list like:
Considerations for Azure Virtual Machines DBMS deployment for SAP workload
SQL Server Azure Virtual Machines DBMS deployment for SAP NetWeaver
Oracle Azure Virtual Machines DBMS deployment for SAP workload
IBM DB2 Azure Virtual Machines DBMS deployment for SAP workload
SAP ASE Azure Virtual Machines DBMS deployment for SAP workload
SAP MaxDB, liveCache, and Content Server deployment on Azure VMs
For SAP HANA databases on Azure, check the section SAP HANA on Azure Virtual Machines.
For high availability of SAP workload on Azure the entry document is:
Azure Virtual Machines high availability for SAP NetWeaver
The entry document points to various other architecture and scenario documents. In subsequent scenario
documents, links to detailed technical documents explaining the deployment and configuration of the different high
availability methods are provided. The different documents of establishing and configuring high availability for
SAP NetWeaver workload are covering Linux as well as Windows operating systems.
For integration between Azure Active Directory (AAD ) and SAP Services and Single-Sign-On, the documents list
like:
Tutorial: Azure Active Directory integration with SAP Cloud for Customer
Tutorial: Azure Active Directory integration with SAP Cloud Platform Identity Authentication
Tutorial: Azure Active Directory integration with SAP Cloud Platform
Tutorial: Azure Active Directory integration with SAP NetWeaver
Tutorial: Azure Active Directory integration with SAP Business ByDesign
Tutorial: Azure Active Directory integration with SAP HANA
Your S/4HANA environment – Fiori Launchpad SAML Single Sign-On with Azure AD
For integration of Azure Services into SAP components the list of documents looks like:
Use SAP HANA in Power BI Desktop
DirectQuery and SAP HANA
Use the SAP BW Connector in Power BI Desktop
Azure Data Factory offers SAP HANA and Business Warehouse data integration
Create MATLAB Distributed Computing Server
clusters on Azure VMs
Use Microsoft Azure virtual machines to create one or more MATLAB Distributed Computing Server clusters to
run your compute-intensive parallel MATLAB workloads. Install your MATLAB Distributed Computing Server
software on a VM to use as a base image and use an Azure quickstart template or Azure PowerShell script
(available on GitHub) to deploy and manage the cluster. After deployment, connect to the cluster to run your
workloads.
About MATLAB and MATLAB Distributed Computing Server

The MATLAB platform is optimized for solving engineering and scientific problems. MATLAB users with large-
scale simulations and data processing tasks can use MathWorks parallel computing products to speed up their
compute-intensive workloads by taking advantage of compute clusters and grid services. Parallel Computing
Toolbox lets MATLAB users parallelize applications and take advantage of multi-core processors, GPUs, and
compute clusters. MATLAB Distributed Computing Server enables MATLAB users to utilize many computers in a
compute cluster.
By using Azure virtual machines, you can create MATLAB Distributed Computing Server clusters that have all the
same mechanisms available to submit parallel work as on-premises clusters, such as interactive jobs, batch jobs,
independent tasks, and communicating tasks. Using Azure in conjunction with the MATLAB platform has many
benefits compared to provisioning and using traditional on-premises hardware: a range of virtual machine sizes,
creation of clusters on-demand so you pay only for the compute resources you use, and the ability to test models at
scale.
Prerequisites
Client computer - You'll need a Windows-based client computer to communicate with Azure and the MATLAB
Distributed Computing Server cluster after deployment.
Azure PowerShell - See How to install and configure Azure PowerShell to install it on your client computer.
Azure subscription - If you don't have a subscription, you can create a free account in just a couple of minutes.
For larger clusters, consider a pay-as-you-go subscription or other purchase options.
vCPUs quota - You might need to increase the vCPU quota to deploy a large cluster or more than one
MATLAB Distributed Computing Server cluster. To increase a quota, open an online customer support request
at no charge.
MATLAB, Parallel Computing Toolbox, and MATLAB Distributed Computing Server licenses - The
scripts assume that the MathWorks Hosted License Manager is used for all licenses.
MATLAB Distributed Computing Server software - Will be installed on a VM that will be used as the base
VM image for the cluster VMs.
High level steps

To use Azure virtual machines for your MATLAB Distributed Computing Server clusters, the following high-level
steps are required. Detailed instructions are in the documentation accompanying the quickstart template and
scripts on GitHub.
1. Create a base VM image
Download and install MATLAB Distributed Computing Server software onto this VM.
NOTE
This process can take a couple of hours, but you only have to do it once for each version of MATLAB you use.
2. Create one or more clusters

Use the supplied PowerShell script or use the quickstart template to create a cluster from the base VM
image.
Manage the clusters using the supplied PowerShell script which allows you to list, pause, resume, and
delete clusters.
Cluster configurations
Currently, the cluster creation script and template enable you to create a single MATLAB Distributed Computing
Server topology. If you want, create one or more additional clusters, with each cluster having a different number of
worker VMs, using different VM sizes, and so on.
MATLAB client and cluster in Azure
The MATLAB client node, MATLAB Job Scheduler node, and MATLAB Distributed Computing Server "worker"
nodes are all configured as Azure VMs in a virtual network, as shown in the following figure.
To use the cluster, connect by Remote Desktop to the client node. The client node runs the MATLAB client.
The client node has a file share that can be accessed by all workers.
MathWorks Hosted License Manager is used for the license checks for all MATLAB software.
By default, one MATLAB Distributed Computing Server worker per vCPU is created on the worker VMs, but
you can specify any number.
Use an Azure-based Cluster

As with other types of MATLAB Distributed Computing Server clusters, you need to use the Cluster Profile
Manager in the MATLAB client (on the client VM ) to create a MATLAB Job Scheduler cluster profile.
Next steps
For detailed instructions to deploy and manage MATLAB Distributed Computing Server clusters in Azure, see
the GitHub repository containing the templates and scripts.
Go to the MathWorks site for detailed documentation for MATLAB and MATLAB Distributed Computing
Server.
Visual Studio images on Azure
Using Visual Studio in a preconfigured Azure virtual machine (VM ) is a quick, easy way to go from nothing to an
up-and-running development environment. System images with different Visual Studio configurations are
available in the Azure Marketplace.
New to Azure? Create a free Azure account.
What configurations and versions are available?

Images for the most recent major versions, Visual Studio 2017 and Visual Studio 2015, can be found in the Azure
Marketplace. For each major version, you see the originally released (RTW ) version and the latest updated versions.
Each of these versions offers the Visual Studio Enterprise and the Visual Studio Community editions. These images
are updated at least every month to include the latest Visual Studio and Windows updates. While the names of the
images remain the same, each image's description includes the installed product version and the image's "as of"
date.
RELEASE VERSION EDITIONS PRODUCT VERSION
Visual Studio 2017: Latest (Version Enterprise, Community Version 15.8.4

15.8)
Visual Studio 2017: Latest Preview Enterprise, Community Version 15.9.0

(Version 15.9, Preview 2)
Visual Studio 2017: RTW Enterprise, Community Version 15.0.18
Visual Studio 2015: Latest (Update 3) Enterprise, Community Version 14.0.25431.01
Visual Studio 2015: RTW None (Expired for servicing)
NOTE
In accordance with Microsoft servicing policy, the originally released (RTW) version of Visual Studio 2015 has expired for
servicing. Visual Studio 2015 Update 3 is the only remaining version offered for the Visual Studio 2015 product line.
For more information, see the Visual Studio Servicing Policy.
What features are installed?

Each image contains the recommended feature set for that Visual Studio edition. Generally, the installation
includes:
All available workloads, including each workload’s recommended optional components
.NET 4.6.2 and .NET 4.7 SDKs, Targeting Packs, and Developer Tools
Visual F#
GitHub Extension for Visual Studio
LINQ to SQL Tools
The command line used to install Visual Studio when building the images is as follows:
vs_enterprise.exe --allWorkloads --includeRecommended --passive ^

add Microsoft.Net.Component.4.7.SDK ^
add Microsoft.Net.Component.4.7.TargetingPack ^
add Microsoft.Net.Component.4.6.2.SDK ^
add Microsoft.Net.Component.4.6.2.TargetingPack ^
add Microsoft.Net.ComponentGroup.4.7.DeveloperTools ^
add Microsoft.VisualStudio.Component.FSharp ^
add Component.GitHub.VisualStudio ^
add Microsoft.VisualStudio.Component.LinqToSql
If the images don't include a Visual Studio feature that you require, provide feedback through the feedback tool in
the upper-right corner of the page.
What size VM should I choose?

Azure offers a full range of virtual machine sizes. Because Visual Studio is a powerful, multi-threaded application,
you want a VM size that includes at least two processors and 7 GB of memory. We recommend the following VM
sizes for the Visual Studio images:
Standard_D2_v3
Standard_D2s_v3
Standard_D4_v3
Standard_D4s_v3
Standard_D2_v2
Standard_D2S_v2
Standard_D3_v2
For more information on the latest machine sizes, see Sizes for Windows virtual machines in Azure.
With Azure, you can rebalance your initial choice by resizing the VM. You can either provision a new VM with a
more appropriate size, or resize your existing VM to different underlying hardware. For more information, see
Resize a Windows VM.
After the VM is running, what's next?

Visual Studio follows the “bring your own license” model in Azure. As with an installation on proprietary hardware,
one of the first steps is licensing your Visual Studio installation. To unlock Visual Studio, either:
Sign in with a Microsoft account that’s associated with a Visual Studio subscription
Unlock Visual Studio with the product key that came with your initial purchase
For more information, see Sign in to Visual Studio and How to unlock Visual Studio.
How do I save the development VM for future or team use?

The spectrum of development environments is huge, and there’s real cost associated with building out the more
complex environments. Regardless of your environment’s configuration, you can save, or capture, your configured
VM as a "base image" for future use or for other members of your team. Then, when booting a new VM, you
provision it from the base image rather than the Azure Marketplace image.
A quick summary: Use the System Preparation tool (Sysprep) and shut down the running VM, and then capture
(Figure 1 ) the VM as an image through the UI in the Azure portal. Azure saves the .vhd file that contains the
image in the storage account of your choosing. The new image then shows up as an Image resource in your
subscription’s list of resources.
(Figure 1 ) Capture an image through the Azure portal UI.
For more information, see Create a managed image of a generalized VM in Azure.
IMPORTANT
Don’t forget to use Sysprep to prepare the VM. If you miss that step, Azure can't provision a VM from the image.
NOTE
You still incur some cost for storage of the images, but that incremental cost can be insignificant compared to the overhead
costs to rebuild the VM from scratch for each team member who needs one. For instance, it costs a few dollars to create and
store a 127-GB image for a month that's reusable by your entire team. However, these costs are insignificant compared to
hours each employee invests to build out and validate a properly configured dev box for their individual use.
Additionally, your development tasks or technologies might need more scale, like varieties of development
configurations and multiple machine configurations. You can use Azure DevTest Labs to create recipes that
automate construction of your "golden image." You can also use DevTest Labs to manage policies for your team’s
running VMs. Using Azure DevTest Labs for developers is the best source for more information on DevTest Labs.
Next steps
Now that you know about the preconfigured Visual Studio images, the next step is to create a new VM:
Create a VM through the Azure portal
Windows Virtual Machines overview
Options with HPC Pack to create and manage a
cluster for Windows HPC workloads in Azure
Create and manage a cloud-based high-performance computing (HPC ) cluster by taking advantage of Microsoft
HPC Pack and Azure compute and infrastructure services. HPC Pack, available for free download, is built on
Microsoft Azure and Windows Server technologies and supports a wide range of HPC workloads.
Get the latest HPC Pack 2016 Update 1 downloads.
For more HPC options in Azure, see Technical resources for batch and high-performance computing.
This article focuses on Azure options to create HPC Pack clusters to run Windows workloads. There are also
options for creating HPC Pack clusters to run Linux HPC workloads.
HPC Pack cluster in Azure VMs and VM scale sets

Azure Resource Manager templates
(GitHub) HPC Pack 2016 Update 1 cluster templates
(GitHub) HPC Pack 2012 R2 cluster templates
(Quickstart) Create an HPC Pack 2012 R2 cluster
(Quickstart) Create an HPC Pack 2012 R2 cluster with custom compute node image
Azure VM images
Browse HPC Pack images in the Azure Marketplace if you want to build your own HPC Pack cluster in Azure.
Tutorials
Tutorial: Deploy an HPC Pack 2016 cluster in Azure
Manage an HPC Pack 2016 cluster in Azure with Azure Active Directory
HPC Pack 2012 R2 cluster deployment in the classic deployment model

Create an HPC cluster with the HPC Pack IaaS deployment script
Tutorial: Get started with an HPC Pack cluster in Azure to run Excel and SOA workloads
Manage compute nodes in an HPC Pack cluster in Azure
Grow and shrink Azure compute resources in an HPC Pack cluster
Set up a Windows RDMA cluster with HPC Pack to run MPI applications
Burst to Azure from HPC Pack 2012 R2

Burst to Azure Worker Instances with Microsoft HPC Pack
Burst to Azure Batch with HPC Pack
Tutorial: Set up a hybrid cluster with HPC Pack in Azure
Job submission
Submit jobs to an HPC Pack cluster in Azure
Deploy an HPC Pack 2016 cluster in Azure
Follow the steps in this article to deploy a Microsoft HPC Pack 2016 Update 1 cluster in Azure virtual machines.
HPC Pack is Microsoft's free HPC solution built on Microsoft Azure and Windows Server technologies and
supports a wide range of HPC workloads.
Use one of the Azure Resource Manager templates to deploy the HPC Pack 2016 cluster. You have several choices
of cluster topology with different numbers and types of cluster head nodes and compute nodes.
Prerequisites
PFX certificate
A Microsoft HPC Pack 2016 cluster requires a Personal Information Exchange (PFX) certificate to secure the
communication between the HPC nodes. The certificate must meet the following requirements:
It must have a private key capable of key exchange
Key usage includes Digital Signature and Key Encipherment
Enhanced key usage includes Client Authentication and Server Authentication
If you don’t already have a certificate that meets these requirements, you can request the certificate from a
certification authority. Alternatively, use the following commands to generate the self-signed certificate based on
the operating system on which you run the command. Then, export the certificate as a password-protected PFX
file with private key.
For Windows 10 or Windows Server 2016, run the built-in New-SelfSignedCertificate PowerShell
cmdlet as follows:
New-SelfSignedCertificate -Subject "CN=HPC Pack 2016 Communication" -KeySpec KeyExchange -TextExtension

@("2.5.29.37={text}1.3.6.1.5.5.7.3.1,1.3.6.1.5.5.7.3.2") -CertStoreLocation cert:\CurrentUser\My -
KeyExportPolicy Exportable -NotAfter (Get-Date).AddYears(5)
For operating systems earlier than Windows 10 or Windows Server 2016, download the self-signed
certificate generator from the Microsoft Script Center. Extract its contents and run the following commands
at a PowerShell prompt:
Import-Module -Name c:\ExtractedModule\New-SelfSignedCertificateEx.ps1
New-SelfSignedCertificateEx -Subject "CN=HPC Pack 2016 Communication" -KeySpec Exchange -KeyUsage

"DigitalSignature,KeyEncipherment" -EnhancedKeyUsage "Server Authentication","Client Authentication" -
StoreLocation CurrentUser -Exportable -NotAfter (Get-Date).AddYears(5)
After the certificate is created in the Current User store, use the Certificates snap-in to export the certificate as a
password-protected PFX file with private key. You can also export the certificate using the Export-Pfxcertificate
PowerShell cmdlet.
Upload certificate to an Azure key vault
Before deploying the HPC cluster, upload the PFX certificate to an Azure key vault as a secret, and record the
following information for use during the deployment: Vault name, Vault resource group, Certificate URL, and
Certificate thumbprint.
A sample PowerShell script to upload the certificate, create the key vault, and generate the required information
follows. For more information about uploading a certificate to an Azure key vault, see Get started with Azure Key
Vault.
#Give the following values

$VaultName = "mytestvault"
$SecretName = "hpcpfxcert"
$VaultRG = "myresourcegroup"
$location = "westus"
$PfxFile = "c:\Temp\mytest.pfx"
$Password = "yourpfxkeyprotectionpassword"
#Validate the pfx file
try {
$pfxCert = New-Object System.Security.Cryptography.X509Certificates.X509Certificate2 -ArgumentList
$PfxFile, $Password
}
catch [System.Management.Automation.MethodInvocationException]
{
throw $_.Exception.InnerException
}
$thumbprint = $pfxCert.Thumbprint
$pfxCert.Dispose()
# Create and encode the JSON object
$pfxContentBytes = Get-Content $PfxFile -Encoding Byte
$pfxContentEncoded = [System.Convert]::ToBase64String($pfxContentBytes)
$jsonObject = @"
{
"data": "$pfxContentEncoded",
"dataType": "pfx",
"password": "$Password"
}
"@
$jsonObjectBytes = [System.Text.Encoding]::UTF8.GetBytes($jsonObject)
$jsonEncoded = [System.Convert]::ToBase64String($jsonObjectBytes)
#Create an Azure key vault and upload the certificate as a secret
$secret = ConvertTo-SecureString -String $jsonEncoded -AsPlainText -Force
$rg = Get-AzureRmResourceGroup -Name $VaultRG -Location $location -ErrorAction SilentlyContinue
if($null -eq $rg)
{
$rg = New-AzureRmResourceGroup -Name $VaultRG -Location $location
}
$hpcKeyVault = New-AzureRmKeyVault -VaultName $VaultName -ResourceGroupName $VaultRG -Location $location -
EnabledForDeployment -EnabledForTemplateDeployment
$hpcSecret = Set-AzureKeyVaultSecret -VaultName $VaultName -Name $SecretName -SecretValue $secret
"The following Information will be used in the deployment template"
"Vault Name : $VaultName"
"Vault Resource Group : $VaultRG"
"Certificate URL : $($hpcSecret.Id)"
"Certificate Thumbprint : $thumbprint"
Supported topologies
Choose one of the Azure Resource Manager templates to deploy the HPC Pack 2016 cluster. Following are high-
level architectures of three example cluster topologies. High-availability topologies include multiple cluster head
nodes.
1. High-availability cluster with Active Directory domain
2. High-availability cluster without Active Directory domain
3. Cluster with a single head node

Deploy a cluster
To create the cluster, choose a template and click Deploy to Azure. In the Azure portal, specify parameters for the
template as described in the following steps. Each template creates all Azure resources required for the HPC
cluster infrastructure. Resources include an Azure virtual network, public IP address, load balancer (only for a
high-availability cluster), network interfaces, availability sets, storage accounts, and virtual machines.
Step 1: Select the subscription, location, and resource group
The Subscription and the Location must be same that you specified when you uploaded your PFX certificate
(see Prerequisites). We recommend that you create a different Resource group for the deployment.
Step 2: Specify the parameter settings
Enter or modify values for the template parameters. Click the icon next to each parameter for help information.
Also see the guidance for available VM sizes.
Specify the values you recorded in the Prerequisites for the following parameters: Vault name, Vault resource
group, Certificate URL, and Certificate thumbprint.
Step 3. Review terms and create
Review the terms and conditions associated with the template. If you agree, click Purchase to start the
deployment.
Depending on the cluster topology, deployment can take 30 minutes or longer to complete.
Connect to the cluster

1. After the HPC Pack cluster is deployed, go to the Azure portal. Click Resource groups, and find the
resource group in which the cluster was deployed. You can find the head node virtual machines.
2. Click one head node (in a high-availability cluster, click any of the head nodes). In Overview, you can find
the public IP address or full DNS name of the cluster.
3. Click Connect to log on to any of the head nodes using Remote Desktop with your specified administrator
user name. If the cluster you deployed is in an Active Directory Domain, the user name is of the form
<privateDomainName>\<adminUsername> (for example, hpc.local\hpcadmin).
Next steps
Submit jobs to your cluster. See Submit jobs to HPC an HPC Pack cluster in Azure and Manage an HPC Pack
2016 cluster in Azure using Azure Active Directory.
Manage an HPC Pack cluster in Azure using Azure
Active Directory
Microsoft HPC Pack 2016 supports integration with Azure Active Directory (Azure AD ) for administrators who
deploy an HPC Pack cluster in Azure.
Follow the steps in this article for the following high level tasks:
Manually integrate your HPC Pack cluster with your Azure AD tenant
Manage and schedule jobs in your HPC Pack cluster in Azure
Integrating an HPC Pack cluster solution with Azure AD follows standard steps to integrate other applications and
services. This article assumes you are familiar with basic user management in Azure AD. For more information
and background, see the Azure Active Directory documentation and the following section.
Benefits of integration
Azure Active Directory (Azure AD ) is a multi-tenant cloud-based directory and identity management service that
provides single sign-on (SSO ) access to cloud solutions.
Integration of an HPC Pack cluster with Azure AD can help you achieve the following goals:
Remove the traditional Active Directory domain controller from the HPC Pack cluster. This can help reduce the
costs of maintaining the cluster if this is not necessary for your business, and speed-up the deployment
process.
Leverage the following benefits that are brought by Azure AD:
Single sign-on
Using a local AD identity for the HPC Pack cluster in Azure
Prerequisites
HPC Pack 2016 cluster deployed in Azure virtual machines - For steps, see Deploy an HPC Pack 2016
cluster in Azure. You need the DNS name of the head node and the credentials of a cluster administrator to
complete the steps in this article.
NOTE
Azure Active Directory integration is not supported in versions of HPC Pack before HPC Pack 2016.
Client computer - You need a Windows or Windows Server client computer to run HPC Pack client
utilities. If you only want to use the HPC Pack web portal or REST API to submit jobs, you can use any
client computer of your choice.
HPC Pack client utilities - Install the HPC Pack client utilities on the client computer, using the free
installation package available from the Microsoft Download Center.
Step 1: Register the HPC cluster server with your Azure AD tenant
2. If your account gives you access to more than one Azure AD tenant, click your account in the top right corner.
Then set your portal session to the desired tenant. You must have permission to access resources in the
directory.
3. Click Azure Active Directory in the left Services navigation pane, click Users and groups, and make sure
there are user accounts already created or configured.
4. In Azure Active Directory, click App registrations > New application registration. Enter the following
information:
Name - HPCPackClusterServer
Application type - Select Web app / API
Sign-on URL - The base URL for the sample, which is by default https://hpcserver
Click Create.
5. After the app is added, select it in the App registrations list. Then click Settings > Properties. Enter the
following information:
Select Yes for Multi-tenanted.
Change App ID URI to https://<Directory_name>/<application_name> . Replace <Directory_name > with
the full name of your Azure AD tenant, for example, hpclocal.onmicrosoft.com , and replace
<application_name> with the name you chose previously.
6. Click Save. When saving completes, on the app page, click Manifest. Edit the manifest by locating the
appRoles setting and adding the following application role, and then click Save:
"appRoles": [
{
"allowedMemberTypes": [
"User",
"Application"
],
"displayName": "HpcAdminMirror",
"id": "61e10148-16a8-432a-b86d-ef620c3e48ef",
"isEnabled": true,
"description": "HpcAdminMirror",
"value": "HpcAdminMirror"
},
{
"allowedMemberTypes": [
"User",
"Application"
],
"description": "HpcUsers",
"displayName": "HpcUsers",
"id": "91e10148-16a8-432a-b86d-ef620c3e48ef",
"isEnabled": true,
"value": "HpcUsers"
}
],
7. In Azure Active Directory, click Enterprise applications > All applications. Select
HPCPackClusterServer from the list.
8. Click Properties, and change User assignment required to Yes. Click Save.
9. Click Users and groups > Add user. Select a user and select a role, and then click Assign. Assign one of the
available roles (HpcUsers or HpcAdminMirror) to the user. Repeat this step with additional users in the
directory. For background information about cluster users, see Managing Cluster Users.
Step 2: Register the HPC cluster client with your Azure AD tenant
2. If your account gives you access to more than one Azure AD tenant, click your account in the top right corner.
Then set your portal session to the desired tenant. You must have permission to access resources in the
directory.
3. In Azure Active Directory, click App registrations > New application registration. Enter the following
information:
Name - HPCPackClusterClient
Application type - Select Native
Redirect URI - http://hpcclient
Click Create
4. After the app is added, select it in the App registrations list. Copy the Application ID value and save it.
You need this later when configuring your application.
5. Click Settings > Required permissions > Add > Select an API. Search and select the
HpcPackClusterServer application (created in Step 1).
6. In the Enable Access page, select Access HpcClusterServer. Then click Done.
Step 3: Configure the HPC cluster

1. Connect to the HPC Pack 2016 head node in Azure.
2. Start HPC PowerShell.
3. Run the following command:
Set-HpcClusterRegistry -SupportAAD true -AADInstance https://login.microsoftonline.com/ -AADAppName

HpcPackClusterServer -AADTenant <your AAD tenant name> -AADClientAppId <client ID> -
AADClientAppRedirectUri http://hpcclient
where
AADTenant specifies the Azure AD tenant name, such as hpclocal.onmicrosoft.com
AADClientAppId specifies the Application ID for the app created in Step 2.
4. Do one of the following, depending on the head node configuration:
In a single head node HPC Pack cluster, restart the HpcScheduler service.
In an HPC Pack cluster with multiple head nodes, run the following PowerShell commands on the
head node to restart the HpcSchedulerStateful service:
Connect-ServiceFabricCluster
Move-ServiceFabricPrimaryReplica –ServiceName "fabric:/HpcApplication/SchedulerStatefulService"

Step 4: Manage and submit jobs from the client
To install the HPC Pack client utilities on your computer, download the HPC Pack 2016 setup files (full installation)
from the Microsoft Download Center. When you begin the installation, choose the setup option for the HPC Pack
client utilities.
To prepare the client computer, install the certificate used during HPC cluster setup on the client computer. Use
standard Windows certificate management procedures to install the public certificate to the Certificates –
Current user > Trusted Root Certification Authorities store.
You can now run the HPC Pack commands or use the HPC Pack Job manager GUI to submit and manage cluster
jobs by using the Azure AD account. For job submission options, see Submit HPC jobs to an HPC Pack cluster in
Azure.
NOTE
When you try to connect to the HPC Pack cluster in Azure for the first time, a popup windows appears. Enter your Azure AD
credentials to log in. The token is then cached. Later connections to the cluster in Azure use the cached token unless
authentication changes or the cache is cleared.
For example, after completing the previous steps, you can query for jobs from an on-premises client as follows:
Get-HpcJob –State All –Scheduler https://<Azure load balancer DNS name> -Owner <Azure AD account>
Useful cmdlets for job submission with Azure AD integration

Manage the local token cache
HPC Pack 2016 provides the following HPC PowerShell cmdlets to manage the local token cache. These cmdlets
are useful for submitting jobs non-interactively. See the following example:
Remove-HpcTokenCache
$SecurePassword = "<password>" | ConvertTo-SecureString -AsPlainText -Force
Set-HpcTokenCache -UserName <AADUsername> -Password $SecurePassword -scheduler https://<Azure load balancer

DNS name>
Set the credentials for submitting jobs using the Azure AD account
Sometimes, you may want to run the job under the HPC cluster user (for a domain-joined HPC cluster, run as one
domain user; for a non-domain-joined HPC cluster, run as one local user on the head node).
1. Use the following commands to set the credentials:
$localUser = "<username>"
$localUserPassword="<password>"
$secpasswd = ConvertTo-SecureString $localUserPassword -AsPlainText -Force
$mycreds = New-Object System.Management.Automation.PSCredential ($localUser, $secpasswd)
Set-HpcJobCredential -Credential $mycreds -Scheduler https://<Azure load balancer DNS name>
2. Then submit the job as follows. The job/task runs under $localUser on the compute nodes.
$emptycreds = New-Object System.Management.Automation.PSCredential ($localUser, (new-object
System.Security.SecureString))
...
$job = New-HpcJob –Scheduler https://<Azure load balancer DNS name>
Add-HpcTask -Job $job -CommandLine "ping localhost" -Scheduler https://<Azure load balancer DNS name>
Submit-HpcJob -Job $job -Scheduler https://<Azure load balancer DNS name> -Credential $emptycreds
If –Credential is not specified with Submit-HpcJob , the job or task runs under a local mapped user as the
Azure AD account. (The HPC cluster creates a local user with the same name as the Azure AD account to
run the task.)
3. Set extended data for the Azure AD account. This is useful when running an MPI job on Linux nodes using
the Azure AD account.
Set extended data for the Azure AD account itself
Set-HpcJobCredential -Scheduler https://<Azure load balancer DNS name> -ExtendedData <data> -

AadUser
Set extended data and run as HPC cluster user
Set-HpcJobCredential -Credential $mycreds -Scheduler https://<Azure load balancer DNS name> -

ExtendedData <data>
Create the head node of an HPC Pack cluster in an
Azure VM with a Marketplace image
Use a Microsoft HPC Pack 2012 R2 virtual machine image from the Azure Marketplace and the Azure portal to
create the head node of an HPC cluster. This HPC Pack VM image is based on Windows Server 2012 R2
Datacenter with HPC Pack 2012 R2 Update 3 pre-installed. Use this head node for a proof of concept deployment
of HPC Pack in Azure. You can then add compute nodes to the cluster to run HPC workloads.
TIP
To deploy a complete HPC Pack 2012 R2 cluster in Azure that includes the head node and compute nodes, we recommend
that you use an automated method. Options include the HPC Pack IaaS deployment script and Resource Manager templates
such as the HPC Pack cluster for Windows workloads. Resource Manager templates are also available for Microsoft HPC Pack
2016 clusters.
Planning considerations
As shown in the following figure, you deploy the HPC Pack head node in an Active Directory domain in an Azure
virtual network.
Active Directory domain: The HPC Pack 2012 R2 head node must be joined to an Active Directory
domain in Azure before you start the HPC services on the VM. As shown in this article, for a proof of
concept deployment, you can promote the VM you create for the head node as a domain controller before
starting the HPC services. Another option is to deploy a separate domain controller and forest in Azure to
which you join the head node VM.
Deployment model: For most new deployments, Microsoft recommends that you use the Resource
Manager deployment model. This article assumes that you use this deployment model.
Azure virtual network: When you use the Resource Manager deployment model to deploy the head node,
you specify or create an Azure virtual network. You use the virtual network if you need to join the head node
to an existing Active Directory domain. You also need it later to add compute node VMs to the cluster.
Steps to create the head node

Following are high-level steps to use the Azure portal to create an Azure VM for the HPC Pack head node by using
the Resource Manager deployment model.
1. If you want to create a new Active Directory forest in Azure with separate domain controller VMs, one option is
to use a Resource Manager template. For a simple proof of concept deployment, it's fine to omit this step and
configure the head node VM itself as a domain controller. This option is described later in this article.
2. On the HPC Pack 2012 R2 on Windows Server 2012 R2 page in the Azure Marketplace, click Create Virtual
Machine.
3. In the portal, on the HPC Pack 2012 R2 on Windows Server 2012 R2 page, select the Resource
Manager deployment model and then click Create.
4. Use the portal to configure the settings and create the VM. If you're new to Azure, follow the tutorial Create
a Windows virtual machine in the Azure portal. For a proof of concept deployment, you can usually accept
the default or recommended settings.
NOTE
If you want to join the head node to an existing Active Directory domain in Azure, make sure you specify the virtual
network for that domain when creating the VM.
5. After you create the VM and the VM is running, connect to the VM by Remote Desktop.
6. Join the VM to an Active Directory domain forest by choosing one of the following options:
If you created the VM in an Azure virtual network with an existing domain forest, join the VM to the
forest by using standard Server Manager or Windows PowerShell tools. Then restart.
If you created the VM in a new virtual network (without an existing domain forest), then promote the VM
as a domain controller. Use standard steps to install and configure the Active Directory Domain Services
role on the head node. For detailed steps, see Install a New Windows Server 2012 Active Directory
Forest.
7. After the VM is running and is joined to an Active Directory forest, start the HPC Pack services as follows:
a. Connect to the head node VM using a domain account that is a member of the local Administrators
group. For example, use the administrator account you set up when you created the head node VM.
b. For a default head node configuration, start Windows PowerShell as an administrator and type the
following:
& $env:CCP_HOME\bin\HPCHNPrepare.ps1 –DBServerInstance ".\ComputeCluster"
It can take several minutes for the HPC Pack services to start.
For additional head node configuration options, type get-help HPCHNPrepare.ps1 .
Next steps
You can now work with the head node of your HPC Pack cluster. For example, start HPC Cluster Manager, and
complete the Deployment To-do List.
If you want to increase the cluster compute capacity on-demand, add Azure burst nodes in a cloud service.
Try running a test workload on the cluster. For an example, see the HPC Pack getting started guide.
Submit HPC jobs from an on-premises computer to
an HPC Pack cluster deployed in Azure
NOTE
Configure an on-premises client computer to submit jobs to a Microsoft HPC Pack cluster in Azure. This article
shows you how to set up a local computer with client tools to submit job over HTTPS to the cluster in Azure. In
this way, several cluster users can submit jobs to a cloud-based HPC Pack cluster, but without connecting directly
to the head node VM or accessing an Azure subscription.
Prerequisites
HPC Pack head node deployed in an Azure VM - We recommend that you use automated tools such as an
Azure quickstart template to deploy the head node and cluster. You need the DNS name of the head node and
the credentials of a cluster administrator to complete the steps in this article.
Client computer - You need a Windows or Windows Server client computer that can run HPC Pack client
utilities (see system requirements). If you only want to use the HPC Pack web portal or REST API to submit
jobs, you can use any client computer of your choice.
HPC Pack installation media - To install the HPC Pack client utilities, the free installation package for the
latest version of HPC Pack is available from the Microsoft Download Center. Make sure that you download the
same version of HPC Pack that is installed on the head node VM.
Step 1: Install and configure the web components on the head node
To enable a REST interface to submit jobs to the cluster over HTTPS, ensure that the HPC Pack web components
are configured on the HPC Pack head node. If they aren't already installed, first install the web components by
running the HpcWebComponents.msi installation file. Then, configure the components by running the HPC
PowerShell script Set-HPCWebComponents.ps1.
For detailed procedures, see Install the Microsoft HPC Pack Web Components.
TIP
Certain Azure quickstart templates for HPC Pack clusters install and configure the web components automatically.
To install the web components

1. Connect to the head node VM by using the credentials of a cluster administrator.
2. From the HPC Pack Setup folder, run HpcWebComponents.msi on the head node.
3. Follow the steps in the wizard to install the web components
To configure the web components
1. On the head node, start HPC PowerShell as an administrator.
2. To change directory to the location of the configuration script, type the following command:
cd $env:CCP_HOME\bin
3. To configure the REST interface and start the HPC Web Service, type the following command:
.\Set-HPCWebComponents.ps1 –Service REST –enable
4. When prompted to select a certificate, choose the certificate that corresponds to the public DNS name of
the head node. For example, if you deploy the head node VM using the classic deployment model, the
certificate name looks like CN=<HeadNodeDnsName>.cloudapp.net. If you use the Resource Manager
deployment model, the certificate name looks like CN=<HeadNodeDnsName>.
<region>.cloudapp.azure.com.
NOTE
You select this certificate later when you submit jobs to the head node from an on-premises computer. Don't select
or configure a certificate that corresponds to the computer name of the head node in the Active Directory domain
(for example, CN=MyHPCHeadNode.HpcAzure.local).
5. To configure the web portal for job submission, type the following command:
.\Set-HPCWebComponents.ps1 –Service Portal -enable
6. After the script completes, stop and restart the HPC Job Scheduler Service by typing the following
commands:
net stop hpcscheduler

net start hpcscheduler
Step 2: Install the HPC Pack client utilities on an on-premises computer

If you want to install the HPC Pack client utilities on your computer, download the HPC Pack setup files (full
installation) from the Microsoft Download Center. When you begin the installation, choose the setup option for
the HPC Pack client utilities.
To use the HPC Pack client tools to submit jobs to the head node VM, you also need to export a certificate from
the head node and install it on the client computer. The certificate must be in .CER format.
To export the certificate from the head node
1. On the head node, add the Certificates snap-in to a Microsoft Management Console for the Local Computer
account. For steps to add the snap-in, see Add the Certificates Snap-in to an MMC.
2. In the console tree, expand Certificates – Local Computer > Personal, and then click Certificates.
3. Locate the certificate that you configured for the HPC Pack web components in Step 1: Install and configure the
web components on the head node (for example, CN=<HeadNodeDnsName>.cloudapp.net).
4. Right-click the certificate, and click All Tasks > Export.
5. In the Certificate Export Wizard, click Next, and ensure that No, do not export the private key is selected.
6. Follow the remaining steps of the wizard to export the certificate in DER encoded binary X.509 (.CER ) format.
To import the certificate on the client computer
1. Copy the certificate that you exported from the head node to a folder on the client computer.
2. On the client computer, run certmgr.msc.
3. In Certificate Manager, expand Certificates – Current user > Trusted Root Certification Authorities, right-
click Certificates, and then click All Tasks > Import.
4. In the Certificate Import Wizard, click Next and follow the steps to import the certificate that you exported
from the head node to the Trusted Root Certification Authorities store.
TIP
You might see a security warning, because the certification authority on the head node isn't recognized by the client
computer. For testing purposes, you can ignore this warning and complete the certificate import.
Step 3: Run test jobs on the cluster

To verify your configuration, try running jobs on the cluster in Azure from the on-premises computer. For example,
you can use HPC Pack GUI tools or command-line commands to submit jobs to the cluster. You can also use a
web-based portal to submit jobs.
To run job submission commands on the client computer
1. On a client computer where the HPC Pack client utilities are installed, start a Command Prompt.
2. Type a sample command. For example, to list all jobs on the cluster, type a command similar to one of the
following, depending on the full DNS name of the head node:
job list /scheduler:https://<HeadNodeDnsName>.cloudapp.net /all
or
job list /scheduler:https://<HeadNodeDnsName>.<region>.cloudapp.azure.com /all
TIP
Use the full DNS name of the head node, not the IP address, in the scheduler URL. If you specify the IP address, an
error appears similar to "The server certificate needs to either have a valid chain of trust or to be placed in the
trusted root store."
3. When prompted, type the user name (in the form <DomainName>\<UserName>) and password of the
HPC cluster administrator or another cluster user that you configured. You can choose to store the
credentials locally for more job operations.
A list of jobs appears.
To use HPC Job Manager on the client computer
1. If you didn't previously store domain credentials for a cluster user when submitting a job, you can add the
credentials in Credential Manager.
a. In Control Panel on the client computer, start Credential Manager.
b. Click Windows Credentials > Add a generic credential.
c. Specify the Internet address (for example, https://<HeadNodeDnsName>.cloudapp.net/HpcScheduler or
https://<HeadNodeDnsName>.<region>.cloudapp.azure.com/HpcScheduler), and the user name
(<DomainName>\<UserName>) and password of the cluster administrator or another cluster user that
you configured.
2. On the client computer, start HPC Job Manager.
3. In the Select Head Node dialog box, type the URL to the head node in Azure (for example,
https://<HeadNodeDnsName>.cloudapp.net or https://<HeadNodeDnsName>.
<region>.cloudapp.azure.com).
HPC Job Manager opens and shows a list of jobs on the head node.
To use the web portal running on the head node
1. Start a web browser on the client computer, and enter one of the following addresses, depending on the full
DNS name of the head node:
https://<HeadNodeDnsName>.cloudapp.net/HpcPortal
or
https://<HeadNodeDnsName>.<region>.cloudapp.azure.com/HpcPortal
2. In the security dialog box that appears, type the domain credentials of the HPC cluster administrator. (You
can also add other cluster users in different roles. See Managing Cluster Users.)
The web portal opens to the job list view.
3. To submit a sample job that returns the string “Hello World” from the cluster, click New job in the left-hand
navigation.
4. On the New Job page, under From submission pages, click HelloWorld. The job submission page appears.
5. Click Submit. If prompted, provide the domain credentials of the HPC cluster administrator. The job is
submitted, and the job ID appears on the My Jobs page.
6. To view the results of the job that you submitted, click the job ID, and then click View Tasks to view the
command output (under Output).
Next steps
You can also submit jobs to the Azure cluster with the HPC Pack REST API.
If you want to submit cluster jobs from a Linux client, see the Python sample in the HPC Pack 2012 R2 SDK
and Sample Code.
Get started running Excel and SOA workloads on an
HPC Pack cluster in Azure
This article shows you how to deploy a Microsoft HPC Pack 2012 R2 cluster on Azure virtual machines by using
an Azure quickstart template, or optionally an Azure PowerShell deployment script. The cluster uses Azure
Marketplace VM images designed to run Microsoft Excel or service-oriented architecture (SOA) workloads with
HPC Pack. You can use the cluster to run Excel HPC and SOA services from an on-premises client computer. The
Excel HPC services include Excel workbook offloading and Excel user-defined functions, or UDFs.
IMPORTANT
This article is based on features, templates, and scripts for HPC Pack 2012 R2. This scenario is not currently supported in
HPC Pack 2016.
NOTE
At a high level, the following diagram shows the HPC Pack cluster you create.
Prerequisites
Client computer - You need a Windows-based client computer to submit sample Excel and SOA jobs to the
cluster. You also need a Windows computer to run the Azure PowerShell cluster deployment script (if you
choose that deployment method).
Azure subscription - If you don't have an Azure subscription, you can create a free account in just a couple of
minutes.
Cores quota - You might need to increase the quota of cores, especially if you deploy several cluster nodes
with multicore VM sizes. If you are using an Azure quickstart template, the cores quota in Resource Manager is
per Azure region. In that case, you might need to increase the quota in a specific region. See Azure subscription
limits, quotas, and constraints. To increase a quota, open an online customer support request at no charge.
Microsoft Office license - If you deploy compute nodes using a Marketplace HPC Pack 2012 R2 VM image
with Microsoft Excel, a 30-day evaluation version of Microsoft Excel Professional Plus 2013 is installed. After
the evaluation period, you need to provide a valid Microsoft Office license to activate Excel to continue to run
workloads. See Excel activation later in this article.
Step 1. Set up an HPC Pack cluster in Azure
We show two options to set up the HPC Pack 2012 R2 cluster: first, using an Azure quickstart template and the
Azure portal; and second, using an Azure PowerShell deployment script.
Option 1. Use a quickstart template
Use an Azure quickstart template to quickly deploy an HPC Pack cluster in the Azure portal. When you open the
template in the portal, you get a simple UI where you enter the settings for your cluster. Here are the steps.
TIP
If you want, use an Azure Marketplace template that creates a similar cluster specifically for Excel workloads. The steps differ
slightly from the following.
1. Visit the Create HPC Cluster template page on GitHub. If you want, review information about the template and
the source code.
2. Click Deploy to Azure to start a deployment with the template in the Azure portal.
3. In the portal, follow these steps to enter the parameters for the HPC cluster template.
a. On the Parameters page, enter or modify values for the template parameters. (Click the icon next to each
setting for help information.) Sample values are shown in the following screen. This example creates a
cluster named hpc01 in the hpc.local domain consisting of a head node and 2 compute nodes. The compute
nodes are created from an HPC Pack VM image that includes Microsoft Excel.
NOTE
The head node VM is created automatically from the latest Marketplace image of HPC Pack 2012 R2 on Windows
Server 2012 R2. Currently the image is based on HPC Pack 2012 R2 Update 3.
Compute node VMs are created from the latest image of the selected compute node family. Select the
ComputeNodeWithExcel option for the latest HPC Pack compute node image that includes an evaluation version of
Microsoft Excel Professional Plus 2013. To deploy a cluster for general SOA sessions or for Excel UDF offloading,
choose the ComputeNode option (without Excel installed).
b. Choose the subscription.

c. Create a resource group for the cluster, such as hpc01RG.
d. Choose a location for the resource group, such as Central US.
e. On the Legal terms page, review the terms. If you agree, click Purchase. Then, when you are finished
setting the values for the template, click Create.
4. When the deployment completes (it typically takes around 30 minutes), export the cluster certificate file
from the cluster head node. In a later step, you import this public certificate on the client computer to
provide the server-side authentication for secure HTTP binding.
a. In the Azure portal, go to the dashboard, select the head node, and click Connect at the top of the page to
connect using Remote Desktop.
b. Use standard procedures in Certificate Manager to export the head node certificate (located under
Cert:\LocalMachine\My) without the private key. In this example, export CN =
hpc01.eastus.cloudapp.azure.com.
Option 2. Use the HPC Pack IaaS Deployment script

The HPC Pack IaaS deployment script provides another versatile way to deploy an HPC Pack cluster. It creates a
cluster in the classic deployment model, whereas the template uses the Azure Resource Manager deployment
model. Also, the script is compatible with a subscription in either the Azure Global or Azure China service.
Additional prerequisites
Azure PowerShell - Install and configure Azure PowerShell (version 0.8.10 or later) on your client computer.
HPC Pack IaaS deployment script - Download and unpack the latest version of the script from the Microsoft
Download Center. Check the version of the script by running New-HPCIaaSCluster.ps1 –Version . This article is
based on version 4.5.0 or later of the script.
Create the configuration file
The HPC Pack IaaS deployment script uses an XML configuration file as input that describes the infrastructure of
the HPC cluster. To deploy a cluster consisting of a head node and 18 compute nodes created from the compute
node image that includes Microsoft Excel, substitute values for your environment into the following sample
configuration file. For more information about the configuration file, see the Manual.rtf file in the script folder and
Create an HPC cluster with the HPC Pack IaaS deployment script.
<?xml version="1.0" encoding="utf-8"?>
<IaaSClusterConfig>
<Subscription>
<SubscriptionName>MySubscription</SubscriptionName>
<StorageAccount>hpc01</StorageAccount>
</Subscription>
<Location>West US</Location>
<VNet>
<VNetName>hpc-vnet01</VNetName>
<SubnetName>Subnet-1</SubnetName>
</VNet>
<Domain>
<DCOption>NewDC</DCOption>
<DomainFQDN>hpc.local</DomainFQDN>
<DomainController>
<VMName>HPCExcelDC01</VMName>
<ServiceName>HPCExcelDC01</ServiceName>
<VMSize>Medium</VMSize>
</DomainController>
</Domain>
<Database>
<DBOption>LocalDB</DBOption>
</Database>
<HeadNode>
<VMName>HPCExcelHN01</VMName>
<ServiceName>HPCExcelHN01</ServiceName>
<VMSize>Large</VMSize>
<EnableRESTAPI/>
<EnableWebPortal/>
<PostConfigScript>C:\tests\PostConfig.ps1</PostConfigScript>
</HeadNode>
<ComputeNodes>
<VMNamePattern>HPCExcelCN%00%</VMNamePattern>
<ServiceName>HPCExcelCN01</ServiceName>
<VMSize>Medium</VMSize>
<NodeCount>18</NodeCount>
<ImageName>HPCPack2012R2_ComputeNodeWithExcel</ImageName>
</ComputeNodes>
</IaaSClusterConfig>
Notes about the configuration file

The VMName of the head node MUST be the same as the ServiceName, or SOA jobs fail to run.
Make sure you specify EnableWebPortal so that the head node certificate is generated and exported.
The file specifies a post-configuration PowerShell script PostConfig.ps1 that runs on the head node. THe
following sample script configures the Azure storage connection string, removes the compute node role from
the head node, and brings all nodes online when they are deployed.
# add the HPC Pack powershell cmdlets
Add-PSSnapin Microsoft.HPC
# set the Azure storage connection string for the cluster

Set-HpcClusterProperty -AzureStorageConnectionString 'DefaultEndpointsProtocol=https;AccountName=
<yourstorageaccountname>;AccountKey=<yourstorageaccountkey>'
# remove the compute node role for head node to make sure the Excel workbook won’t run on head node
Get-HpcNode -GroupName HeadNodes | Set-HpcNodeState -State offline | Set-HpcNode -Role BrokerNode
# total number of nodes in the deployment including the head node and compute nodes, which should match
the number specified in the XML configuration file
$TotalNumOfNodes = 19
$ErrorActionPreference = 'SilentlyContinue'
# bring nodes online when they are deployed until all nodes are online
while ($true)
{
Get-HpcNode -State Offline | Set-HpcNodeState -State Online -Confirm:$false
$OnlineNodes = @(Get-HpcNode -State Online)
if ($OnlineNodes.Count -eq $TotalNumOfNodes)
{
break
}
sleep 60
}
Run the script

1. Open the PowerShell console on the client computer as an administrator.
2. Change directory to the script folder (E:\IaaSClusterScript in this example).
cd E:\IaaSClusterScript
3. To deploy the HPC Pack cluster, run the following command. This example assumes that the configuration
file is located in E:\HPCDemoConfig.xml.
.\New-HpcIaaSCluster.ps1 –ConfigFile E:\HPCDemoConfig.xml –AdminUserName MyAdminName
The HPC Pack deployment script runs for some time. One thing the script does is to export and download the
cluster certificate and save it in the current user’s Documents folder on the client computer. The script generates a
message similar to the following. In a following step, you import the certificate in the appropriate certificate store.
You have enabled REST API or web portal on HPC Pack head node. Please import the following certificate in the
Trusted Root Certification Authorities certificate store on the computer where you are submitting job or
accessing the HPC web portal:
C:\Users\hpcuser\Documents\HPCWebComponent_HPCExcelHN004_20150707162011.cer
Step 2. Offload Excel workbooks and run UDFs from an on-premises

client
Excel activation
When using the ComputeNodeWithExcel VM image for production workloads, you need to provide a valid
Microsoft Office license key to activate Excel on the compute nodes. Otherwise, the evaluation version of Excel
expires after 30 days, and running Excel workbooks will fail with the COMException (0x800AC472).
You can rearm Excel for another 30 days of evaluation time: Log on to the head node and clusrun
%ProgramFiles(x86)%\Microsoft Office\Office15\OSPPREARM.exe on all Excel compute nodes via HPC Cluster
Manager. You can rearm a maximum of two times. After that, you must provide a valid Office license key.
The Office Professional Plus 2013 installed on the VM image is a volume edition with a Generic Volume License
Key (GVLK). You can activate it via Key Management Service (KMS )/Active Directory-Based Activation (AD -BA) or
Multiple Activation Key (MAK).
* To use KMS/AD-BA, use an existing KMS server or set up a new one by using the Microsoft Office 2013 Volume
License Pack. (If you want to, set up the server on the head node.) Then, activate the KMS host key via the
Internet or telephone. Then clusrun `ospp.vbs` to set the KMS server and port and activate Office on all the
Excel compute nodes.
* To use MAK, first clusrun `ospp.vbs` to input the key and then activate all the Excel compute nodes via the
Internet or telephone.
NOTE
Retail product keys for Office Professional Plus 2013 cannot be used with this VM image. If you have valid keys and
installation media for Office or Excel editions other than this Office Professional Plus 2013 volume edition, you can use them
instead. First uninstall this volume edition and install the edition that you have. The reinstalled Excel compute node can be
captured as a customized VM image to use in a deployment at scale.
Offload Excel workbooks

Follow these steps to offload an Excel workbook so that it runs on the HPC Pack cluster in Azure. To do this, you
must have Excel 2010 or 2013 already installed on the client computer.
1. Use one of the options in Step 1 to deploy an HPC Pack cluster with the Excel compute node image. Obtain the
cluster certificate file (.cer) and cluster username and password.
2. On the client computer, import the cluster certificate under Cert:\CurrentUser\Root.
3. Make sure Excel is installed. Create an Excel.exe.config file with the following contents in the same folder as
Excel.exe on the client computer. This step ensures that the HPC Pack 2012 R2 Excel COM add-in loads
successfully.
<?xml version="1.0"?>
<configuration>
<startup useLegacyV2RuntimeActivationPolicy="true">
<supportedRuntime version="v4.0" sku=".NETFramework,Version=v4.0"/>
</startup>
</configuration>
4. Set up the client to submit jobs to the HPC Pack cluster. One option is to download the full HPC Pack 2012 R2
Update 3 installation and install the HPC Pack client. Alternatively, download and install the HPC Pack 2012 R2
Update 3 client utilities and the appropriate Visual C++ 2010 redistributable for your computer (x64, x86).
5. In this example, we use a sample Excel workbook named ConvertiblePricing_Complete.xlsb. You can download
it here.
6. Copy the Excel workbook to a working folder such as D:\Excel\Run.
7. Open the Excel workbook. On the Develop ribbon, click COM Add-Ins and confirm that the HPC Pack
Excel COM add-in is loaded successfully.
8. Edit the VBA macro HPCControlMacros in Excel by changing the commented lines as shown in the
following script. Substitute appropriate values for your environment.
'Private Const HPC_ClusterScheduler = "HEADNODE_NAME"

Private Const HPC_ClusterScheduler = "hpc01.eastus.cloudapp.azure.com"
'Private Const HPC_NetworkShare = "\\PATH\TO\SHARE\DIRECTORY"

Private Const HPC_DependFiles =
"D:\Excel\Upload\ConvertiblePricing_Complete.xlsb=ConvertiblePricing_Complete.xlsb"
'HPCExcelClient.Initialize ActiveWorkbook
HPCExcelClient.Initialize ActiveWorkbook, HPC_DependFiles
'HPCWorkbookPath = HPC_NetworkShare & Application.PathSeparator & ActiveWorkbook.name

HPCWorkbookPath = "ConvertiblePricing_Complete.xlsb"
'HPCExcelClient.OpenSession headNode:=HPC_ClusterScheduler, remoteWorkbookPath:=HPCWorkbookPath

HPCExcelClient.OpenSession headNode:=HPC_ClusterScheduler, remoteWorkbookPath:=HPCWorkbookPath,
UserName:="hpc\azureuser", Password:="<YourPassword>"
9. Copy the Excel workbook to an upload directory such as D:\Excel\Upload. This directory is specified in the
HPC_DependsFiles constant in the VBA macro.
10. To run the workbook on the cluster in Azure, click the Cluster button on the worksheet.
Run Excel UDFs
To run Excel UDFs, follow the preceding steps 1 – 3 to set up the client computer. For Excel UDFs, you don't need
to have the Excel application installed on compute nodes. So, when creating your cluster compute nodes, you could
choose a normal compute node image instead of the compute node image with Excel.
NOTE
There is a 34 character limit in the Excel 2010 and 2013 cluster connector dialog box. You use this dialog box to specify the
cluster that runs the UDFs. If the full cluster name is longer (for example, hpcexcelhn01.southeastasia.cloudapp.azure.com), it
does not fit in the dialog box. The workaround is to set a machine-wide variable such as CCP_IAASHN with the value of the
long cluster name. Then, enter %CCP_IAASHN% in the dialog box as the cluster head node name.
After the cluster is successfully deployed, continue with the following steps to run a sample built-in Excel UDF. For
customized Excel UDFs, see these resources to build the XLLs and deploy them on the IaaS cluster.
1. Open a new Excel workbook. On the Develop ribbon, click Add-Ins. Then, in the dialog box, click Browse,
navigate to the %CCP_HOME%Bin\XLL32 folder, and select the sample ClusterUDF32.xll. If the
ClusterUDF32 doesn't exist on the client machine, copy it from the %CCP_HOME%Bin\XLL32 folder on the
head node.
2. Click File > Options > Advanced. Under Formulas, check Allow user-defined XLL functions to run a
compute cluster. Then click Options and enter the full cluster name in Cluster head node name. (As
noted previously this input box is limited to 34 characters, so a long cluster name may not fit. You may use a
machine-wide variable here for a long cluster name.)
3. To run the UDF calculation on the cluster, click the cell with value =XllGetComputerNameC () and press
Enter. The function simply retrieves the name of the compute node on which the UDF runs. For the first run,
a credentials dialog box prompts for the username and password to connect to the IaaS cluster.
When there are many cells to calculate, press Alt-Shift-Ctrl + F9 to run the calculation on all cells.
Step 3. Run a SOA workload from an on-premises client

To run general SOA applications on the HPC Pack IaaS cluster, first use one of the methods in Step 1 to deploy the
cluster. Specify a generic compute node image in this case, because you do not need Excel on the compute nodes.
Then follow these steps.
1. After retrieving the cluster certificate, import it on the client computer under Cert:\CurrentUser\Root.
2. Install the HPC Pack 2012 R2 Update 3 SDK and HPC Pack 2012 R2 Update 3 client utilities. These tools
enable you to develop and run SOA client applications.
3. Download the HelloWorldR2 sample code. Open the HelloWorldR2.sln in Visual Studio 2010 or 2012. (This
sample is not currently compatible with more recent versions of Visual Studio.)
4. Build the EchoService project first. Then, deploy the service to the IaaS cluster in the same way you deploy to
an on-premises cluster. For detailed steps, see the Readme.doc in HelloWordR2. Modify and build the
HellWorldR2 and other projects as described in the following section to generate the SOA client applications
that run on an Azure IaaS cluster.
Use Http binding with Azure storage queue
To use Http binding with an Azure storage queue, make a few changes to the sample code.
Update the cluster name.
// Before
const string headnode = "[headnode]";
// After e.g.
const string headnode = "hpc01.eastus.cloudapp.azure.com";
or
const string headnode = "hpc01.cloudapp.net";
Optionally, use the default TransportScheme in SessionStartInfo or explicitly set it to Http.
info.TransportScheme = TransportScheme.Http;
Use default binding for the BrokerClient.

// Before
using (BrokerClient<IService1> client = new BrokerClient<IService1>(session, binding))
// After
using (BrokerClient<IService1> client = new BrokerClient<IService1>(session))
Or set explicitly using the basicHttpBinding.
BasicHttpBinding binding = new BasicHttpBinding(BasicHttpSecurityMode.TransportWithMessageCredential);

binding.Security.Message.ClientCredentialType = BasicHttpMessageCredentialType.UserName;
binding.Security.Transport.ClientCredentialType = HttpClientCredentialType.None;
Optionally, set the UseAzureQueue flag to true in SessionStartInfo. If not set, it will be set to true by default
when the cluster name has Azure domain suffixes and the TransportScheme is Http.
info.UseAzureQueue = true;
Use Http binding without Azure storage queue

To use Http binding without an Azure storage queue, explicitly set the UseAzureQueue flag to false in the
SessionStartInfo.
info.UseAzureQueue = false;
Use NetTcp binding

To use NetTcp binding, the configuration is similar to connecting to an on-premises cluster. You need to open a few
endpoints on the head node VM. If you used the HPC Pack IaaS deployment script to create the cluster, for
example, set the endpoints in the Azure portal as follows.
1. Stop the VM.
2. Add the TCP ports 9090, 9087, 9091, 9094 for the Session, Broker, Broker worker, and Data services,
respectively
3. Start the VM.

The SOA client application requires no changes except altering the head name to the IaaS cluster full name.
Next steps
See these resources for more information about running Excel workloads with HPC Pack.
See Managing SOA Services in Microsoft HPC Pack for more about deploying and managing SOA services
with HPC Pack.
Frequently asked questions about Azure IaaS VM
disks and managed and unmanaged premium disks
This article answers some frequently asked questions about Azure Managed Disks and Azure Premium SSD disks.
Managed Disks
What is Azure Managed Disks?
Managed Disks is a feature that simplifies disk management for Azure IaaS VMs by handling storage account
management for you. For more information, see the Managed Disks overview.
If I create a standard managed disk from an existing VHD that's 80 GB, how much will that cost me?
A standard managed disk created from an 80-GB VHD is treated as the next available standard disk size, which is
an S10 disk. You're charged according to the S10 disk pricing. For more information, see the pricing page.
Are there any transaction costs for standard managed disks?
Yes. You're charged for each transaction. For more information, see the pricing page.
For a standard managed disk, will I be charged for the actual size of the data on the disk or for the
provisioned capacity of the disk?
You're charged based on the provisioned capacity of the disk. For more information, see the pricing page.
How is pricing of premium managed disks different from unmanaged disks?
The pricing of premium managed disks is the same as unmanaged premium disks.
Can I change the storage account type (Standard or Premium ) of my managed disks?
Yes. You can change the storage account type of your managed disks by using the Azure portal, PowerShell, or the
Azure CLI.
Can I use a VHD file in an Azure storage account to create a managed disk with a different subscription?
Yes.
Can I use a VHD file in an Azure storage account to create a managed disk in a different region?
No.
Are there any scale limitations for customers that use managed disks?
Managed Disks eliminates the limits associated with storage accounts. However, the maximum limit is 50,000
managed disks per region and per disk type for a subscription.
Can I take an incremental snapshot of a managed disk?
No. The current snapshot capability makes a full copy of a managed disk.
Can VMs in an availability set consist of a combination of managed and unmanaged disks?
No. The VMs in an availability set must use either all managed disks or all unmanaged disks. When you create an
availability set, you can choose which type of disks you want to use.
Is Managed Disks the default option in the Azure portal?
Yes.
Can I create an empty managed disk?
Yes. You can create an empty disk. A managed disk can be created independently of a VM, for example, without
attaching it to a VM.
What is the supported fault domain count for an availability set that uses Managed Disks?
Depending on the region where the availability set that uses Managed Disks is located, the supported fault domain
count is 2 or 3.
How is the standard storage account for diagnostics set up?
You set up a private storage account for VM diagnostics.
What kind of Role-Based Access Control support is available for Managed Disks?
Managed Disks supports three key default roles:
Owner: Can manage everything, including access
Contributor: Can manage everything except access
Reader: Can view everything, but can't make changes
Is there a way that I can copy or export a managed disk to a private storage account?
You can generate a read-only shared access signature (SAS ) URI for the managed disk and use it to copy the
contents to a private storage account or on-premises storage. You can use the SAS URI using the Azure portal,
Azure PowerShell, the Azure CLI, or AzCopy
Can I create a copy of my managed disk?
Customers can take a snapshot of their managed disks and then use the snapshot to create another managed disk.
Are unmanaged disks still supported?
Yes, both unmanaged and managed disks are supported. We recommend that you use managed disks for new
workloads and migrate your current workloads to managed disks.
If I create a 128-GB disk and then increase the size to 130 GiB, will I be charged for the next disk size
(256 GiB )?
Yes.
Can I create locally redundant storage, geo-redundant storage, and zone-redundant storage managed
disks?
Azure Managed Disks currently supports only locally redundant storage managed disks.
Can I shrink or downsize my managed disks?
No. This feature is not supported currently.
Can I break a lease on my disk?
No. This is not supported currently as a lease is present to prevent accidental deletion when the disk is being used.
Can I change the computer name property when a specialized (not created by using the System
Preparation tool or generalized) operating system disk is used to provision a VM?
No. You can't update the computer name property. The new VM inherits it from the parent VM, which was used to
create the operating system disk.
Where can I find sample Azure Resource Manager templates to create VMs with managed disks?
List of templates using Managed Disks
https://github.com/chagarw/MDPP
Can I co-locate unmanaged and managed disks on the same VM?
No.
When creating a disk from a blob, is there any continually existing relationship with that source blob?
No, when the new disk is created it is a full standalone copy of that blob at that time and there is no connection
between the two. If you like, once you've created the disk, the source blob may be deleted without affecting the
newly created disk in any way.
Can I rename a managed or unmanaged disk after it has been created?
For managed disks you cannot rename them. However, you may rename an unmanaged disk as long as it is not
currently attached to a VHD or VM.
Standard SSD disks

What are Azure Standard SSD disks? Standard SSD disks are standard disks backed by solid-state media,
optimized as cost effective storage for workloads that need consistent performance at lower IOPS levels.
What are the regions currently supported for Standard SSD disks? All Azure regions now support Standard
SSD disks.
Is Azure Backup available when using Standard SSDs? Yes, Azure Backup is now available.
How do I create Standard SSD disks? You can create Standard SSD disks using Azure Resource Manager
templates, SDK, PowerShell or CLI. Below are the parameters needed in the Resource Manager template to create
Standard SSD Disks:
apiVersion for Microsoft.Compute must be set as 2018-04-01 (or later)
Specify managedDisk.storageAccountType as StandardSSD_LRS
The following example shows the properties.storageProfile.osDisk section for a VM that uses Standard SSD Disks:
"osDisk": {
"name": "myOsDisk",
"createOption": "FromImage",
"managedDisk": {
"storageAccountType": "StandardSSD_LRS"
}
}
For a complete template example of how to create a Standard SSD disk with a template, see Create a VM from a
Windows Image with Standard SSD Data Disks.
Can I convert my existing disks to Standard SSD? Yes, you can. Refer to Convert Azure managed disks storage
from standard to premium, and vice versa for the general guidelines for converting Managed Disks. And, use the
following value to update the disk type to Standard SSD. -AccountType StandardSSD_LRS
What is the benefit of using Standard SSD disks instead of HDD? Standard SSD disks deliver better latency,
consistency, availability and reliability compared to HDD disks. Application workloads run a lot more smoothly on
Standard SSD because of that. Note, Premium SSD disks are the recommended solution for most IO -intensive
production workloads.
Can I use Standard SSDs as Unmanaged Disks? No, Standard SSDs disks are only available as Managed Disks.
Do Standard SSD Disks support "single instance VM SLA"? No, Standard SSDs do not have single instance
VM SLA. Use Premium SSD disks for single instance VM SLA.
Migrate to Managed Disks

What changes are required in a pre-existing Azure Backup service configuration prior/after migration to
Managed Disks?
No changes are required.
Will my VM backups created via Azure Backup service before the migration continue to work?
Yes, backups work seamlessly.
What changes are required in a pre-existing Azure Disks Encryption configuration prior/after migration
to Managed Disks?
No changes are required.
Is automated migration of an existing virtual machine scale sets from unmanaged disks to Managed
Disks supported?
No. You can create a new scale set with Managed Disks using the image from your old scale set with unmanaged
disks.
Can I create a Managed Disk from a page blob snapshot taken before migrating to Managed Disks?
No. You can export a page blob snapshot as a page blob and then create a Managed Disk from the exported page
blob.
Can I fail over my on-premises machines protected by Azure Site Recovery to a VM with Managed
Disks?
Yes, you can choose to failover to a VM with Managed Disks.
Is there any impact of migration on Azure VMs protected by Azure Site Recovery via Azure to Azure
replication?
Yes. Currently, Azure Site Recovery Azure to Azure protection for VMs with Managed Disks is only available as a
public preview service.
Can I migrate VMs with unmanaged disks that are located on storage accounts that are or were
previously encrypted to managed disks?
Yes
Managed Disks and Storage Service Encryption

Is Azure Storage Service Encryption enabled by default when I create a managed disk?
Yes.
Who manages the encryption keys?
Microsoft manages the encryption keys.
Can I disable Storage Service Encryption for my managed disks?
No.
Is Storage Service Encryption only available in specific regions?
No. It's available in all the regions where Managed Disks are available. Managed Disks is available in all public
regions and Germany. It is also available in China, however, only for Microsoft managed keys, not customer
managed keys.
How can I find out if my managed disk is encrypted?
You can find out the time when a managed disk was created from the Azure portal, the Azure CLI, and PowerShell.
If the time is after June 9, 2017, then your disk is encrypted.
How can I encrypt my existing disks that were created before June 10, 2017?
As of June 10, 2017, new data written to existing managed disks is automatically encrypted. We are also planning
to encrypt existing data, and the encryption will happen asynchronously in the background. If you must encrypt
existing data now, create a copy of your disk. New disks will be encrypted.
Copy managed disks by using the Azure CLI
Copy managed disks by using PowerShell
Are managed snapshots and images encrypted?
Yes. All managed snapshots and images created after June 9, 2017, are automatically encrypted.
Can I convert VMs with unmanaged disks that are located on storage accounts that are or were
previously encrypted to managed disks?
Yes
Will an exported VHD from a managed disk or a snapshot also be encrypted?
No. But if you export a VHD to an encrypted storage account from an encrypted managed disk or snapshot, then
it's encrypted.
Premium disks: Managed and unmanaged

If a VM uses a size series that supports Premium SSD disks, such as a DSv2, can I attach both premium
and standard data disks?
Yes.
Can I attach both premium and standard data disks to a size series that doesn't support Premium SSD
disks, such as D, Dv2, G, or F series?
No. You can attach only standard data disks to VMs that don't use a size series that supports Premium SSD disks.
If I create a premium data disk from an existing VHD that was 80 GB, how much will that cost?
A premium data disk created from an 80-GB VHD is treated as the next-available premium disk size, which is a
P10 disk. You're charged according to the P10 disk pricing.
Are there transaction costs to use Premium SSD disks?
There is a fixed cost for each disk size, which comes provisioned with specific limits on IOPS and throughput. The
other costs are outbound bandwidth and snapshot capacity, if applicable. For more information, see the pricing
page.
What are the limits for IOPS and throughput that I can get from the disk cache?
The combined limits for cache and local SSD for a DS series are 4,000 IOPS per core and 33 MiB per second per
core. The GS series offers 5,000 IOPS per core and 50 MiB per second per core.
Is the local SSD supported for a Managed Disks VM?
The local SSD is temporary storage that is included with a Managed Disks VM. There is no extra cost for this
temporary storage. We recommend that you do not use this local SSD to store your application data because it
isn't persisted in Azure Blob storage.
Are there any repercussions for the use of TRIM on premium disks?
There is no downside to the use of TRIM on Azure disks on either premium or standard disks.
New disk sizes: Managed and unmanaged

What is the largest Managed disk size supported for operating system and data disks?
The partition type that Azure supports for an operating system disk is the master boot record (MBR ). The MBR
format supports a disk size up to 2 TiB. The largest size that Azure supports for an operating system disk is 2 TiB.
Azure supports up to 32 TiB for managed data disks. Managed Disk sizes larger than 4 TiB are in preview. For
more information on them see our blog post.
What is the largest Unmanaged Disk size supported for operating system and data disks?
The partition type that Azure supports for an operating system disk is the master boot record (MBR ). The MBR
format supports a disk size up to 2 TiB. The largest size that Azure supports for an operating system Unmanaged
disk is 2 TiB. Azure supports up to 4 TiB for data Unmanaged disks.
What is the largest page blob size that's supported?
The largest page blob size that Azure supports is 8 TiB (8,191 GiB ). The maximum page blog size when attached to
a VM as data or operating system disks is 4 TiB (4,095 GiB ).
Do I need to use a new version of Azure tools to create, attach, resize, and upload disks larger than 1
TiB?
You don't need to upgrade your existing Azure tools to create, attach, or resize disks larger than 1 TiB. To upload
your VHD file from on-premises directly to Azure as a page blob or unmanaged disk, you need to use the latest
tool sets listed below. We only support VHD uploads of up to 8 TiB.
AZURE TOOLS SUPPORTED VERSIONS
Azure PowerShell Version number 4.1.0: June 2017 release or later
Azure CLI v1 Version number 0.10.13: May 2017 release or later
Azure CLI v2 Version number 2.0.12: July 2017 release or later
AzCopy Version number 6.1.0: June 2017 release or later
Are P4 and P6 disk sizes supported for unmanaged disks or page blobs?
P4 (32 GiB ) and P6 (64 GiB ) disk sizes are not supported as the default disk tiers for unmanaged disks and page
blobs. You need to explicitly set the Blob Tier to P4 and P6 to have your disk mapped to these tiers. If you deploy a
unmanaged disk or page blob with the disk size or content length less than 32 GiB or between 32 GiB to 64 GiB
without setting the Blob Tier, you will continue to land on P10 with 500 IOPS and 100 MiB/s and the mapped
pricing tier.
If my existing premium managed disk less than 64 GiB was created before the small disk was enabled
(around June 15, 2017), how is it billed?
Existing small premium disks less than 64 GiB continue to be billed according to the P10 pricing tier.
How can I switch the disk tier of small premium disks less than 64 GiB from P10 to P4 or P6?
You can take a snapshot of your small disks and then create a disk to automatically switch the pricing tier to P4 or
P6 based on the provisioned size.
Can you resize existing Managed Disks from sizes less than 4 TiB to new newly introduced disk sizes up
to 32 TiB?
New Managed Disk sizes 8 TiB, 16 TiB, and 32 TiB are currently in Preview. We don’t yet support resizing existing
disk sizes to the new disk sizes.
What is the largest disk sizes supported by Azure Backup and Azure Site Recovery service?
The largest disk size supported by Azure Backup and Azure Site Recovery service is 4 TiB.
What are the recommended VM sizes for large disk sizes (>4TiB ) for Standard SSD and Standard HDD
disks to achieve optimized disk IOPS and Bandwidth?
To achieve the disk throughput of Standard SSD and Standard HDD large disk sizes (>4TB ) beyond 500 IOPS and
60 MiB/s, you should use one of the following VM sizes to optimize your performance: B -series, DSv2-series,
Dsv3-Series, ESv3-Series, Fs-series, Fsv2-series, M -series, GS -series, NCv2-series, NCv3-series, or Ls-Series VMs.
What regions are the managed disk sizes larger than 4 TiB supported in?
At this time in the preview, the managed disk sizes are supported in West US Central only.
Do we support enabling Host Caching on the newer disk sizes?
We support Host Caching of ReadOnly and Read/Write on disk sizes less than 4TiB. For disk sizes more than 4 TiB,
we don’t support setting caching option other than None. We recommend leveraging caching for smaller disk sizes
where you can expect to observe better performance boost with data cached to the VM.
What if my question isn't answered here?

If your question isn't listed here, let us know and we'll help you find an answer. You can post a question at the end
of this article in the comments. To engage with the Azure Storage team and other community members about this
article, use the MSDN Azure Storage forum.
To request features, submit your requests and ideas to the Azure Storage feedback forum.
Attach a data disk to a Windows VM with PowerShell
This article shows you how to attach both new and existing disks to a Windows virtual machine by using
PowerShell.
First, review these tips:
The size of the virtual machine controls how many data disks you can attach. For more information, see Sizes
for virtual machines.
To use Premium storage, you'll need a Premium Storage-enabled VM type like the DS -series or GS -series
virtual machine. For more information, see Premium Storage: High-Performance Storage for Azure Virtual
Machine Workloads.

To install and use the PowerShell locally, this tutorial requires the Azure PowerShell module version 6.0.0 or later.
Run Get-Module -ListAvailable AzureRM to find the version. If you need to upgrade, see Install Azure PowerShell
module. If you are running PowerShell locally, you'll also need to run Connect-AzureRmAccount to create a
Add an empty data disk to a virtual machine

This example shows how to add an empty data disk to an existing virtual machine.
Using managed disks
$rgName = 'myResourceGroup'
$vmName = 'myVM'
$location = 'East US'
$storageType = 'Premium_LRS'
$dataDiskName = $vmName + '_datadisk1'
$diskConfig = New-AzureRmDiskConfig -SkuName $storageType -Location $location -CreateOption Empty -DiskSizeGB

128
$dataDisk1 = New-AzureRmDisk -DiskName $dataDiskName -Disk $diskConfig -ResourceGroupName $rgName

$vm = Add-AzureRmVMDataDisk -VM $vm -Name $dataDiskName -CreateOption Attach -ManagedDiskId $dataDisk1.Id -Lun
1
Update-AzureRmVM -VM $vm -ResourceGroupName $rgName
Using managed disks in an Availability Zone

To create a disk in an Availability Zone, use New -AzureRmDiskConfig with the -Zone parameter. The following
example creates a disk in zone 1.
$vmName = 'myVM'
$location = 'East US 2'
$dataDiskName = $vmName + '_datadisk1'
$diskConfig = New-AzureRmDiskConfig -SkuName $storageType -Location $location -CreateOption Empty -DiskSizeGB

128 -Zone 1
$dataDisk1 = New-AzureRmDisk -DiskName $dataDiskName -Disk $diskConfig -ResourceGroupName $rgName

$vm = Add-AzureRmVMDataDisk -VM $vm -Name $dataDiskName -CreateOption Attach -ManagedDiskId $dataDisk1.Id -Lun
1
Initialize the disk

After you add an empty disk, you'll need to initialize it. To initialize the disk, you can sign in to a VM and use disk
management. If you enabled WinRM and a certificate on the VM when you created it, you can use remote
PowerShell to initialize the disk. You can also use a custom script extension:
$location = "location-name"
$scriptName = "script-name"
$fileName = "script-file-name"
Set-AzureRmVMCustomScriptExtension -ResourceGroupName $rgName -Location $locName -VMName $vmName -Name
$scriptName -TypeHandlerVersion "1.4" -StorageAccountName "mystore1" -StorageAccountKey "primary-key" -
FileName $fileName -ContainerName "scripts"
The script file can contain code to initialize the disks, for example:
$disks = Get-Disk | Where partitionstyle -eq 'raw' | sort number
$letters = 70..89 | ForEach-Object { [char]$_ }

$count = 0
$labels = "data1","data2"
foreach ($disk in $disks) {

$driveLetter = $letters[$count].ToString()
$disk |
Initialize-Disk -PartitionStyle MBR -PassThru |
New-Partition -UseMaximumSize -DriveLetter $driveLetter |
Format-Volume -FileSystem NTFS -NewFileSystemLabel $labels[$count] -Confirm:$false -Force
$count++
}
Attach an existing data disk to a VM

You can attach an existing managed disk to a VM as a data disk.
$vmName = "myVM"
$dataDiskName = "myDisk"
$disk = Get-AzureRmDisk -ResourceGroupName $rgName -DiskName $dataDiskName
$vm = Add-AzureRmVMDataDisk -CreateOption Attach -Lun 0 -VM $vm -ManagedDiskId $disk.Id
Next steps
Create a snapshot.
Attach a managed data disk to a Windows VM by
using the Azure portal
This article shows you how to attach a new managed data disk to a Windows virtual machine (VM ) by using the
Azure portal. The size of the VM determines how many data disks you can attach. For more information, see Sizes
for virtual machines.
Add a data disk

1. In the Azure portal, from the menu on the left, select Virtual machines.
2. Select a virtual machine from the list.
3. On the Virtual machine page, select Disks.
4. On the Disks page, select Add data disk.
5. In the drop-down for the new disk, select Create disk.
6. In the Create managed disk page, type in a name for the disk and adjust the other settings as necessary.
When you're done, select Create.
7. In the Disks page, select Save to save the new disk configuration for the VM.
8. After Azure creates the disk and attaches it to the virtual machine, the new disk is listed in the virtual machine's
disk settings under Data disks.
Initialize a new data disk

2. Select the Windows Start menu inside the running VM and enter diskmgmt.msc in the search box. The Disk
Management console opens.
3. Disk Management recognizes that you have a new, uninitialized disk and the Initialize Disk window appears.
4. Verify the new disk is selected and then select OK to initialize it.
5. The new disk appears as unallocated. Right-click anywhere on the disk and select New simple volume. The
New Simple Volume Wizard window opens.
6. Proceed through the wizard, keeping all of the defaults, and when you're done select Finish.
7. Close Disk Management.
8. A pop-up window appears notifying you that you need to format the new disk before you can use it. Select
Format disk.
9. In the Format new disk window, check the settings, and then select Start.
10. A warning appears notifying you that formatting the disks erases all of the data. Select OK.
11. When the formatting is complete, select OK.
Use TRIM with standard storage

If you use standard storage (HDD ), you should enable the TRIM command. The TRIM command discards
unused blocks on the disk so that you're billed only for storage that you're actually using. By using TRIM, you can
save on costs if you create large files and then later delete them.
To check the TRIM setting, open a command prompt on your Windows VM and enter the following command:
fsutil behavior query DisableDeleteNotify
If the command returns 0, TRIM is enabled correctly. Otherwise, if it returns 1, run the following command to
enable TRIM:
fsutil behavior set DisableDeleteNotify 0
After you delete data from your disk, you can ensure the TRIM operations flush properly by running defrag with
TRIM:
defrag.exe <volume:> -l
You can also format the volume to ensure the entire volume is trimmed.
Next steps
You can also attach a data disk by using PowerShell.
If your application needs to use the D: drive to store data, you can change the drive letter of the Windows
temporary disk.
How to detach a data disk from a Windows virtual
machine
When you no longer need a data disk that's attached to a virtual machine, you can easily detach it. This removes
the disk from the virtual machine, but doesn't remove it from storage.
WARNING
If you detach a disk it is not automatically deleted. If you have subscribed to Premium storage, you will continue to incur
storage charges for the disk. For more information, see Pricing and Billing when using Premium Storage.
If you want to use the existing data on the disk again, you can reattach it to the same virtual machine, or another
one.
Detach a data disk using PowerShell

You can hot remove a data disk using PowerShell, but make sure nothing is actively using the disk before detaching
it from the VM.
In this example, we remove the disk named myDisk from the VM myVM in the myResourceGroup resource
group. First you remove the disk using the Remove-AzureRmVMDataDisk cmdlet. Then, you update the state of
the virtual machine, using the Update-AzureRmVM cmdlet, to complete the process of removing the data disk.
$VirtualMachine = Get-AzureRmVM -ResourceGroupName "myResourceGroup" -Name "myVM"

Remove-AzureRmVMDataDisk -VM $VirtualMachine -Name "myDisk"
Update-AzureRmVM -ResourceGroupName "myResourceGroup" -VM $VirtualMachine
The disk stays in storage but is no longer attached to a virtual machine.
Detach a data disk using the portal

1. In the left menu, select Virtual Machines.
2. Select the virtual machine that has the data disk you want to detach and click Stop to deallocate the VM.
3. In the virtual machine pane, select Disks.
4. At the top of the Disks pane, select Edit.
5. In the Disks pane, to the far right of the data disk that you would like to detach, click the detach button.
6. After the disk has been removed, click Save on the top of the pane.
7. In the virtual machine pane, click Overview and then click the Start button at the top of the pane to restart the
VM.
The disk stays in storage but is no longer attached to a virtual machine.
Next steps
If you want to reuse the data disk, you can just attach it to another VM
How to expand the OS drive of a virtual machine
When you create a new virtual machine (VM ) in a Resource Group by deploying an image from Azure
Marketplace, the default OS drive is often 127 GB (some images have smaller OS disk sizes by default). Even
though it’s possible to add data disks to the VM (how many depending upon the SKU you’ve chosen) and
moreover it’s recommended to install applications and CPU intensive workloads on these addendum disks,
oftentimes customers need to expand the OS drive to support certain scenarios such as following:
Support legacy applications that install components on OS drive.
Migrate a physical PC or virtual machine from on-premises with a larger OS drive.
IMPORTANT
Resizing the OS Disk of an Azure Virtual Machine will cause it to restart.
After expanding the disks, you need to expand the volume within the OS to take advantage of the larger disk.
Resize a managed disk

Open your Powershell ISE or Powershell window in administrative mode and follow the steps below:
1. Sign in to your Microsoft Azure account in resource management mode and select your subscription as
follows:
Select-AzureRmSubscription –SubscriptionName 'my-subscription-name'
2. Set your resource group name and VM name as follows:
$rgName = 'my-resource-group-name'
$vmName = 'my-vm-name'
3. Obtain a reference to your VM as follows:
$vm = Get-AzureRmVM -ResourceGroupName $rgName -Name $vmName
4. Stop the VM before resizing the disk as follows:
Stop-AzureRmVM -ResourceGroupName $rgName -Name $vmName
5. Obtain a reference to the managed OS disk. Set the size of the managed OS disk to the desired value and
update the Disk as follows:
$disk= Get-AzureRmDisk -ResourceGroupName $rgName -DiskName $vm.StorageProfile.OsDisk.Name

$disk.DiskSizeGB = 1023
Update-AzureRmDisk -ResourceGroupName $rgName -Disk $disk -DiskName $disk.Name
WARNING
The new size should be greater than the existing disk size. The maximum allowed is 2048 GB for OS disks. (It is
possible to expand the VHD blob beyond that size, but the OS will only be able to work with the first 2048 GB of
space.)
6. Updating the VM may take a few seconds. Once the command finishes executing, restart the VM as follows:
Start-AzureRmVM -ResourceGroupName $rgName -Name $vmName
And that’s it! Now RDP into the VM, open Computer Management (or Disk Management) and expand the drive
using the newly allocated space.
Resize an unmanaged disk

Open your Powershell ISE or Powershell window in administrative mode and follow the steps below:
1. Sign in to your Microsoft Azure account in resource management mode and select your subscription as
follows:
Select-AzureRmSubscription –SubscriptionName 'my-subscription-name'
2. Set your resource group name and VM name as follows:
3. Obtain a reference to your VM as follows:
4. Stop the VM before resizing the disk as follows:
5. Set the size of the unmanaged OS disk to the desired value and update the VM as follows:
$vm.StorageProfile.OSDisk.DiskSizeGB = 1023
Update-AzureRmVM -ResourceGroupName $rgName -VM $vm
WARNING
The new size should be greater than the existing disk size. The maximum allowed is 2048 GB for OS disks. (It is
possible to expand the VHD blob beyond that size, but the OS will only be able to work with the first 2048 GB of
space.)
6. Updating the VM may take a few seconds. Once the command finishes executing, restart the VM as follows:
Scripts for OS disk

Below is the complete script for your reference for both managed and unmanaged disks:
Managed disks
Select-AzureRmSubscription -SubscriptionName 'my-subscription-name'
$disk= Get-AzureRmDisk -ResourceGroupName $rgName -DiskName $vm.StorageProfile.OsDisk.Name
Update-AzureRmDisk -ResourceGroupName $rgName -Disk $disk -DiskName $disk.Name
Unmanaged disks
Select-AzureRmSubscription -SubscriptionName 'my-subscription-name'
$vm.StorageProfile.OSDisk.DiskSizeGB = 1023
Update-AzureRmVM -ResourceGroupName $rgName -VM $vm
Resizing data disks

This article is focused primarily on expanding the OS disk of the VM, but the script can also be used for expanding
the data disks attached to the VM. For example, to expand the first data disk attached to the VM, replace the
OSDisk object of StorageProfile with DataDisks array and use a numeric index to obtain a reference to first
attached data disk, as shown below:
Managed disk
$disk= Get-AzureRmDisk -ResourceGroupName $rgName -DiskName $vm.StorageProfile.DataDisks[0].Name

Unmanaged disk
$vm.StorageProfile.DataDisks[0].DiskSizeGB = 1023
Similarly you may reference other data disks attached to the VM, either by using an index as shown above or the
Name property of the disk:
Managed disk
(Get-AzureRmDisk -ResourceGroupName $rgName -DiskName ($vm.StorageProfile.DataDisks | Where ({$_.Name -eq 'my-

second-data-disk'})).Name).DiskSizeGB = 1023
Unmanaged disk
($vm.StorageProfile.DataDisks | Where ({$_.Name -eq 'my-second-data-disk'}).DiskSizeGB = 1023
Expand the volume within the OS

Once you have expanded the disk for the VM, you need to go into the OS and expand the volume to encompass
the new space. There are several methods for expanding a partition. This section covers connecting the VM using
an RDP connection to expand the partition using DiskPart.
1. Open an RDP connection to your VM.
2. Open a command prompt and type diskpart.
3. At the DISKPART prompt, type list volume . Make note of the volume you want to extend.
4. At the DISKPART prompt, type select volume <volumenumber> . This selects the volume volumenumber that
you want to extend into contiguous, empty space on the same disk.
5. At the DISKPART prompt, type extend [size=<size>] . This extends the selected volume by size in
megabytes (MB ).
Next steps
You can also attach disks using the Azure portal.
Create a snapshot
A snapshot is a full, read-only copy of a virtual hard drive (VHD ). You can take a snapshot of an OS or data disk
VHD to use as a backup, or to troubleshoot virtual machine (VM ) issues.
Use the Azure portal

2. From the left menu, select Create a resource, and then search for and select snapshot.
3. In the Snapshot window, select Create. The Create snapshot window appears.
4. Enter a Name for the snapshot.
5. Select an existing Resource group or enter the name of a new one.
6. Select an Azure datacenter Location.
7. For Source disk, select the managed disk to snapshot.
8. Select the Account type to use to store the snapshot. Select Standard_HDD, unless you need the snapshot to
be stored on a high-performing disk.
9. Select Create.
Use PowerShell
The following steps show how to copy the VHD disk, create the snapshot configuration, and take a snapshot of the
disk by using the New -AzureRmSnapshot cmdlet.
Before you begin, ensure you have the latest version of the AzureRM.Compute PowerShell module, which must
be version 5.7.0 or later. Run Get-Module -ListAvailable AzureRM to find the version. If you need to upgrade, see
Install Azure PowerShell module. If you're running PowerShell locally, run Connect-AzureRmAccount to create a
1. Set some parameters:
$location = 'eastus'
$vmName = 'myVM'
$snapshotName = 'mySnapshot'
2. Get the VM:
$vm = get-azurermvm `
-Name $vmName
3. Create the snapshot configuration. For this example, the snapshot is of the OS disk:
$snapshot = New-AzureRmSnapshotConfig
-SourceUri $vm.StorageProfile.OsDisk.ManagedDisk.Id
-Location $location
-CreateOption copy
NOTE
If you would like to store your snapshot in zone-resilient storage, create it in a region that supports availability zones
and include the -SkuName Standard_ZRS parameter.
4. Take the snapshot:
New-AzureRmSnapshot
-Snapshot $snapshot
-SnapshotName $snapshotName
Next steps
Create a virtual machine from a snapshot by creating a managed disk from a snapshot and then attaching the new
managed disk as the OS disk. For more information, see the sample in Create a VM from a snapshot with
PowerShell.
Back up Azure unmanaged VM disks with incremental
snapshots
Overview
Azure Storage provides the capability to take snapshots of blobs. Snapshots capture the blob state at that point in
time. In this article, we describe a scenario in which you can maintain backups of virtual machine disks using
snapshots. You can use this methodology when you choose not to use Azure Backup and Recovery Service, and
wish to create a custom backup strategy for your virtual machine disks.
Azure virtual machine disks are stored as page blobs in Azure Storage. Since we are describing a backup strategy
for virtual machine disks in this article, we refer to snapshots in the context of page blobs. To learn more about
snapshots, refer to Creating a Snapshot of a Blob.
What is a snapshot?
A blob snapshot is a read-only version of a blob that is captured at a point in time. Once a snapshot has been
created, it can be read, copied, or deleted, but not modified. Snapshots provide a way to back up a blob as it
appears at a moment in time. Until REST version 2015-04-05, you had the ability to copy full snapshots. With the
REST version 2015-07-08 and above, you can also copy incremental snapshots.
Full snapshot copy

Snapshots can be copied to another storage account as a blob to keep backups of the base blob. You can also copy
a snapshot over its base blob, which is like restoring the blob to an earlier version. When a snapshot is copied from
one storage account to another, it occupies the same space as the base page blob. Therefore, copying whole
snapshots from one storage account to another is slow and consumes much space in the target storage account.
NOTE
If you copy the base blob to another destination, the snapshots of the blob are not copied along with it. Similarly, if you
overwrite a base blob with a copy, snapshots associated with the base blob are not affected and stay intact under the base
blob name.
Back up disks using snapshots

As a backup strategy for your virtual machine disks, you can take periodic snapshots of the disk or page blob, and
copy them to another storage account using tools like Copy Blob operation or AzCopy. You can copy a snapshot to
a destination page blob with a different name. The resulting destination page blob is a writeable page blob and not
a snapshot. Later in this article, we describe steps to take backups of virtual machine disks using snapshots.
Restore disks using snapshots
When it is time to restore your disk to a stable version that was previously captured in one of the backup
snapshots, you can copy a snapshot over the base page blob. After the snapshot is promoted to the base page blob,
the snapshot remains, but its source is overwritten with a copy that can be both read and written. Later in this
article we describe steps to restore a previous version of your disk from its snapshot.
Implementing full snapshot copy
You can implement a full snapshot copy by doing the following,
First, take a snapshot of the base blob using the Snapshot Blob operation.
Then, copy the snapshot to a target storage account using Copy Blob.
Repeat this process to maintain backup copies of your base blob.
Incremental snapshot copy

The new feature in the GetPageRanges API provides a much better way to back up the snapshots of your page
blobs or disks. The API returns the list of changes between the base blob and the snapshots, which reduces the
amount of storage space used on the backup account. The API supports page blobs on Premium Storage as well as
Standard Storage. Using this API, you can build faster and more efficient backup solutions for Azure VMs. This API
will be available with the REST version 2015-07-08 and higher.
Incremental Snapshot Copy allows you to copy from one storage account to another the difference between,
Base blob and its Snapshot OR
Any two snapshots of the base blob
Provided the following conditions are met,
The blob was created on Jan-1-2016 or later.
The blob was not overwritten with PutPage or Copy Blob between two snapshots.
Note: This feature is available for Premium and Standard Azure Page Blobs.
When you have a custom backup strategy using snapshots, copying the snapshots from one storage account to
another can be slow and can consume much storage space. Instead of copying the entire snapshot to a backup
storage account, you can write the difference between consecutive snapshots to a backup page blob. This way, the
time to copy and the space to store backups is substantially reduced.
Implementing Incremental Snapshot Copy
You can implement incremental snapshot copy by doing the following,
Take a snapshot of the base blob using Snapshot Blob.
Copy the snapshot to the target backup storage account using Copy Blob. This is the backup page blob. Take a
snapshot of the backup page blob and store it in the backup account.
Take another snapshot of the base blob using Snapshot Blob.
Get the difference between the first and second snapshots of the base blob using GetPageRanges. Use the new
parameter prevsnapshot, to specify the DateTime value of the snapshot you want to get the difference with.
When this parameter is present, the REST response includes only the pages that were changed between target
snapshot and previous snapshot including clear pages.
Use PutPage to apply these changes to the backup page blob.
Finally, take a snapshot of the backup page blob and store it in the backup storage account.
In the next section, we will describe in more detail how you can maintain backups of disks using Incremental
Snapshot Copy
Scenario
In this section, we describe a scenario that involves a custom backup strategy for virtual machine disks using
snapshots.
Consider a DS -series Azure VM with a premium storage P30 disk attached. The P30 disk called mypremiumdisk is
stored in a premium storage account called mypremiumaccount. A standard storage account called
mybackupstdaccount is used for storing the backup of mypremiumdisk. We would like to keep a snapshot of
mypremiumdisk every 12 hours.
To learn about creating a storage account, see Create a storage account.
To learn about backing up Azure VMs, refer to Plan Azure VM backups.
Steps to maintain backups of a disk using incremental snapshots

The following steps describe how to take snapshots of mypremiumdisk and maintain the backups in
mybackupstdaccount. The backup is a standard page blob called mybackupstdpageblob. The backup page blob
always reflects the same state as the last snapshot of mypremiumdisk.
1. Create the backup page blob for your premium storage disk, by taking a snapshot of mypremiumdisk called
mypremiumdisk_ss1.
2. Copy this snapshot to mybackupstdaccount as a page blob called mybackupstdpageblob.
3. Take a snapshot of mybackupstdpageblob called mybackupstdpageblob_ss1, using Snapshot Blob and store it
in mybackupstdaccount.
4. During the backup window, create another snapshot of mypremiumdisk, say mypremiumdisk_ss2, and store it in
mypremiumaccount.
5. Get the incremental changes between the two snapshots, mypremiumdisk_ss2 and mypremiumdisk_ss1, using
GetPageRanges on mypremiumdisk_ss2 with the prevsnapshot parameter set to the timestamp of
mypremiumdisk_ss1. Write these incremental changes to the backup page blob mybackupstdpageblob in
mybackupstdaccount. If there are deleted ranges in the incremental changes, they must be cleared from the
backup page blob. Use PutPage to write incremental changes to the backup page blob.
6. Take a snapshot of the backup page blob mybackupstdpageblob, called mybackupstdpageblob_ss2. Delete the
previous snapshot mypremiumdisk_ss1 from premium storage account.
7. Repeat steps 4-6 every backup window. In this way, you can maintain backups of mypremiumdisk in a standard
storage account.
Steps to restore a disk from snapshots

The following steps, describe how to restore the premium disk, mypremiumdisk to an earlier snapshot from the
backup storage account mybackupstdaccount.
1. Identify the point in time that you wish to restore the premium disk to. Let's say that it is snapshot
mybackupstdpageblob_ss2, which is stored in the backup storage account mybackupstdaccount.
2. In mybackupstdaccount, promote the snapshot mybackupstdpageblob_ss2 as the new backup base page blob
mybackupstdpageblobrestored.
3. Take a snapshot of this restored backup page blob, called mybackupstdpageblobrestored_ss1.
4. Copy the restored page blob mybackupstdpageblobrestored from mybackupstdaccount to mypremiumaccount
as the new premium disk mypremiumdiskrestored.
5. Take a snapshot of mypremiumdiskrestored, called mypremiumdiskrestored_ss1 for making future incremental
backups.
6. Point the DS series VM to the restored disk mypremiumdiskrestored and detach the old mypremiumdisk from
the VM.
7. Begin the Backup process described in previous section for the restored disk mypremiumdiskrestored, using the
mybackupstdpageblobrestored as the backup page blob.
Next Steps
Use the following links to learn more about creating snapshots of a blob and planning your VM backup
infrastructure.
Creating a Snapshot of a Blob
Plan your VM Backup Infrastructure
Update the storage type of a managed disk
Azure Managed Disks offers three storage type options: Premium SSD, Standard SSD, and Standard HDD. You
can switch a managed disk between storage types with minimal downtime, based on your performance needs.
Switching between storage types is not supported for an unmanaged disk; however, you can easily convert an
unmanaged disk to a managed disk.
This article shows how to convert a managed disk from standard to premium, and vice versa, by using Azure
PowerShell. If you need to install or upgrade PowerShell, see Install and configure Azure PowerShell.
Prerequisites
Because the conversion requires a restart of the virtual machine (VM ), you should schedule the migration of
your disks storage during a pre-existing maintenance window.
If you're using an unmanaged disk, first convert it to a managed disk to allow you to switch it between the
storage types.
The examples in this article require the Azure PowerShell module version 6.0.0 or later. Run
Get-Module -ListAvailable AzureRM to find the version. If you need to upgrade, see Install Azure PowerShell
module. Run Connect-AzureRmAccount to create a connection with Azure.
Convert all the managed disks of a VM from standard to premium

The following example shows how to switch all the disks of a VM from standard to premium storage. To use
premium-managed disks, your VM must use a VM size that supports premium storage. This example also
switches to a size that supports premium storage:
# Name of the resource group that contains the VM
$rgName = 'yourResourceGroup'
# Name of the your virtual machine

$vmName = 'yourVM'
# Choose between StandardLRS and PremiumLRS based on your scenario

# Premium capable size

# Required only if converting storage from standard to premium
$size = 'Standard_DS2_v2'
# Stop and deallocate the VM before changing the size

$vm = Get-AzureRmVM -Name $vmName -resourceGroupName $rgName
# Change the VM size to a size that supports premium storage

# Skip this step if converting storage from premium to standard
$vm.HardwareProfile.VmSize = $size
# Get all disks in the resource group of the VM

$vmDisks = Get-AzureRmDisk -ResourceGroupName $rgName
# For disks that belong to the selected VM, convert to premium storage
foreach ($disk in $vmDisks)
{
if ($disk.ManagedBy -eq $vm.Id)
{
$diskUpdateConfig = New-AzureRmDiskUpdateConfig –AccountType $storageType
Update-AzureRmDisk -DiskUpdate $diskUpdateConfig -ResourceGroupName $rgName `
-DiskName $disk.Name
}
}
Convert a managed disk from standard to premium

For your dev/test workload, you may want a mixture of standard and premium disks to reduce your cost. To dos
so, upgrade to premium storage only those disks that require better performance. The following example shows
how to switch a single disk of a VM from standard to premium storage, and vice versa. To use premium-managed
disks, your VM must use a VM size that supports premium storage. This example also shows how to switch to a
size that supports premium storage:
$diskName = 'yourDiskName'
# resource group that contains the managed disk
$rgName = 'yourResourceGroupName'
# Choose between StandardLRS and PremiumLRS based on your scenario
# Premium capable size
$size = 'Standard_DS2_v2'
$disk = Get-AzureRmDisk -DiskName $diskName -ResourceGroupName $rgName
# Get parent VM resource

$vmResource = Get-AzureRmResource -ResourceId $disk.ManagedBy
# Stop and deallocate the VM before changing the storage type

Stop-AzureRmVM -ResourceGroupName $vmResource.ResourceGroupName -Name $vmResource.Name -Force
$vm = Get-AzureRmVM -ResourceGroupName $vmResource.ResourceGroupName -Name $vmResource.Name
# Change the VM size to a size that supports premium storage

# Skip this step if converting storage from premium to standard
$vm.HardwareProfile.VmSize = $size
# Update the storage type

$diskUpdateConfig = New-AzureRmDiskUpdateConfig -AccountType $storageType -DiskSizeGB $disk.DiskSizeGB
Start-AzureRmVM -ResourceGroupName $vm.ResourceGroupName -Name $vm.Name
Convert a managed disk from standard HDD to standard SSD

The following example shows how to switch a single disk of a VM from standard HDD to standard SSD, and vice
versa:
$diskName = 'yourDiskName'
# resource group that contains the managed disk
$rgName = 'yourResourceGroupName'
# Choose between Standard_LRS and StandardSSD_LRS based on your scenario
$storageType = 'StandardSSD_LRS'
$disk = Get-AzureRmDisk -DiskName $diskName -ResourceGroupName $rgName
# Get parent VM resource

$vmResource = Get-AzureRmResource -ResourceId $disk.ManagedBy
# Stop and deallocate the VM before changing the storage type

Stop-AzureRmVM -ResourceGroupName $vmResource.ResourceGroupName -Name $vmResource.Name -Force
$vm = Get-AzureRmVM -ResourceGroupName $vmResource.ResourceGroupName -Name $vmResource.Name
# Update the storage type

$diskUpdateConfig = New-AzureRmDiskUpdateConfig -AccountType $storageType -DiskSizeGB $disk.DiskSizeGB
Start-AzureRmVM -ResourceGroupName $vm.ResourceGroupName -Name $vm.Name
Next steps
Make a read-only copy of a VM by using a snapshot.
Migrate to Premium Storage by using Azure Site
Recovery
Azure Premium Storage delivers high-performance, low -latency disk support for virtual machines (VMs) that are
running I/O -intensive workloads. This guide helps you migrate your VM disks from a standard storage account to
a premium storage account by using Azure Site Recovery.
Site Recovery is an Azure service that contributes to your strategy for business continuity and disaster recovery by
orchestrating the replication of on-premises physical servers and VMs to the cloud (Azure) or to a secondary
datacenter. When outages occur in your primary location, you fail over to the secondary location to keep
applications and workloads available. You fail back to your primary location when it returns to normal operation.
Site Recovery provides test failovers to support disaster recovery drills without affecting production environments.
You can run failovers with minimal data loss (depending on replication frequency) for unexpected disasters. In the
scenario of migrating to Premium Storage, you can use the failover in Site Recovery to migrate target disks to a
premium storage account.
We recommend migrating to Premium Storage by using Site Recovery because this option provides minimal
downtime. This option also avoids the manual execution of copying disks and creating new VMs. Site Recovery will
systematically copy your disks and create new VMs during failover.
Site Recovery supports a number of types of failover with minimal or no downtime. To plan your downtime and
estimate data loss, see the types of failover in Site Recovery. If you prepare to connect to Azure VMs after failover,
you should be able to connect to the Azure VM by using RDP after failover.
Azure Site Recovery components

These Site Recovery components are relevant to this migration scenario:
Configuration server is an Azure VM that coordinates communication and manages data replication and
recovery processes. On this VM, you run a single setup file to install the configuration server and an
additional component, called a process server, as a replication gateway. Read about configuration server
prerequisites. You set up the configuration server only once, and you can use it for all migrations to the
same region.
Process server is a replication gateway that:
1. Receives replication data from source VMs.
2. Optimizes the data with caching, compression, and encryption.
3. Sends the data to a storage account.
It also handles push installation of the mobility service to source VMs and performs automatic discovery of
source VMs. The default process server is installed on the configuration server. You can deploy additional
standalone process servers to scale your deployment. Read about best practices for process server
deployment and deploying additional process servers. You set up the process server only once, and you can
use it for all migrations to the same region.
Mobility service is a component that is deployed on every standard VM that you want to replicate. It
captures data writes on the standard VM and forwards them to the process server. Read about replicated
machine prerequisites.
This graphic shows how these components interact:
NOTE
Site Recovery does not support the migration of Storage Spaces disks.
For additional components for other scenarios, see Scenario architecture.
Azure essentials
These are the Azure requirements for this migration scenario:
An Azure subscription.
An Azure premium storage account to store replicated data.
An Azure virtual network to which VMs will connect when they're created at failover. The Azure virtual network
must be in the same region as the one in which Site Recovery runs.
An Azure standard storage account to store replication logs. This can be the same storage account for the VM
disks that are being migrated.
Prerequisites
Understand the relevant migration scenario components in the preceding section.
Plan your downtime by learning about failover in Site Recovery.
Setup and migration steps

You can use Site Recovery to migrate Azure IaaS VMs between regions or within same region. The following
instructions are tailored for this migration scenario from the article Replicate VMware VMs or physical servers to
Azure. Please follow the links for detailed steps in addition to the instructions in this article.
Step 1: Create a Recovery Services vault
2. Select Create a resource > Management > Backup and Site Recovery (OMS ). Alternatively, you can select
Browse > Recovery Services Vault > Add. >[!NOTE ] >Backup and Site Recovery was formerly part of the
.
3. Specify a region that VMs will be replicated to. For the purpose of migration in the same region, select the
region where your source VMs and source storage accounts are.
Step 2: Choose your protection goals
1. On the VM where you want to install the configuration server, open the Azure portal.
2. Go to Recovery Services vaults > Settings > Site Recovery > Step 1: Prepare Infrastructure >
Protection goal.
3. Under Protection goal, in the first drop-down list, select To Azure. In the second drop-down list, select
Not virtualized / Other, and then select OK.
Step 3: Set up the source environment (configuration server)
1. Download Azure Site Recovery Unified Setup and the vault registration key by going to the Prepare
infrastructure > Prepare source > Add Server panes.
You will need the vault registration key to run the unified setup. The key is valid for five days after you
generate it.
2. In the Add Server pane, add a configuration server.

3. On the VM that you're using as the configuration server, run Unified Setup to install the configuration server
and the process server. You can walk through the screenshots to complete the installation. You can refer to
the following screenshots for steps specified for this migration scenario.
a. In Before You Begin, select Install the configuration server and process server.
b. In Registration, browse and select the registration key that you downloaded from the vault.
c. In Environment Details, select whether you're going to replicate VMware VMs. For this migration
scenario, choose No.
4. After the installation is complete, do the following in the Microsoft Azure Site Recovery Configuration
Server window:
a. Use the Manage Accounts tab to create the account that Site Recovery can use for automatic discovery.
(In the scenario about protecting physical machines, setting up the account isn't relevant, but you need at
least one account to enable one of the following steps. In this case, you can name the account and
password as any.)
b. Use the Vault Registration tab to upload the vault credential file.
Step 4: Set up the target environment

Select Prepare infrastructure > Target, and specify the deployment model that you want to use for VMs after
failover. You can choose Classic or Resource Manager, depending on your scenario.
Site Recovery checks that you have one or more compatible Azure storage accounts and networks.
NOTE
If you're using a premium storage account for replicated data, you need to set up an additional standard storage account to
store replication logs.
Step 5: Set up replication settings

To verify that your configuration server is successfully associated with the replication policy that you create, follow
Set up replication settings.
Step 6: Plan capacity
1. Use the capacity planner to accurately estimate network bandwidth, storage, and other requirements to meet
your replication needs.
2. When you're done, select Yes, I have done it in Have you completed capacity planning?.
Step 7: Install the mobility service and enable replication
1. You can choose to push installation to your source VMs or to manually install the mobility service on your
source VMs. You can find the requirement of pushing installation and the path of the manual installer in the
provided link. If you're doing a manual installation, you might need to use an internal IP address to find the
configuration server.
The failed-over VM will have two temporary disks: one from the primary VM and the other created during
the provisioning of the VM in the recovery region. To exclude the temporary disk before replication, install
the mobility service before you enable replication. To learn more about how to exclude the temporary disk,
see Exclude disks from replication.
2. Enable replication as follows:
a. Select Replicate Application > Source. After you've enabled replication for the first time, select
+Replicate in the vault to enable replication for additional machines.
b. In step 1, set up Source as your process server.
c. In step 2, specify the post-failover deployment model, a premium storage account to migrate to, a
standard storage account to save logs, and a virtual network to fail to.
d. In step 3, add protected VMs by IP address. (You might need an internal IP address to find them.)
e. In step 4, configure the properties by selecting the accounts that you set up previously on the process
server.
f. In step 5, choose the replication policy that you created previously in "Step 5: Set up replication settings."
g. Select OK.
NOTE
When an Azure VM is deallocated and started again, there is no guarantee that it will get the same IP address. If the
IP address of the configuration server/process server or the protected Azure VMs changes, the replication in this
scenario might not work correctly.
When you design your Azure Storage environment, we recommend that you use separate storage accounts for
each VM in an availability set. We recommend that you follow the best practice in the storage layer to use multiple
storage accounts for each availability set. Distributing VM disks to multiple storage accounts helps to improve
storage availability and distributes the I/O across the Azure storage infrastructure.
If your VMs are in an availability set, instead of replicating disks of all VMs into one storage account, we highly
recommend migrating multiple VMs multiple times. That way, the VMs in the same availability set do not share a
single storage account. Use the Enable Replication pane to set up a destination storage account for each VM, one
at a time.
You can choose a post-failover deployment model according to your need. If you choose Azure Resource Manager
as your post-failover deployment model, you can fail over a VM (Resource Manager) to a VM (Resource Manager),
or you can fail over a VM (classic) to a VM (Resource Manager).
Step 8: Run a test failover
To check whether your replication is complete, select your Site Recovery instance and then select Settings >
Replicated Items. You will see the status and percentage of your replication process.
After initial replication is complete, run a test failover to validate your replication strategy. For detailed steps of a
test failover, see Run a test failover in Site Recovery.
NOTE
Before you run any failover, make sure that your VMs and replication strategy meet the requirements. For more information
about running a test failover, see Test failover to Azure in Site Recovery.
You can see the status of your test failover in Settings > Jobs > YOUR_FAILOVER_PLAN_NAME. In the pane, you
can see a breakdown of the steps and success/failure results. If the test failover fails at any step, select the step to
check the error message.
Step 9: Run a failover
After the test failover is completed, run a failover to migrate your disks to Premium Storage and replicate the VM
instances. Follow the detailed steps in Run a failover.
Be sure to select Shut down VMs and synchronize the latest data. This option specifies that Site Recovery
should try to shut down the protected VMs and synchronize the data so that the latest version of the data will be
failed over. If you don't select this option or the attempt doesn't succeed, the failover will be from the latest
available recovery point for the VM.
Site Recovery will create a VM instance whose type is the same as or similar to a Premium Storage-capable VM.
You can check the performance and price of various VM instances by going to Windows Virtual Machines Pricing
or Linux Virtual Machines Pricing.
Post-migration steps
1. Configure replicated VMs to the availability set if applicable. Site Recovery does not support
migrating VMs along with the availability set. Depending on the deployment of your replicated VM, do one
of the following:
For a VM created through the classic deployment model: Add the VM to the availability set in the Azure
portal. For detailed steps, go to Add an existing virtual machine to an availability set.
For a VM created through the Resource Manager deployment model: Save your configuration of the VM
and then delete and re-create the VMs in the availability set. To do so, use the script at Set Azure
Resource Manager VM Availability Set. Before you run this script, check its limitations and plan your
downtime.
2. Delete old VMs and disks. Make sure that the Premium disks are consistent with source disks and that the
new VMs perform the same function as the source VMs. Delete the VM and delete the disks from your
source storage accounts in the Azure portal. If there's a problem in which the disk is not deleted even though
you deleted the VM, see Troubleshoot storage resource deletion errors.
3. Clean the Azure Site Recovery infrastructure. If Site Recovery is no longer needed, you can clean its
infrastructure. Delete replicated items, the configuration server, and the recovery policy, and then delete the
Azure Site Recovery vault.
Troubleshooting
Monitor and troubleshoot protection for virtual machines and physical servers
Microsoft Azure Site Recovery forum
Next steps
For specific scenarios for migrating virtual machines, see the following resources:
Migrate Azure Virtual Machines between Storage Accounts
Create and upload a Windows Server VHD to Azure
Migrating Virtual Machines from Amazon AWS to Microsoft Azure
Also, see the following resources to learn more about Azure Storage and Azure Virtual Machines:
Azure Storage
Azure Virtual Machines
Premium Storage: High-performance storage for Azure virtual machine workloads
Find and delete unattached Azure managed and
unmanaged disks
When you delete a virtual machine (VM ) in Azure, by default, any disks that are attached to the VM aren't deleted.
This feature helps to prevent data loss due to the unintentional deletion of VMs. After a VM is deleted, you will
continue to pay for unattached disks. This article shows you how to find and delete any unattached disks and
reduce unnecessary costs.
Managed disks: Find and delete unattached disks

The following script looks for unattached managed disks by examining the value of the ManagedBy property.
When a managed disk is attached to a VM, the ManagedBy property contains the resource ID of the VM. When a
managed disk is unattached, the ManagedBy property is null. The script examines all the managed disks in an
Azure subscription. When the script locates a managed disk with the ManagedBy property set to null, the script
determines that the disk is unattached.
IMPORTANT
First, run the script by setting the deleteUnattachedDisks variable to 0. This action lets you find and view all the unattached
managed disks.
After you review all the unattached disks, run the script again and set the deleteUnattachedDisks variable to 1. This action
lets you delete all the unattached managed disks.
# Set deleteUnattachedDisks=1 if you want to delete unattached Managed Disks
# Set deleteUnattachedDisks=0 if you want to see the Id of the unattached Managed Disks
$deleteUnattachedDisks=0
$managedDisks = Get-AzureRmDisk
foreach ($md in $managedDisks) {
# ManagedBy property stores the Id of the VM to which Managed Disk is attached to

# If ManagedBy property is $null then it means that the Managed Disk is not attached to a VM
if($md.ManagedBy -eq $null){
if($deleteUnattachedDisks -eq 1){
Write-Host "Deleting unattached Managed Disk with Id: $($md.Id)"
$md | Remove-AzureRmDisk -Force
Write-Host "Deleted unattached Managed Disk with Id: $($md.Id) "
}else{
$md.Id
Unmanaged disks: Find and delete unattached disks

Unmanaged disks are VHD files that are stored as page blobs in Azure storage accounts. The following script looks
for unattached unmanaged disks (page blobs) by examining the value of the LeaseStatus property. When an
unmanaged disk is attached to a VM, the LeaseStatus property is set to Locked. When an unmanaged disk is
unattached, the LeaseStatus property is set to Unlocked. The script examines all the unmanaged disks in all the
Azure storage accounts in an Azure subscription. When the script locates an unmanaged disk with a LeaseStatus
property set to Unlocked, the script determines that the disk is unattached.
IMPORTANT
First, run the script by setting the deleteUnattachedVHDs variable to 0. This action lets you find and view all the unattached
unmanaged VHDs.
After you review all the unattached disks, run the script again and set the deleteUnattachedVHDs variable to 1. This action
lets you delete all the unattached unmanaged VHDs.
# Set deleteUnattachedVHDs=1 if you want to delete unattached VHDs
# Set deleteUnattachedVHDs=0 if you want to see the Uri of the unattached VHDs
$deleteUnattachedVHDs=0
$storageAccounts = Get-AzureRmStorageAccount
foreach($storageAccount in $storageAccounts){
$storageKey = (Get-AzureRmStorageAccountKey -ResourceGroupName $storageAccount.ResourceGroupName -Name

$storageAccount.StorageAccountName)[0].Value
$context = New-AzureStorageContext -StorageAccountName $storageAccount.StorageAccountName -

StorageAccountKey $storageKey
$containers = Get-AzureStorageContainer -Context $context
foreach($container in $containers){
$blobs = Get-AzureStorageBlob -Container $container.Name -Context $context
#Fetch all the Page blobs with extension .vhd as only Page blobs can be attached as disk to Azure VMs
$blobs | Where-Object {$_.BlobType -eq 'PageBlob' -and $_.Name.EndsWith('.vhd')} | ForEach-Object {
#If a Page blob is not attached as disk then LeaseStatus will be unlocked
if($_.ICloudBlob.Properties.LeaseStatus -eq 'Unlocked'){
if($deleteUnattachedVHDs -eq 1){
Write-Host "Deleting unattached VHD with Uri: $($_.ICloudBlob.Uri.AbsoluteUri)"
$_ | Remove-AzureStorageBlob -Force
Write-Host "Deleted unattached VHD with Uri: $($_.ICloudBlob.Uri.AbsoluteUri)"

}
else{
$_.ICloudBlob.Uri.AbsoluteUri
Next steps
For more information, see Delete storage account and Identify Orphaned Disks Using PowerShell
Use Azure file shares with Windows VMs
You can use Azure file shares as a way to store and access files from your VM. For example, you can store a script
or an application configuration file that you want all your VMs to share. In this article, we show you how to create
and mount an Azure file share, and how to upload and download files.
Connect to a file share from a VM

This section assumes you already have a file share that you want to connect to. If you need to create one, see
Create a file share later in this article.
2. On the left menu, click Storage accounts.
3. Choose your storage account.
4. In the Overview page, under Services, select Files.
5. Select a file share or click + File share to create a new file share to use.
6. Click Connect to open a page that shows the command-line syntax for mounting the file share from Windows
or Linux.
7. In Drive letter, select the letter that you would like to use to identify the drive.
8. Choose which syntax to use and select the copy button on the right to copy it to your clipboard. Paste it some
place where you can easily access it.
9. Connect to your VM and open a command prompt.
10. Paste in the edited connection syntax and hit Enter.
11. When the connection has been created, you get the message The command completed successfully.
12. Check the connection by typing in the drive letter to switch to that drive and then type dir to see the contents of
the file share.
Create a file share

5. In the File Service page, click + File share.
6. Fill in the file share name. File share names can use lowercase letters, numbers and single hyphens. The name
cannot start with a hyphen and you can't use multiple, consecutive hyphens.
7. Fill in a limit on how large the file can be, up to 5120 GB.
8. Click OK to create the file share.
Upload files
5. Select a file share.
6. Click Upload to open the Upload files page.
7. Click on the folder icon to browse your local file system for a file to upload.
8. Click Upload to upload the file to the file share.
Download files
5. Select a file share.
6. Right-click on the file and choose Download to download it to your local machine.
Next steps
You can also create and manage file shares using PowerShell. For more information, see Get started with Azure
File storage on Windows.
Using Managed Disks in Azure Resource Manager
Templates
This document walks through the differences between managed and unmanaged disks when using Azure Resource
Manager templates to provision virtual machines. The examples help you to update existing templates that are
using unmanaged Disks to managed disks. For reference, we are using the 101-vm-simple-windows template as a
guide. You can see the template using both managed Disks and a prior version using unmanaged disks if you'd like
to directly compare them.
Unmanaged Disks template formatting

To begin, let's take a look at how unmanaged disks are deployed. When creating unmanaged disks, you need a
storage account to hold the VHD files. You can create a new storage account or use one that already exists. This
article shows you how to create a new storage account. Create a storage account resource in the resources block as
shown below.
{
"name": "[variables('storageAccountName')]",
"apiVersion": "2016-01-01",
"sku": {
},
"kind": "Storage",
"properties": {}
}
Within the virtual machine object, add a dependency on the storage account to ensure that it's created before the
virtual machine. Within the storageProfile section, specify the full URI of the VHD location, which references the
storage account and is needed for the OS disk and any data disks.
{
"apiVersion": "2015-06-15",
"name": "[variables('vmName')]",
"dependsOn": [
"[resourceId('Microsoft.Storage/storageAccounts/', variables('storageAccountName'))]",
"[resourceId('Microsoft.Network/networkInterfaces/', variables('nicName'))]"
],
"properties": {
"hardwareProfile": {...},
"osProfile": {...},
"storageProfile": {
"imageReference": {
"sku": "[parameters('windowsOSVersion')]",
"version": "latest"
},
"osDisk": {
"name": "osdisk",
"vhd": {
"uri": "[concat(reference(resourceId('Microsoft.Storage/storageAccounts/',
variables('storageAccountName'))).primaryEndpoints.blob, 'vhds/osdisk.vhd')]"
},
},
"dataDisks": [
{
"name": "datadisk1",
"diskSizeGB": 1023,
"lun": 0,
"vhd": {
"uri": "[concat(reference(resourceId('Microsoft.Storage/storageAccounts/',
variables('storageAccountName'))).primaryEndpoints.blob, 'vhds/datadisk1.vhd')]"
},
"createOption": "Empty"
}
]
},
"networkProfile": {...},
"diagnosticsProfile": {...}
}
}
Managed disks template formatting

With Azure Managed Disks, the disk becomes a top-level resource and no longer requires a storage account to be
created by the user. Managed disks were first exposed in the 2016-04-30-preview API version, they are available in
all subsequent API versions and are now the default disk type. The following sections walk through the default
settings and detail how to further customize your disks.
NOTE
It is recommended to use an API version later than 2016-04-30-preview as there were breaking changes between
2016-04-30-preview and 2017-03-30 .
Default managed disk settings

To create a VM with managed disks, you no longer need to create the storage account resource and can update
your virtual machine resource as follows. Specifically note that the apiVersion reflects 2017-03-30 and the
osDisk and dataDisks no longer refer to a specific URI for the VHD. When deploying without specifying
additional properties, the disk will use a Storage type based on the size of the VM. For example, if you are using a
Premium capable VM Size (sizes with "s" in their name such as Standard_D2s_v3) then system will use
Premium_LRS storage. Use the sku setting of the disk to specify a Storage type. If no name is specified, it takes the
format of <VMName>_OsDisk_1_<randomstring> for the OS disk and <VMName>_disk<#>_<randomstring> for each data
disk. By default, Azure disk encryption is disabled; caching is Read/Write for the OS disk and None for data disks.
You may notice in the example below there is still a storage account dependency, though this is only for storage of
diagnostics and is not needed for disk storage.
{
"apiVersion": "2017-03-30",
"dependsOn": [
"[resourceId('Microsoft.Network/networkInterfaces/', variables('nicName'))]"
],
"properties": {
"osProfile": {...},
"storageProfile": {
"imageReference": {
"version": "latest"
},
"osDisk": {
},
"dataDisks": [
{
"diskSizeGB": 1023,
"lun": 0,
}
]
},
}
}
Using a top-level managed disk resource

As an alternative to specifying the disk configuration in the virtual machine object, you can create a top-level disk
resource and attach it as part of the virtual machine creation. For example, you can create a disk resource as
follows to use as a data disk.
{
"type": "Microsoft.Compute/disks",
"name": "[concat(variables('vmName'),'-datadisk1')]",
"apiVersion": "2017-03-30",
"sku": {
},
"properties": {
"creationData": {
},
"diskSizeGB": 1023
}
}
Within the VM object, reference the disk object to be attached. Specifying the resource ID of the managed disk
created in the managedDisk property allows the attachment of the disk as the VM is created. The apiVersion for
the VM resource is set to 2017-03-30 . A dependency on the disk resource is added to ensure it's successfully
created before VM creation.
{
"apiVersion": "2017-03-30",
"dependsOn": [
"[resourceId('Microsoft.Network/networkInterfaces/', variables('nicName'))]",
"[resourceId('Microsoft.Compute/disks/', concat(variables('vmName'),'-datadisk1'))]"
],
"properties": {
"osProfile": {...},
"storageProfile": {
"imageReference": {
"version": "latest"
},
"osDisk": {
},
"dataDisks": [
{
"lun": 0,
"name": "[concat(variables('vmName'),'-datadisk1')]",
"createOption": "attach",
"managedDisk": {
"id": "[resourceId('Microsoft.Compute/disks/', concat(variables('vmName'),'-
datadisk1'))]"
}
}
]
},
}
}
Create managed availability sets with VMs using managed disks

To create managed availability sets with VMs using managed disks, add the sku object to the availability set
resource and set the name property to Aligned . This property ensures that the disks for each VM are sufficiently
isolated from each other to avoid single points of failure. Also note that the apiVersion for the availability set
resource is set to 2017-03-30 .
{
"apiVersion": "2017-03-30",
"type": "Microsoft.Compute/availabilitySets",
"name": "[variables('avSetName')]",
"properties": {
"PlatformUpdateDomainCount": 3,
"PlatformFaultDomainCount": 2
},
"sku": {
"name": "Aligned"
}
}
Standard SSD disks

Below are the parameters needed in the Resource Manager template to create Standard SSD Disks:
apiVersion for Microsoft.Compute must be set as 2018-04-01 (or later)
Specify managedDisk.storageAccountType as StandardSSD_LRS
The following example shows the properties.storageProfile.osDisk section for a VM that uses Standard SSD Disks:
"osDisk": {
"name": "myOsDisk",
"managedDisk": {
"storageAccountType": "StandardSSD_LRS"
}
}
For a complete template example of how to create a Standard SSD disk with a template, see Create a VM from a
Windows Image with Standard SSD Data Disks.
Additional scenarios and customizations
To find full information on the REST API specifications, please review the create a managed disk REST API
documentation. You will find additional scenarios, as well as default and acceptable values that can be submitted to
the API through template deployments.
Next steps
For full templates that use managed disks visit the following Azure Quickstart Repo links.
Windows VM with managed disk
Linux VM with managed disk
Full list of managed disk templates
Visit the Azure Managed Disks Overview document to learn more about managed disks.
Review the template reference documentation for virtual machine resources by visiting the
Microsoft.Compute/virtualMachines template reference document.
Review the template reference documentation for disk resources by visiting the Microsoft.Compute/disks
template reference document.
For information on how to use managed disks in Azure virtual machine scale sets, visit the Use data disks with
scale sets document.
Enable Write Accelerator
Write Accelerator is a disk capability for M -Series Virtual Machines (VMs) on Premium Storage with Azure
Managed Disks exclusively. As the name states, the purpose of the functionality is to improve the I/O latency of
writes against Azure Premium Storage. Write Accelerator is ideally suited where log file updates are required to
persist to disk in a highly performant manner for modern databases.
Write Accelerator is generally available for M -series VMs in the Public Cloud.
Planning for using Write Accelerator

Write Accelerator should be used for the volumes that contain the transaction log or redo logs of a DBMS. It is not
recommended to use Write Accelerator for the data volumes of a DBMS as the feature has been optimized to be
used against log disks.
Write Accelerator only works in conjunction with Azure managed disks.
IMPORTANT
Enabling Write Accelerator for the operating system disk of the VM will reboot the VM.
To enable Write Accelerator to an existing Azure disk that is NOT part of a volume build out of multiple disks with Windows
disk or volume managers, Windows Storage Spaces, Windows Scale-out file server (SOFS), Linux LVM, or MDADM, the
workload accessing the Azure disk needs to be shut down. Database applications using the Azure disk MUST be shut down.
If you want to enable or disable Write Accelerator for an existing volume that is built out of multiple Azure Premium Storage
disks and striped using Windows disk or volume managers, Windows Storage Spaces, Windows Scale-out file server (SOFS),
Linux LVM or MDADM, all disks building the volume must be enabled or disabled for Write Accelerator in separate steps.
Before enabling or disabling Write Accelerator in such a configuration, shut down the Azure VM.
Enabling Write Accelerator for OS disks should not be necessary for SAP -related VM configurations.
Restrictions when using Write Accelerator
When using Write Accelerator for an Azure disk/VHD, these restrictions apply:
The Premium disk caching must be set to 'None' or 'Read Only'. All other caching modes are not supported.
Snapshots on a Write Accelerator enabled disk is not supported yet. This restriction blocks Azure Backup
Service ability to perform an application consistent snapshot of all disks of the virtual machine.
Only smaller I/O sizes (<=32 KiB ) are taking the accelerated path. In workload situations where data is getting
bulk loaded or where the transaction log buffers of the different DBMS are filled to a larger degree before
getting persisted to the storage, chances are that the I/O written to disk is not taking the accelerated path.
There are limits of Azure Premium Storage VHDs per VM that can be supported by Write Accelerator. The current
limits are:
VM SKU NUMBER OF WRITE ACCELERATOR DISKS WRITE ACCELERATOR DISK IOPS PER VM
M128ms, 128s 16 8000
M64ms, M64ls, M64s 8 4000

VM SKU NUMBER OF WRITE ACCELERATOR DISKS WRITE ACCELERATOR DISK IOPS PER VM
M32ms, M32ls, M32ts, M32s 4 2000
M16ms, M16s 2 1000
M8ms, M8s 1 500
The IOPS limits are per VM and not per disk. All Write Accelerator disks share the same IOPS limit per VM.
Enabling Write Accelerator on a specific disk

The next few sections will describe how Write Accelerator can be enabled on Azure Premium Storage VHDs.
Prerequisites
The following prerequisites apply to the usage of Write Accelerator at this point in time:
The disks you want to apply Azure Write Accelerator against need to be Azure managed disks on Premium
Storage.
You must be using an M -series VM
Enabling Azure Write Accelerator using Azure PowerShell

The Azure Power Shell module from version 5.5.0 include the changes to the relevant cmdlets to enable or disable
Write Accelerator for specific Azure Premium Storage disks. In order to enable or deploy disks supported by Write
Accelerator, the following Power Shell commands got changed, and extended to accept a parameter for Write
Accelerator.
A new switch parameter, -WriteAccelerator has been added to the following cmdlets:
Set-AzureRmVMOsDisk
Add-AzureRmVMDataDisk
Set-AzureRmVMDataDisk
Add-AzureRmVmssDataDisk
Not giving the parameter sets the property to false and will deploy disks that have no support by Write
Accelerator.
A new switch parameter, -OsDiskWriteAccelerator was added to the following cmdlets:
Set-AzureRmVmssStorageProfile
Not specifying the parameter sets the property to false by default, returning disks that don't leverage Write
Accelerator.
A new optional Boolean (non-nullable) parameter, -OsDiskWriteAccelerator was added to the following cmdlets:
Update-AzureRmVM
Update-AzureRmVmss
Specify either $true or $false to control support of Azure Write Accelerator with the disks.
Examples of commands could look like:
New-AzureRmVMConfig | Set-AzureRmVMOsDisk | Add-AzureRmVMDataDisk -Name "datadisk1" | Add-AzureRmVMDataDisk -
Name "logdisk1" -WriteAccelerator | New-AzureRmVM
Get-AzureRmVM | Update-AzureRmVM -OsDiskWriteAccelerator $true
New-AzureRmVmssConfig | Set-AzureRmVmssStorageProfile -OsDiskWriteAccelerator | Add-AzureRmVmssDataDisk -Name

"datadisk1" -WriteAccelerator:$false | Add-AzureRmVmssDataDisk -Name "logdisk1" -WriteAccelerator | New-
AzureRmVmss
Get-AzureRmVmss | Update-AzureRmVmss -OsDiskWriteAccelerator:$false
Two main scenarios can be scripted as shown in the following sections.

Adding a new disk supported by Write Accelerator using PowerShell
You can use this script to add a new disk to your VM. The disk created with this script uses Write Accelerator.
Replace myVM , myWAVMs , log001 , size of the disk, and LunID of the disk with values appropriate for your specific
deployment.
# Specify your VM Name

$vmName="myVM"
#Specify your Resource Group
$rgName = "myWAVMs"
#data disk name
$datadiskname = "log001"
#LUN Id
$lunid=8
#size
$size=1023
#Pulls the VM info for later
$vm=Get-AzurermVM -ResourceGroupName $rgname -Name $vmname
#add a new VM data disk
Add-AzureRmVMDataDisk -CreateOption empty -DiskSizeInGB $size -Name $vmname-$datadiskname -VM $vm -Caching None
-WriteAccelerator:$true -lun $lunid
#Updates the VM with the disk config - does not require a reboot
Update-AzureRmVM -ResourceGroupName $rgname -VM $vm
Enabling Write Accelerator on an existing Azure disk using PowerShell

You can use this script to enable Write Accelerator on an existing disk. Replace myVM , myWAVMs , and test-log001
with values appropriate for your specific deployment. The script adds Write Accelerator to an existing disk where
the value for $newstatus is set to '$true'. Using the value '$false' will disable Write Accelerator on a given disk.
#Specify your VM Name

$vmName="myVM"
#Specify your Resource Group
$rgName = "myWAVMs"
#data disk name
$datadiskname = "test-log001"
#new Write Accelerator status ($true for enabled, $false for disabled)
$newstatus = $true
#Pulls the VM info for later
$vm=Get-AzurermVM -ResourceGroupName $rgname -Name $vmname
#add a new VM data disk
Set-AzureRmVMDataDisk -VM $vm -Name $datadiskname -Caching None -WriteAccelerator:$newstatus
#Updates the VM with the disk config - does not require a reboot
Update-AzureRmVM -ResourceGroupName $rgname -VM $vm
NOTE
Executing the script above will detach the disk specified, enable Write Accelerator against the disk, and then attach the disk
again
Enabling Write Accelerator using the Azure portal

You can enable Write Accelerator via the portal where you specify your disk caching settings:
Enabling Write Accelerator using the Azure CLI

You can use the Azure CLI to enable Write Accelerator.
To enable Write Accelerator on an existing disk, use az vm update, you may use the following examples if you
replace the diskName, VMName, and ResourceGroup with your own values:
az vm update -g group1 -n vm1 -write-accelerator 1=true
To attach a disk with Write Accelerator enabled use az vm disk attach, you may use the following example if you
substitute in your own values: az vm disk attach -g group1 -vm-name vm1 -disk d1 --enable-write-accelerator
To disable Write Accelerator, use az vm update, setting the properties to false:
az vm update -g group1 -n vm1 -write-accelerator 0=false 1=false
Enabling Write Accelerator using Rest APIs

To deploy through Azure Rest API, you need to install the Azure armclient.
Install armclient
To run armclient, you need to install it through Chocolatey. You can install it through cmd.exe or powershell. Use
elevated rights for these commands (“Run as Administrator”).
Using cmd.exe, run the following command:
@"%SystemRoot%\System32\WindowsPowerShell\v1.0\powershell.exe" -NoProfile -InputFormat None -ExecutionPolicy
Bypass -Command "iex ((New-Object System.Net.WebClient).DownloadString('https://chocolatey.org/install.ps1'))"
&& SET "PATH=%PATH%;%ALLUSERSPROFILE%\chocolatey\bin"
Using Power Shell, run the following command:

Set-ExecutionPolicy Bypass -Scope Process -Force; iex ((New-Object
System.Net.WebClient).DownloadString('https://chocolatey.org/install.ps1'))
Now you can install the armclient by using the following command in either cmd.exe or PowerShell
choco install armclient
Getting your current VM configuration

To change the attributes of your disk configuration, you first need to get the current configuration in a JSON file.
You can get the current configuration by executing the following command:
armclient GET /subscriptions/<<subscription-
ID<</resourceGroups/<<ResourceGroup>>/providers/Microsoft.Compute/virtualMachines/<<virtualmachinename>>?api-
version=2017-12-01 > <<filename.json>>
Replace the terms within '<< >>' with your data, including the file name the JSON file should have.
The output could look like:
{
"properties": {
"vmId": "2444c93e-f8bb-4a20-af2d-1658d9dbbbcb",
"vmSize": "Standard_M64s"
},
"storageProfile": {
"imageReference": {
"publisher": "SUSE",
"offer": "SLES-SAP",
"sku": "12-SP3",
"version": "latest"
},
"osDisk": {
"osType": "Linux",
"name": "mylittlesap_OsDisk_1_754a1b8bb390468e9b4c429b81cc5f5a",
"managedDisk": {
"storageAccountType": "Premium_LRS",
"id":
"/subscriptions/XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX/resourceGroups/mylittlesap/providers/Microsoft.Compute/dis
ks/mylittlesap_OsDisk_1_754a1b8bb390468e9b4c429b81cc5f5a"
},
"diskSizeGB": 30
},
"dataDisks": [
{
"lun": 0,
"name": "data1",
"createOption": "Attach",
"caching": "None",
"managedDisk": {
"id":
ks/data1"
},
"diskSizeGB": 1023
},
{
"lun": 1,
"name": "log1",
"caching": "None",
"managedDisk": {
"id":
ks/data2"
ks/data2"
},
"diskSizeGB": 1023
}
]
},
"osProfile": {
"computerName": "mylittlesapVM",
"adminUsername": "pl",
"linuxConfiguration": {
"disablePasswordAuthentication": false
},
"secrets": []
},
"networkProfile": {
{
"id":
"/subscriptions/XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX/resourceGroups/mylittlesap/providers/Microsoft.Network/net
workInterfaces/mylittlesap518"
}
]
},
"diagnosticsProfile": {
"bootDiagnostics": {
"enabled": true,
"storageUri": "https://mylittlesapdiag895.blob.core.windows.net/"
}
},
"provisioningState": "Succeeded"
},
"location": "westeurope",
"id":
"/subscriptions/XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX/resourceGroups/mylittlesap/providers/Microsoft.Compute/vir
tualMachines/mylittlesapVM",
"name": "mylittlesapVM"
Next, update the JSON file and to enable Write Accelerator on the disk called 'log1'. This can be accomplished by
adding this attribute into the JSON file after the cache entry of the disk.
{
"lun": 1,
"name": "log1",
"caching": "None",
"writeAcceleratorEnabled": true,
"managedDisk": {
"id":
ks/data2"
},
"diskSizeGB": 1023
}
Then update the existing deployment with this command:

armclient PUT /subscriptions/<<subscription-
ID<</resourceGroups/<<ResourceGroup>>/providers/Microsoft.Compute/virtualMachines/<<virtualmachinename>>?api-
version=2017-12-01 @<<filename.json>>
The output should look like the one below. You can see that Write Accelerator enabled for one disk.
{
"properties": {
"vmSize": "Standard_M64s"
},
"storageProfile": {
"imageReference": {
"publisher": "SUSE",
"offer": "SLES-SAP",
"sku": "12-SP3",
"version": "latest"
},
"osDisk": {
"osType": "Linux",
"name": "mylittlesap_OsDisk_1_754a1b8bb390468e9b4c429b81cc5f5a",
"managedDisk": {
"id":
ks/mylittlesap_OsDisk_1_754a1b8bb390468e9b4c429b81cc5f5a"
},
"diskSizeGB": 30
},
"dataDisks": [
{
"lun": 0,
"name": "data1",
"caching": "None",
"managedDisk": {
"id":
ks/data1"
},
"diskSizeGB": 1023
},
{
"lun": 1,
"name": "log1",
"caching": "None",
"writeAcceleratorEnabled": true,
"managedDisk": {
"id":
ks/data2"
},
"diskSizeGB": 1023
}
]
},
"osProfile": {
"computerName": "mylittlesapVM",
"adminUsername": "pl",
"linuxConfiguration": {
"disablePasswordAuthentication": false
},
"secrets": []
},
"networkProfile": {
{
"id":
"/subscriptions/XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX/resourceGroups/mylittlesap/providers/Microsoft.Network/net
workInterfaces/mylittlesap518"
}
]
]
},
"diagnosticsProfile": {
"bootDiagnostics": {
"enabled": true,
"storageUri": "https://mylittlesapdiag895.blob.core.windows.net/"
}
},
"provisioningState": "Succeeded"
},
"location": "westeurope",
"id":
"/subscriptions/XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX/resourceGroups/mylittlesap/providers/Microsoft.Compute/vir
tualMachines/mylittlesapVM",
"name": "mylittlesapVM"
Once you've made this change, the drive should be supported by Write Accelerator.
Enabling Azure Ultra SSDs
Azure Ultra SSD delivers high throughput, high IOPS, and consistent low latency disk storage for Azure IaaS VMs.
This new offering provides top of the line performance at the same availability levels as our existing disks offerings.
Additional benefits of Ultra SSD include the ability to dynamically change the performance of the disk along with
your workloads without the need to restart your virtual machines. Ultra SSD is suited for data-intensive workloads
such as SAP HANA, top tier databases, and transaction-heavy workloads.
Currently, Ultra SSds are in preview and you must enroll in the preview in order to access them.
Once approved, run one of the following commands to determine which zone in East US 2 to deploy your Ultra
SSD to:
PowerShell:
Get-AzureRmComputeResourceSku | where {$_.ResourceType -eq "disks" -and $_.Name -eq "UltraSSD_LRS" }
CLI: az vm list-skus --resource-type disks --query “[?name==’UltraSSD_LRS’]”
The response will be similar to the form below, where X is the Zone to use for deploying in East US 2. X could be
either 1, 2, or 3.
RESOURCETYPE NAME LOCATION ZONES RESTRICTION CAPABILITY VALUE
disks UltraSSD_LRS eastus2 X
If there was no response from the command, that means your registration to the feature is either still pending, or
not approved yet.
Now that you know which zone to deploy to, follow the deployment steps in this article to get your first VMs
deployed with Ultra SSD disks.
Deploying an Ultra SSD

First, determine the VM Size to deploy. As part of this preview, only DsV3 and EsV3 VM families are supported.
Refer to the second table on this blog for additional details about these VM sizes. Also refer to the sample Create a
VM with multiple Ultra SSD disks, which shows how to create a VM with multiple Ultra SSD disks.
The following describe the new/modified Resource Manager template changes: apiVersion for
Microsoft.Compute/virtualMachines and Microsoft.Compute/Disks must be set as 2018-06-01 (or later ).
Specify Disk Sku UltraSSD_LRS, disk capacity, IOPS, and throughput in MBps to create an Ultra SSD disk. The
following is an example that creates a disk with 1,024 GiB (GiB = 2^30 Bytes), 80,000 IOPS, and 1,200 MBps
(MBps = 10^6 Bytes per second):
"properties": {
"creationData": {
},
"diskSizeGB": 1024,
"diskIOPSReadWrite": 80000,
"diskMBpsReadWrite": 1200,
}
Add an additional capability on the properties of the VM to indicate its Ultra SSD Enabled (refer to the sample for
the full Resource Manager template):
{
"apiVersion": "2018-06-01",
"properties": {
"hardwareProfile": {},
"additionalCapabilities" : {
"ultraSSDEnabled" : "true"
},
"osProfile": {},
"storageProfile": {},
"networkProfile": {}
}
}
Once the VM is provisioned, you can partition and format the data disks and configure them for your workloads.
Additional Ultra SSD scenarios

During VM Creation, Ultra SSDs can be implicitly created as well. However, these disks will receive a default
value for IOPS (500) and throughput (8 MiB/s).
Additional Ultra SSDs can be attached to Ultra SSD compatible VMs.
Ultra SSD supports adjusting the disk performance attributes (IOPS and throughput) at runtime without
detaching the disk from the virtual machine. ONce a disk performance resize operation has been issued on a
disk, it can take up to an hour for the change to actually take effect.
Growing the disk capacity does require a virtual machine to be de-allocated.
Next steps
If you would like to try the new disk type and haven't signed up for the preview yet, request access through this
survey.
Migrate Azure VMs to Managed Disks in Azure
Azure Managed Disks simplifies your storage management by removing the need to separately manage storage
accounts. You can also migrate your existing Azure VMs to Managed Disks to benefit from better reliability of VMs
in an Availability Set. It ensures that the disks of different VMs in an Availability Set is sufficiently isolated from
each other to avoid single point of failures. It automatically places disks of different VMs in an Availability Set in
different Storage scale units (stamps) which limits the impact of single Storage scale unit failures caused due to
hardware and software failures. Based on your needs, you can choose from two types of storage options:
Premium Managed Disks are Solid State Drive (SSD ) based storage media which delivers high
performance, low -latency disk support for virtual machines running I/O -intensive workloads. You can take
advantage of the speed and performance of these disks by migrating to Premium Managed Disks.
Standard Managed Disks use Hard Disk Drive (HDD ) based storage media and are best suited for Dev/Test
and other infrequent access workloads that are less sensitive to performance variability.
You can migrate to Managed Disks in following scenarios:
MIGRATE... DOCUMENTATION LINK
Convert stand alone VMs and VMs in an availability set to Convert VMs to use managed disks
managed disks
A single VM from classic to Resource Manager on managed Migrate a single VM

disks
All the VMs in a vNet from classic to Resource Manager on Migrate IaaS resources from classic to Resource Manager and
managed disks then Convert a VM from unmanaged disks to managed disks
Plan for the conversion to Managed Disks

This section helps you to make the best decision on VM and disk types.
Location
Pick a location where Azure Managed Disks are available. If you are moving to Premium Managed Disks, also
ensure that Premium storage is available in the region where you are planning to move to. See Azure Services by
Region for up-to-date information on available locations.
VM sizes
If you are migrating to Premium Managed Disks, you have to update the size of the VM to Premium Storage
capable size available in the region where VM is located. Review the VM sizes that are Premium Storage capable.
The Azure VM size specifications are listed in Sizes for virtual machines. Review the performance characteristics of
virtual machines that work with Premium Storage and choose the most appropriate VM size that best suits your
workload. Make sure that there is sufficient bandwidth available on your VM to drive the disk traffic.
Disk sizes
Premium Managed Disks
There are seven types of premium managed disks that can be used with your VM and each has specific IOPs and
throughput limits. Take into consideration these limits when choosing the Premium disk type for your VM based
on the needs of your application in terms of capacity, performance, scalability, and peak loads.
PREMIUM
DISKS
TYPE P4 P6 P10 P15 P20 P30 P40 P50
Disk size 32 GB 64 GB 128 GB 256 GB 512 GB 1024 GB 2048 GB 4095 GB

(1 TB) (2 TB) (4 TB)
IOPS per 120 240 500 1100 2300 5000 7500 7500
disk
Throughp 25 MB 50 MB 100 MB 125 MB 150 MB 200 MB 250 MB 250 MB

ut per per per per per per per per per
disk second second second second second second second second
Standard Managed Disks

There are seven types of standard managed disks that can be used with your VM. Each of them have different
capacity but have same IOPS and throughput limits. Choose the type of Standard Managed disks based on the
capacity needs of your application.
STANDAR
D DISK
TYPE S4 S6 S10 S15 S20 S30 S40 S50
Disk size 30 GB 64 GB 128 GB 256 GB 512 GB 1024 GB 2048 GB 4095 GB

(1 TB) (2TB) (4 TB)
IOPS per 500 500 500 500 500 500 500 500
disk
Throughp 60 MB 60 MB 60 MB 60 MB 60 MB 60 MB 60 MB 60 MB
ut per per per per per per per per per
disk second second second second second second second second
Disk caching policy

By default, disk caching policy is Read -Only for all the Premium data disks, and Read -Write for the Premium
operating system disk attached to the VM. This configuration setting is recommended to achieve the optimal
performance for your application’s IOs. For write-heavy or write-only data disks (such as SQL Server log files),
disable disk caching so that you can achieve better application performance.
Pricing
Review the pricing for Managed Disks. Pricing of Premium Managed Disks is same as the Premium Unmanaged
Disks. But pricing for Standard Managed Disks is different than Standard Unmanaged Disks.
Next steps
Learn more about Managed Disks
Convert a Windows virtual machine from
unmanaged disks to managed disks
If you have existing Windows virtual machines (VMs) that use unmanaged disks, you can convert the VMs to use
managed disks through the Azure Managed Disks service. This process converts both the OS disk and any
attached data disks.
This article shows you how to convert VMs by using Azure PowerShell. If you need to install or upgrade it, see
Install and configure Azure PowerShell.
Before you begin

Review Plan for the migration to Managed Disks.
Review the FAQ about migration to Managed Disks.
The conversion requires a restart of the VM, so schedule the migration of your VMs during a pre-existing
maintenance window.
The conversion is not reversible.
Be aware that any users with the Virtual Machine Contributor role will not be able to change the VM size
(as they could pre-conversion). This is because VMs with managed disks require the user to have the
Microsoft.Compute/disks/write permission on the OS disks.
Be sure to test the conversion. Migrate a test virtual machine before you perform the migration in
production.
During the conversion, you deallocate the VM. The VM receives a new IP address when it is started after
the conversion. If needed, you can assign a static IP address to the VM.
The original VHDs and the storage account used by the VM before conversion are not deleted. They
continue to incur charges. To avoid being billed for these artifacts, delete the original VHD blobs after you
verify that the conversion is complete.
Review the minimum version of the Azure VM agent required to support the conversion process. For
information on how to check and update your agent version, see Minimum version support for VM agents
in Azure
Convert single-instance VMs

This section covers how to convert single-instance Azure VMs from unmanaged disks to managed disks. (If your
VMs are in an availability set, see the next section.)
1. Deallocate the VM by using the Stop-AzureRmVM cmdlet. The following example deallocates the VM
named myVM in the resource group named myResourceGroup :
$vmName = "myVM"
2. Convert the VM to managed disks by using the ConvertTo-AzureRmVMManagedDisk cmdlet. The
following process converts the previous VM, including the OS disk and any data disks, and starts the
Virtual Machine:
ConvertTo-AzureRmVMManagedDisk -ResourceGroupName $rgName -VMName $vmName
Convert VMs in an availability set

If the VMs that you want to convert to managed disks are in an availability set, you first need to convert the
availability set to a managed availability set.
1. Convert the availability set by using the Update-AzureRmAvailabilitySet cmdlet. The following example
updates the availability set named myAvailabilitySet in the resource group named myResourceGroup :
$avSetName = 'myAvailabilitySet'
$avSet = Get-AzureRmAvailabilitySet -ResourceGroupName $rgName -Name $avSetName

Update-AzureRmAvailabilitySet -AvailabilitySet $avSet -Sku Aligned
If the region where your availability set is located has only 2 managed fault domains but the number of
unmanaged fault domains is 3, this command shows an error similar to "The specified fault domain count
3 must fall in the range 1 to 2." To resolve the error, update the fault domain to 2 and update Sku to
Aligned as follows:
$avSet.PlatformFaultDomainCount = 2
Update-AzureRmAvailabilitySet -AvailabilitySet $avSet -Sku Aligned
2. Deallocate and convert the VMs in the availability set. The following script deallocates each VM by using
the Stop-AzureRmVM cmdlet, converts it by using ConvertTo-AzureRmVMManagedDisk, and restarts it
automatically as apart of the conversion process:
$avSet = Get-AzureRmAvailabilitySet -ResourceGroupName $rgName -Name $avSetName
foreach($vmInfo in $avSet.VirtualMachinesReferences)
{
$vm = Get-AzureRmVM -ResourceGroupName $rgName | Where-Object {$_.Id -eq $vmInfo.id}
Stop-AzureRmVM -ResourceGroupName $rgName -Name $vm.Name -Force
ConvertTo-AzureRmVMManagedDisk -ResourceGroupName $rgName -VMName $vm.Name
}
Troubleshooting
If there is an error during conversion, or if a VM is in a failed state because of issues in a previous conversion, run
the ConvertTo-AzureRmVMManagedDisk cmdlet again. A simple retry usually unblocks the situation. Before
converting, make sure all the VM extensions are in the 'Provisioning succeeded' state or the conversion will fail
with the error code 409.
Convert using the Azure portal

You can also convert unmanaged disks to managed disks using the Azure portal.
2. Select the VM from the list of VMs in the portal.
3. In the blade for the VM, select Disks from the menu.
4. At the top of the Disks blade, select Migrate to managed disks.
5. If your VM is in an availability set, there will be a warning on the Migrate to managed disks blade that you
need to convert the availability set first. The warning should have a link you can click to convert the availability
set. Once the availability set is converted or if your VM is not in an availability set, click Migrate to start the
process of migrating your disks to managed disks.
The VM will be stopped and restarted after migration is complete.
Next steps
Convert standard managed disks to premium
Take a read-only copy of a VM by using snapshots.
Manually migrate a Classic VM to a new ARM
Managed Disk VM from the VHD
This section helps you to migrate your existing Azure VMs from the classic deployment model to Managed Disks
in the Resource Manager deployment model.
Plan for the migration to Managed Disks

Location
Pick a location where Azure Managed Disks are available. If you are migrating to Premium Managed Disks, also
ensure that Premium storage is available in the region where you are planning to migrate to. See Azure Services
byRegion for up-to-date information on available locations.
VM sizes
Disk sizes
There are seven types of premium Managed disks that can be used with your VM and each has specific IOPs and
throughput limits. Consider these limits when choosing the Premium disk type for your VM based on the needs of
your application in terms of capacity, performance, scalability, and peak loads.
PREMIUM
DISKS TYPE P4 P6 P10 P20 P30 P40 P50
Disk size 128 GB 512 GB 128 GB 512 GB 1024 GB (1 2048 GB (2 4095 GB (4

TB) TB) TB)
IOPS per 120 240 500 2300 5000 7500 7500

disk
Throughput 25 MB per 50 MB per 100 MB per 150 MB per 200 MB per 250 MB per 250 MB per
per disk second second second second second second second

There are seven types of Standard Managed disks that can be used with your VM. Each of them have different
STANDARD
DISK TYPE S4 S6 S10 S20 S30 S40 S50
Disk size 30 GB 64 GB 128 GB 512 GB 1024 GB (1 2048 GB 4095 GB (4

TB) (2TB) TB)
IOPS per 500 500 500 500 500 500 500

disk
Throughput 60 MB per 60 MB per 60 MB per 60 MB per 60 MB per 60 MB per 60 MB per

per disk second second second second second second second
Disk caching policy

Pricing
Checklist
1. If you are migrating to Premium Managed Disks, make sure it is available in the region you are migrating
to.
2. Decide the new VM series you will be using. It should be a Premium Storage capable if you are migrating to
Premium Managed Disks.
3. Decide the exact VM size you will use which are available in the region you are migrating to. VM size needs
to be large enough to support the number of data disks you have. For example, if you have four data disks,
the VM must have two or more cores. Also, consider processing power, memory and network bandwidth
needs.
4. Have the current VM details handy, including the list of disks and corresponding VHD blobs.
Prepare your application for downtime. To do a clean migration, you have to stop all the processing in the current
system. Only then you can get it to consistent state which you can migrate to the new platform. Downtime
duration depends on the amount of data in the disks to migrate.
Migrate the VM
Prepare your application for downtime. To do a clean migration, you have to stop all the processing in the current
system. Only then you can get it to consistent state which you can migrate to the new platform. Downtime
duration depends the amount of data in the disks to migrate.
This part requires the Azure PowerShell module version 6.0.0 or later. Run Get-Module -ListAvailable AzureRM to
find the version. If you need to upgrade, see Install Azure PowerShell module. You also need to run
Connect-AzureRmAccount to create a connection with Azure.
1. First, set the common parameters:

$resourceGroupName = 'yourResourceGroupName'
$location = 'your location'
$virtualNetworkName = 'yourExistingVirtualNetworkName'
$virtualMachineName = 'yourVMName'
$virtualMachineSize = 'Standard_DS3'
$adminUserName = "youradminusername"
$adminPassword = "yourpassword" | ConvertTo-SecureString -AsPlainText -Force
$imageName = 'yourImageName'
$osVhdUri = 'https://storageaccount.blob.core.windows.net/vhdcontainer/osdisk.vhd'
$dataVhdUri = 'https://storageaccount.blob.core.windows.net/vhdcontainer/datadisk1.vhd'
$dataDiskName = 'dataDisk1'
2. Create a managed OS disk using the VHD from the classic VM.
Ensure that you have provided the complete URI of the OS VHD to the $osVhdUri parameter. Also, enter -
AccountType as Premium_LRS or Standard_LRS based on type of disks (Premium or Standard) you are
migrating to.
$osDisk = New-AzureRmDisk -DiskName $osDiskName -Disk (New-AzureRmDiskConfig '

-AccountType Premium_LRS -Location $location -CreateOption Import -SourceUri $osVhdUri) '
3. Attach the OS disk to the new VM.
$VirtualMachine = New-AzureRmVMConfig -VMName $virtualMachineName -VMSize $virtualMachineSize

$VirtualMachine = Set-AzureRmVMOSDisk -VM $VirtualMachine -ManagedDiskId $osDisk.Id '
-StorageAccountType Premium_LRS -DiskSizeInGB 128 -CreateOption Attach -Windows
4. Create a managed data disk from the data VHD file and add it to the new VM.
$dataDisk1 = New-AzureRmDisk -DiskName $dataDiskName -Disk (New-AzureRmDiskConfig '

-AccountType Premium_LRS -Location $location -CreationDataCreateOption Import '
-SourceUri $dataVhdUri ) -ResourceGroupName $resourceGroupName
$VirtualMachine = Add-AzureRmVMDataDisk -VM $VirtualMachine -Name $dataDiskName '

-CreateOption Attach -ManagedDiskId $dataDisk1.Id -Lun 1
5. Create the new VM by setting public IP, Virtual Network and NIC.
$publicIp = New-AzureRmPublicIpAddress -Name ($VirtualMachineName.ToLower()+'_ip') '
-ResourceGroupName $resourceGroupName -Location $location -AllocationMethod Dynamic
$vnet = Get-AzureRmVirtualNetwork -Name $virtualNetworkName -ResourceGroupName $resourceGroupName
$nic = New-AzureRmNetworkInterface -Name ($VirtualMachineName.ToLower()+'_nic') '

-ResourceGroupName $resourceGroupName -Location $location -SubnetId $vnet.Subnets[0].Id '
-PublicIpAddressId $publicIp.Id
$VirtualMachine = Add-AzureRmVMNetworkInterface -VM $VirtualMachine -Id $nic.Id
New-AzureRmVM -VM $VirtualMachine -ResourceGroupName $resourceGroupName -Location $location
NOTE
There may be additional steps necessary to support your application that is not be covered by this guide.
Next steps
Connect to the virtual machine. For instructions, see How to connect and log on to an Azure virtual machine
running Windows.
Common PowerShell commands for Azure Virtual
Networks
If you want to create a virtual machine, you need to create a virtual network or know about an existing virtual
network in which the VM can be added. Typically, when you create a VM, you also need to consider creating the
resources described in this article.
See How to install and configure Azure PowerShell for information about installing the latest version of Azure
PowerShell, selecting your subscription, and signing in to your account.
Some variables might be useful for you if running more than one of the commands in this article:
$location - The location of the network resources. You can use Get-AzureRmLocation to find a geographical
region that works for you.
$myResourceGroup - The name of the resource group where the network resources are located.
Create network resources

TASK COMMAND
Create subnet configurations $subnet1 = New-AzureRmVirtualNetworkSubnetConfig -

Name "mySubnet1" -AddressPrefix XX.X.X.X/XX
$subnet2 = New-AzureRmVirtualNetworkSubnetConfig -
Name "mySubnet2" -AddressPrefix XX.X.X.X/XX
A typical network might have a subnet for an internet facing

load balancer and a separate subnet for an internal load
balancer.
Create a virtual network $vnet = New-AzureRmVirtualNetwork -Name "myVNet" -

ResourceGroupName $myResourceGroup -Location $location
-AddressPrefix XX.X.X.X/XX -Subnet $subnet1, $subnet2
Test for a unique domain name Test-AzureRmDnsAvailability -DomainNameLabel "myDNS" -

Location $location
You can specify a DNS domain name for a public IP resource,

which creates a mapping for
domainname.location.cloudapp.azure.com to the public IP
address in the Azure-managed DNS servers. The name can
contain only letters, numbers, and hyphens. The first and last
character must be a letter or number and the domain name
must be unique within its Azure location. If True is returned,
your proposed name is globally unique.
Create a public IP address $pip = New-AzureRmPublicIpAddress -Name "myPublicIp" -

ResourceGroupName $myResourceGroup -DomainNameLabel
"myDNS" -Location $location -AllocationMethod Dynamic
The public IP address uses the domain name that you

previously tested and is used by the frontend configuration of
the load balancer.
TASK COMMAND
Create a frontend IP configuration $frontendIP = New-AzureRmLoadBalancerFrontendIpConfig -

Name "myFrontendIP" -PublicIpAddress $pip
The frontend configuration includes the public IP address that

you previously created for incoming network traffic.
Create a backend address pool $beAddressPool = New-

AzureRmLoadBalancerBackendAddressPoolConfig -Name
"myBackendAddressPool"
Provides internal addresses for the backend of the load

balancer that are accessed through a network interface.
Create a probe $healthProbe = New-AzureRmLoadBalancerProbeConfig -

Name "myProbe" -RequestPath 'HealthProbe.aspx' -Protocol
http -Port 80 -IntervalInSeconds 15 -ProbeCount 2
Contains health probes used to check availability of virtual

machines instances in the backend address pool.
Create a load balancing rule $lbRule = New-AzureRmLoadBalancerRuleConfig -Name HTTP

-FrontendIpConfiguration $frontendIP -BackendAddressPool
$beAddressPool -Probe $healthProbe -Protocol Tcp -
FrontendPort 80 -BackendPort 80
Contains rules that assign a public port on the load balancer

to a port in the backend address pool.
Create an inbound NAT rule $inboundNATRule = New-

AzureRmLoadBalancerInboundNatRuleConfig -Name
"myInboundRule1" -FrontendIpConfiguration $frontendIP -
Protocol TCP -FrontendPort 3441 -BackendPort 3389
Contains rules mapping a public port on the load balancer to

a port for a specific virtual machine in the backend address
pool.
Create a load balancer $loadBalancer = New-AzureRmLoadBalancer -

ResourceGroupName $myResourceGroup -Name
"myLoadBalancer" -Location $location -
FrontendIpConfiguration $frontendIP -InboundNatRule
$inboundNATRule -LoadBalancingRule $lbRule -
BackendAddressPool $beAddressPool -Probe $healthProbe
Create a network interface $nic1= New-AzureRmNetworkInterface -ResourceGroupName

$myResourceGroup -Name "myNIC" -Location $location -
PrivateIpAddress XX.X.X.X -Subnet $subnet2 -
LoadBalancerBackendAddressPool
$loadBalancer.BackendAddressPools[0] -
LoadBalancerInboundNatRule
$loadBalancer.InboundNatRules[0]
Create a network interface using the public IP address and

virtual network subnet that you previously created.
Get information about network resources

TASK COMMAND
List virtual networks Get-AzureRmVirtualNetwork -ResourceGroupName

$myResourceGroup
Lists all the virtual networks in the resource group.
Get information about a virtual network Get-AzureRmVirtualNetwork -Name "myVNet" -

ResourceGroupName $myResourceGroup
List subnets in a virtual network Get-AzureRmVirtualNetwork -Name "myVNet" -

ResourceGroupName $myResourceGroup | Select Subnets
Get information about a subnet Get-AzureRmVirtualNetworkSubnetConfig -Name

"mySubnet1" -VirtualNetwork $vnet
Gets information about the subnet in the specified virtual

network. The $vnet value represents the object returned by
Get-AzureRmVirtualNetwork.
List IP addresses Get-AzureRmPublicIpAddress -ResourceGroupName

$myResourceGroup
Lists the public IP addresses in the resource group.
List load balancers Get-AzureRmLoadBalancer -ResourceGroupName

$myResourceGroup
Lists all the load balancers in the resource group.
List network interfaces Get-AzureRmNetworkInterface -ResourceGroupName

$myResourceGroup
Lists all the network interfaces in the resource group.
Get information about a network interface Get-AzureRmNetworkInterface -Name "myNIC" -

Gets information about a specific network interface.
Get the IP configuration of a network interface Get-AzureRmNetworkInterfaceIPConfig -Name "myNICIP" -

NetworkInterface $nic
Gets information about the IP configuration of the specified

network interface. The $nic value represents the object
returned by Get-AzureRmNetworkInterface.
Manage network resources

TASK COMMAND
Add a subnet to a virtual network Add-AzureRmVirtualNetworkSubnetConfig -AddressPrefix

XX.X.X.X/XX -Name "mySubnet1" -VirtualNetwork $vnet
Adds a subnet to an existing virtual network. The $vnet value

represents the object returned by Get-
AzureRmVirtualNetwork.
TASK COMMAND
Delete a virtual network Remove-AzureRmVirtualNetwork -Name "myVNet" -

Removes the specified virtual network from the resource

group.
Delete a network interface Remove-AzureRmNetworkInterface -Name "myNIC" -

Removes the specified network interface from the resource

group.
Delete a load balancer Remove-AzureRmLoadBalancer -Name "myLoadBalancer" -

Removes the specified load balancer from the resource group.
Delete a public IP address Remove-AzureRmPublicIpAddress-Name "myIPAddress" -

Removes the specified public IP address from the resource

group.
Next Steps
Use the network interface that you just created when you create a VM.
Learn about how you can create a VM with multiple network interfaces.
2 minutes to read
How to open ports to a virtual machine with the
Azure portal
You open a port, or create an endpoint, to a virtual machine (VM ) in Azure by creating a network filter on a subnet
or a VM network interface. You place these filters, which control both inbound and outbound traffic, on a network
security group attached to the resource that receives the traffic.
The example in this article demonstrates how to create a network filter that uses the standard TCP port 80 (it's
assumed you've already started the appropriate services and opened any OS firewall rules on the VM ).
After you've created a VM that's configured to serve web requests on the standard TCP port 80, you can:
1. Create a network security group.
2. Create an inbound security rule allowing traffic and assign values to the following settings:
Destination port ranges: 80
Source port ranges: * (allows any source port)
Priority value: Enter a value that is less than 65,500 and higher in priority than the default catch-all
deny inbound rule.
3. Associate the network security group with the VM network interface or subnet.
Although this example uses a simple rule to allow HTTP traffic, you can also use network security groups and rules
to create more complex network configurations.
Sign in to Azure
Create a network security group

1. Search for and select the resource group for the VM, choose Add, then search for and select Network
security group.
2. Select Create.
The Create network security group window opens.
3. Enter a name for your network security group.
4. Select or create a resource group, then select a location.
5. Select Create to create the network security group.
Create an inbound security rule

1. Select your new network security group.
2. Select Inbound security rules, then select Add.
3. Select Advanced.
4. Choose a common Service from the drop-down menu, such as HTTP. You can also select Custom if you
want to provide a specific port to use.
5. Optionally, change the Priority or Name. The priority affects the order in which rules are applied: the lower
the numerical value, the earlier the rule is applied.
6. Select Add to create the rule.
Associate your network security group with a subnet

Your final step is to associate your network security group with a subnet or a specific network interface. For this
example, we'll associate the network security group with a subnet.
1. Select Subnets, then select Associate.
2. Select your virtual network, and then select the appropriate subnet.
Any VMs you connect to that subnet are now reachable on port 80.
Additional information
You can also perform the steps in this article by using Azure PowerShell.
The commands described in this article allow you to quickly get traffic flowing to your VM. Network security
groups provide many great features and granularity for controlling access to your resources. For more
information, see Filter network traffic with a network security group.
For highly available web applications, consider placing your VMs behind an Azure load balancer. The load balancer
distributes traffic to VMs, with a network security group that provides traffic filtering. For more information, see
Load balance Windows virtual machines in Azure to create a highly available application.
Next steps
In this article, you created a network security group, created an inbound rule that allows HTTP traffic on port 80,
and then associated that rule with a subnet.
You can find information on creating more detailed environments in the following articles:
Security groups
How to open ports and endpoints to a VM in Azure
using PowerShell
You open a port, or create an endpoint, to a virtual machine (VM ) in Azure by creating a network filter on a subnet
or a VM network interface. You place these filters, which control both inbound and outbound traffic, on a network
security group attached to the resource that receives the traffic.
The example in this article demonstrates how to create a network filter that uses the standard TCP port 80 (it's
assumed you've already started the appropriate services and opened any OS firewall rules on the VM ).
After you've created a VM that's configured to serve web requests on the standard TCP port 80, you can:
1. Create a network security group.
2. Create an inbound security rule allowing traffic and assign values to the following settings:
Destination port ranges: 80
Source port ranges: * (allows any source port)
Priority value: Enter a value that is less than 65,500 and higher in priority than the default catch-all
deny inbound rule.
3. Associate the network security group with the VM network interface or subnet.
Although this example uses a simple rule to allow HTTP traffic, you can also use network security groups and
rules to create more complex network configurations.
Quick commands
To create a Network Security Group and ACL rules you need the latest version of Azure PowerShell installed. You
can also perform these steps using the Azure portal.
Log in to your Azure account:
In the following examples, replace parameter names with your own values. Example parameter names included
myResourceGroup, myNetworkSecurityGroup, and myVnet.
Create a rule with New -AzureRmNetworkSecurityRuleConfig. The following example creates a rule named
myNetworkSecurityGroupRule to allow tcp traffic on port 80:
$httprule = New-AzureRmNetworkSecurityRuleConfig `
-Name "myNetworkSecurityGroupRule" `
-Description "Allow HTTP" `
-Access "Allow" `
-Protocol "Tcp" `
-Direction "Inbound" `
-Priority "100" `
-SourceAddressPrefix "Internet" `
-DestinationPortRange 80
Next, create your Network Security group with New -AzureRmNetworkSecurityGroup and assign the HTTP rule
you just created as follows. The following example creates a Network Security Group named
myNetworkSecurityGroup:
-Name "myNetworkSecurityGroup" `
-SecurityRules $httprule
Now let's assign your Network Security Group to a subnet. The following example assigns an existing virtual
network named myVnet to the variable $vnet with Get-AzureRmVirtualNetwork:
$vnet = Get-AzureRmVirtualNetwork `
-Name "myVnet"
Associate your Network Security Group with your subnet with Set-AzureRmVirtualNetworkSubnetConfig. The
following example associates the subnet named mySubnet with your Network Security Group:
$subnetPrefix = $vnet.Subnets|?{$_.Name -eq 'mySubnet'}
Set-AzureRmVirtualNetworkSubnetConfig `
-Name "mySubnet" `
-AddressPrefix $subnetPrefix.AddressPrefix `
-NetworkSecurityGroup $nsg
Finally, update your virtual network with Set-AzureRmVirtualNetwork in order for your changes to take effect:
Set-AzureRmVirtualNetwork -VirtualNetwork $vnet
More information on Network Security Groups

The quick commands here allow you to get up and running with traffic flowing to your VM. Network Security
Groups provide many great features and granularity for controlling access to your resources. You can read more
about creating a Network Security Group and ACL rules here.
For highly available web applications, you should place your VMs behind an Azure Load Balancer. The load
balancer distributes traffic to VMs, with a Network Security Group that provides traffic filtering. For more
information, see How to load balance Linux virtual machines in Azure to create a highly available application.
Next steps
In this example, you created a simple rule to allow HTTP traffic. You can find information on creating more
detailed environments in the following articles:
What is a network security group?
Azure Resource Manager Overview for Load Balancers
Create a virtual machine with a static public IP
address using the Azure portal
You can create a virtual machine with a static public IP address. A public IP address enables you to communicate to
a virtual machine from the internet. Assign a static public IP address, rather than a dynamic address, to ensure that
the address never changes. Learn more about static public IP addresses. To change a public IP address assigned to
an existing virtual machine from dynamic to static, or to work with private IP addresses, see Add, change, or
remove IP addresses. Public IP addresses have a nominal charge, and there is a limit to the number of public IP
addresses that you can use per subscription.
Sign in to Azure
Create a virtual machine

1. Select + Create a resource found on the upper, left corner of the Azure portal.
2. Select Compute, and then select Windows Server 2016 VM, or another operating system of your choosing.
3. Enter, or select, the following information, accept the defaults for the remaining settings, and then select OK:
SETTING VALUE
Name myVM
User name Enter a user name of your choosing.
Password Enter a password of your choosing. The password must be

at least 12 characters long and meet the defined
complexity requirements.
Subscription Select your subscription.
Resource group Select Use existing and select myResourceGroup.
Location Select East US
4. Select a size for the VM and then select Select.

5. Under Settings, select Public IP address.
6. Enter myPublicIpAddress, select Static, and then select OK, as shown in the following picture:
If the public IP address must be a standard SKU, select Standard under SKU. Learn more about Public IP
address SKUs. If the virtual machine will be added to the back-end pool of a public Azure Load Balancer, the
SKU of the virtual machine's public IP address must match the SKU of the load balancer's public IP address.
For details, see Azure Load Balancer.
7. Select a port, or no ports under Select public inbound ports. Portal 3389 is selected, to enable remote
access to the Windows Server virtual machine from the internet. Opening port 3389 from the internet is not
recommended for production workloads.
8. Accept the remaining default settings and select OK.

9. On the Summary page, select Create. The virtual machine takes a few minutes to deploy.
10. Once the virtual machine is deployed, enter myPublicIpAddress in the search box at the top of the portal. When
myPublicIpAddress appears in the search results, select it.
11. You can view the public IP address that is assigned, and that the address is assigned to the myVM virtual
machine, as shown in the following picture:
Azure assigned a public IP address from addresses used in the region you created the virtual machine in.
You can download the list of ranges (prefixes) for the Azure Public, US government, China, and Germany
clouds.
12. Select Configuration to confirm that the assignment is Static.
WARNING
Do not modify the IP address settings within the virtual machine's operating system. The operating system is unaware of
Azure public IP addresses. Though you can add private IP address settings to the operating system, we recommend not
doing so unless necessary, and not until after reading Add a private IP address to an operating system.
Clean up resources
When no longer needed, delete the resource group and all of the resources it contains:
1. Enter myResourceGroup in the Search box at the top of the portal. When you see myResourceGroup in the
search results, select it.
2. Select Delete resource group.
3. Enter myResourceGroup for TYPE THE RESOURCE GROUP NAME: and select Delete.
Next steps
Learn more about public IP addresses in Azure
Learn more about all public IP address settings
Learn more about private IP addresses and assigning a static private IP address to an Azure virtual machine
Learn more about creating Linux and Windows virtual machines
Create and manage a Windows virtual machine that
has multiple NICs
Virtual machines (VMs) in Azure can have multiple virtual network interface cards (NICs) attached to them. A
common scenario is to have different subnets for front-end and back-end connectivity. You can associate multiple
NICs on a VM to multiple subnets, but those subnets must all reside in the same virtual network (vNet). This article
details how to create a VM that has multiple NICs attached to it. You also learn how to add or remove NICs from
an existing VM. Different VM sizes support a varying number of NICs, so size your VM accordingly.
Prerequisites
Make sure that you have the latest Azure PowerShell version installed and configured.
In the following examples, replace example parameter names with your own values. Example parameter names
include myResourceGroup, myVnet, and myVM.
Create a VM with multiple NICs

First, create a resource group. The following example creates a resource group named myResourceGroup in the
EastUs location:
New-AzureRmResourceGroup -Name "myResourceGroup" -Location "EastUS"
Create virtual network and subnets

A common scenario is for a virtual network to have two or more subnets. One subnet may be for front-end traffic,
the other for back-end traffic. To connect to both subnets, you then use multiple NICs on your VM.
1. Define two virtual network subnets with New -AzureRmVirtualNetworkSubnetConfig. The following
example defines the subnets for mySubnetFrontEnd and mySubnetBackEnd:
$mySubnetFrontEnd = New-AzureRmVirtualNetworkSubnetConfig -Name "mySubnetFrontEnd" `

-AddressPrefix "192.168.1.0/24"
$mySubnetBackEnd = New-AzureRmVirtualNetworkSubnetConfig -Name "mySubnetBackEnd" `
2. Create your virtual network and subnets with New -AzureRmVirtualNetwork. The following example creates
a virtual network named myVnet:
$myVnet = New-AzureRmVirtualNetwork -ResourceGroupName "myResourceGroup" `

-Location "EastUs" `
-Name "myVnet" `
-AddressPrefix "192.168.0.0/16" `
-Subnet $mySubnetFrontEnd,$mySubnetBackEnd
Create multiple NICs

Create two NICs with New -AzureRmNetworkInterface. Attach one NIC to the front-end subnet and one NIC to
the back-end subnet. The following example creates NICs named myNic1 and myNic2:
$frontEnd = $myVnet.Subnets|?{$_.Name -eq 'mySubnetFrontEnd'}
$myNic1 = New-AzureRmNetworkInterface -ResourceGroupName "myResourceGroup" `
-Name "myNic1" `
-SubnetId $frontEnd.Id
$backEnd = $myVnet.Subnets|?{$_.Name -eq 'mySubnetBackEnd'}

-Name "myNic2" `
-SubnetId $backEnd.Id
Typically you also create a network security group to filter network traffic to the VM and a load balancer to
distribute traffic across multiple VMs.
Now start to build your VM configuration. Each VM size has a limit for the total number of NICs that you can add
to a VM. For more information, see Windows VM sizes.
1. Set your VM credentials to the $cred variable as follows:
2. Define your VM with New -AzureRmVMConfig. The following example defines a VM named myVM and
uses a VM size that supports more than two NICs (Standard_DS3_v2):
$vmConfig = New-AzureRmVMConfig -VMName "myVM" -VMSize "Standard_DS3_v2"
3. Create the rest of your VM configuration with Set-AzureRmVMOperatingSystem and Set-

AzureRmVMSourceImage. The following example creates a Windows Server 2016 VM:
$vmConfig = Set-AzureRmVMOperatingSystem -VM $vmConfig `

-Windows `
-ComputerName "myVM" `
-Credential $cred `
-ProvisionVMAgent `
-EnableAutoUpdate
$vmConfig = Set-AzureRmVMSourceImage -VM $vmConfig `
-PublisherName "MicrosoftWindowsServer" `
-Offer "WindowsServer" `
-Skus "2016-Datacenter" `
-Version "latest"
4. Attach the two NICs that you previously created with Add-AzureRmVMNetworkInterface:
$vmConfig = Add-AzureRmVMNetworkInterface -VM $vmConfig -Id $myNic1.Id -Primary

$vmConfig = Add-AzureRmVMNetworkInterface -VM $vmConfig -Id $myNic2.Id
5. Create your VM with New -AzureRmVM:
New-AzureRmVM -VM $vmConfig -ResourceGroupName "myResourceGroup" -Location "EastUs"
6. Add routes for secondary NICs to the OS by completing the steps in Configure the operating system for
multiple NICs.
Add a NIC to an existing VM
To add a virtual NIC to an existing VM, you deallocate the VM, add the virtual NIC, then start the VM. Different
VM sizes support a varying number of NICs, so size your VM accordingly. If needed, you can resize a VM.
1. Deallocate the VM with Stop-AzureRmVM. The following example deallocates the VM named myVM in
myResourceGroup:
Stop-AzureRmVM -Name "myVM" -ResourceGroupName "myResourceGroup"
2. Get the existing configuration of the VM with Get-AzureRmVm. The following example gets information for
the VM named myVM in myResourceGroup:
$vm = Get-AzureRmVm -Name "myVM" -ResourceGroupName "myResourceGroup"
3. The following example creates a virtual NIC with New -AzureRmNetworkInterface named myNic3 that is
attached to mySubnetBackEnd. The virtual NIC is then attached to the VM named myVM in
myResourceGroup with Add-AzureRmVMNetworkInterface:
# Get info for the back end subnet

$myVnet = Get-AzureRmVirtualNetwork -Name "myVnet" -ResourceGroupName "myResourceGroup"
$backEnd = $myVnet.Subnets|?{$_.Name -eq 'mySubnetBackEnd'}
# Create a virtual NIC

-Name "myNic3" `
-SubnetId $backEnd.Id
# Get the ID of the new virtual NIC and add to VM

$nicId = (Get-AzureRmNetworkInterface -ResourceGroupName "myResourceGroup" -Name "MyNic3").Id
Add-AzureRmVMNetworkInterface -VM $vm -Id $nicId | Update-AzureRmVm -ResourceGroupName "myResourceGroup"
Primary virtual NICs

One of the NICs on a multi-NIC VM needs to be primary. If one of the existing virtual NICs on the VM is
already set as primary, you can skip this step. The following example assumes that two virtual NICs are now
present on a VM and you wish to add the first NIC ( [0] ) as the primary:
# List existing NICs on the VM and find which one is primary

$vm.NetworkProfile.NetworkInterfaces
# Set NIC 0 to be primary

$vm.NetworkProfile.NetworkInterfaces[0].Primary = $true
$vm.NetworkProfile.NetworkInterfaces[1].Primary = $false
# Update the VM state in Azure

Update-AzureRmVM -VM $vm -ResourceGroupName "myResourceGroup"
4. Start the VM with Start-AzureRmVm:
Start-AzureRmVM -ResourceGroupName "myResourceGroup" -Name "myVM"
5. Add routes for secondary NICs to the OS by completing the steps in Configure the operating system for
multiple NICs.
Remove a NIC from an existing VM
To remove a virtual NIC from an existing VM, you deallocate the VM, remove the virtual NIC, then start the VM.
1. Deallocate the VM with Stop-AzureRmVM. The following example deallocates the VM named myVM in
myResourceGroup:
Stop-AzureRmVM -Name "myVM" -ResourceGroupName "myResourceGroup"
2. Get the existing configuration of the VM with Get-AzureRmVm. The following example gets information for
the VM named myVM in myResourceGroup:
$vm = Get-AzureRmVm -Name "myVM" -ResourceGroupName "myResourceGroup"
3. Get information about the NIC remove with Get-AzureRmNetworkInterface. The following example gets
information about myNic3:
# List existing NICs on the VM if you need to determine NIC name

$vm.NetworkProfile.NetworkInterfaces
$nicId = (Get-AzureRmNetworkInterface -ResourceGroupName "myResourceGroup" -Name "myNic3").Id
4. Remove the NIC with Remove-AzureRmVMNetworkInterface and then update the VM with Update-
AzureRmVm. The following example removes myNic3 as obtained by $nicId in the preceding step:
Remove-AzureRmVMNetworkInterface -VM $vm -NetworkInterfaceIDs $nicId | `

Update-AzureRmVm -ResourceGroupName "myResourceGroup"
5. Start the VM with Start-AzureRmVm:
Start-AzureRmVM -Name "myVM" -ResourceGroupName "myResourceGroup"
Create multiple NICs with templates

Azure Resource Manager templates provide a way to create multiple instances of a resource during deployment,
such as creating multiple NICs. Resource Manager templates use declarative JSON files to define your
environment. For more information, see overview of Azure Resource Manager. You can use copy to specify the
number of instances to create:
"copy": {
"name": "multiplenics",
"count": "[parameters('count')]"
}
For more information, see creating multiple instances by using copy.

You can also use copyIndex() to append a number to a resource name. You can then create myNic1, MyNic2 and
so on. The following code shows an example of appending the index value:
"name": "[concat('myNic', copyIndex())]",

You can read a complete example of creating multiple NICs by using Resource Manager templates.
Add routes for secondary NICs to the OS by completing the steps in Configure the operating system for multiple
NICs.
Configure guest OS for multiple NICs

Azure assigns a default gateway to the first (primary) network interface attached to the virtual machine. Azure does
not assign a default gateway to additional (secondary) network interfaces attached to a virtual machine. Therefore,
you are unable to communicate with resources outside the subnet that a secondary network interface is in, by
default. Secondary network interfaces can, however, communicate with resources outside their subnet, though the
steps to enable communication are different for different operating systems.
1. From a Windows command prompt, run the route print command, which returns output similar to the
following output for a virtual machine with two attached network interfaces:
===========================================================================
Interface List
3...00 0d 3a 10 92 ce ......Microsoft Hyper-V Network Adapter #3
7...00 0d 3a 10 9b 2a ......Microsoft Hyper-V Network Adapter #4
===========================================================================
In this example, Microsoft Hyper-V Network Adapter #4 (interface 7) is the secondary network interface
that doesn't have a default gateway assigned to it.
2. From a command prompt, run the ipconfig command to see which IP address is assigned to the secondary
network interface. In this example, 192.168.2.4 is assigned to interface 7. No default gateway address is
returned for the secondary network interface.
3. To route all traffic destined for addresses outside the subnet of the secondary network interface to the
gateway for the subnet, run the following command:
route add -p 0.0.0.0 MASK 0.0.0.0 192.168.2.1 METRIC 5015 IF 7
The gateway address for the subnet is the first IP address (ending in .1) in the address range defined for the
subnet. If you don't want to route all traffic outside the subnet, you could add individual routes to specific
destinations, instead. For example, if you only wanted to route traffic from the secondary network interface
to the 192.168.3.0 network, you enter the command:
route add -p 192.168.3.0 MASK 255.255.255.0 192.168.2.1 METRIC 5015 IF 7
4. To confirm successful communication with a resource on the 192.168.3.0 network, for example, enter the
following command to ping 192.168.3.4 using interface 7 (192.168.2.4):
ping 192.168.3.4 -S 192.168.2.4
You may need to open ICMP through the Windows firewall of the device you're pinging with the following
command:
netsh advfirewall firewall add rule name=Allow-ping protocol=icmpv4 dir=in action=allow
5. To confirm the added route is in the route table, enter the route print command, which returns output
similar to the following text:
===========================================================================
Active Routes:
Network Destination Netmask Gateway Interface Metric
0.0.0.0 0.0.0.0 192.168.1.1 192.168.1.4 15
0.0.0.0 0.0.0.0 192.168.2.1 192.168.2.4 5015
The route listed with 192.168.1.1 under Gateway, is the route that is there by default for the primary
network interface. The route with 192.168.2.1 under Gateway, is the route you added.
Next steps
Review Windows VM sizes when you're trying to create a VM that has multiple NICs. Pay attention to the
maximum number of NICs that each VM size supports.
Create a Windows virtual machine with Accelerated
Networking
In this tutorial, you learn how to create a Windows virtual machine (VM ) with Accelerated Networking. To create a
Linux VM with Accelerated Networking, see Create a Linux VM with Accelerated Networking. Accelerated
networking enables single root I/O virtualization (SR -IOV ) to a VM, greatly improving its networking performance.
This high-performance path bypasses the host from the datapath, reducing latency, jitter, and CPU utilization, for
use with the most demanding network workloads on supported VM types. The following picture shows
communication between two VMs with and without accelerated networking:
Without accelerated networking, all networking traffic in and out of the VM must traverse the host and the virtual
switch. The virtual switch provides all policy enforcement, such as network security groups, access control lists,
isolation, and other network virtualized services to network traffic. To learn more about virtual switches, read the
Hyper-V network virtualization and virtual switch article.
With accelerated networking, network traffic arrives at the VM's network interface (NIC ), and is then forwarded to
the VM. All network policies that the virtual switch applies are now offloaded and applied in hardware. Applying
policy in hardware enables the NIC to forward network traffic directly to the VM, bypassing the host and the
virtual switch, while maintaining all the policy it applied in the host.
The benefits of accelerated networking only apply to the VM that it is enabled on. For the best results, it is ideal to
enable this feature on at least two VMs connected to the same Azure Virtual Network (VNet). When
communicating across VNets or connecting on-premises, this feature has minimal impact to overall latency.
Benefits
Lower Latency / Higher packets per second (pps): Removing the virtual switch from the datapath removes
the time packets spend in the host for policy processing and increases the number of packets that can be
processed inside the VM.
Reduced jitter: Virtual switch processing depends on the amount of policy that needs to be applied and the
workload of the CPU that is doing the processing. Offloading the policy enforcement to the hardware removes
that variability by delivering packets directly to the VM, removing the host to VM communication and all
software interrupts and context switches.
Decreased CPU utilization: Bypassing the virtual switch in the host leads to less CPU utilization for
processing network traffic.
Limitations and Constraints

Supported operating systems
The following distributions are supported out of the box from the Azure Gallery:
Windows Server 2016 Datacenter
Windows Server 2012 R2 Datacenter
Supported VM instances
Accelerated Networking is supported on most general purpose and compute-optimized instance sizes with 2 or
more vCPUs. These supported series are: D/DSv2 and F/Fs
On instances that support hyperthreading, Accelerated Networking is supported on VM instances with 4 or more
vCPUs. Supported series are: D/DSv3, E/ESv3, Fsv2, and Ms/Mms
For more information on VM instances, see Windows VM sizes.
Regions
Available in all public Azure regions and Azure Government Cloud.
Enabling Accelerated Networking on a running VM
A supported VM size without accelerated networking enabled can only have the feature enabled when it is stopped
and deallocated.
Deployment through Azure Resource Manager
Virtual machines (classic) cannot be deployed with Accelerated Networking.
Create a Windows VM with Azure Accelerated Networking

Though this article provides steps to create a virtual machine with accelerated networking using Azure PowerShell,
you can also create a virtual machine with accelerated networking using the Azure portal . When creating a virtual
machine in the portal, under Settings, select Enabled, under Accelerated networking. The option to enable
accelerated networking doesn't appear in the portal unless you've selected a supported operating system and VM
size. After the virtual machine is created, you need to complete the instructions in Confirm the driver is installed in
the operating system.
Create a virtual network

Install Azure PowerShell version 5.1.1 or later. To find your currently installed version, run
Get-Module -ListAvailable AzureRM . If you need to install or upgrade, install the latest version of the AzureRM
module from the PowerShell Gallery. In a PowerShell session, log in to an Azure account using Connect-
AzureRmAccount.
In the following examples, replace example parameter names with your own values. Example parameter names
included myResourceGroup, myNic, and myVM.
Create a resource group with New -AzureRmResourceGroup. The following example creates a resource group
named myResourceGroup in the centralus location:
New-AzureRmResourceGroup -Name "myResourceGroup" -Location "centralus"
First, create a subnet configuration with New -AzureRmVirtualNetworkSubnetConfig. The following example
creates a subnet named mySubnet:
$subnet = New-AzureRmVirtualNetworkSubnetConfig `
-Name "mySubnet" `
Create a virtual network with New -AzureRmVirtualNetwork, with the mySubnet subnet.
$vnet = New-AzureRmVirtualNetwork -ResourceGroupName "myResourceGroup" `

-Location "centralus" `
-Name "myVnet" `
-AddressPrefix "192.168.0.0/16" `
-Subnet $Subnet
Create a network security group

First, create a network security group rule with New -AzureRmNetworkSecurityRuleConfig.
$rdp = New-AzureRmNetworkSecurityRuleConfig `
-Name 'Allow-RDP-All' `
-Description 'Allow RDP' `
-Access Allow `
-Protocol Tcp `
-Priority 100 `
-SourceAddressPrefix * `
-DestinationPortRange 3389
Create a network security group with New -AzureRmNetworkSecurityGroup and assign the Allow -RDP -All
security rule to it. In addition to the Allow -RDP -All rule, the network security group contains several default rules.
One default rule disables all inbound access from the Internet, which is why the Allow -RDP -All rule is assigned to
the network security group so that you can remotely connect to the virtual machine, once it's created.
-Location centralus `
-Name "myNsg" `
-SecurityRules $rdp
Associate the network security group to the mySubnet subnet with Set-AzureRmVirtualNetworkSubnetConfig.
The rule in the network security group is effective for all resources deployed in the subnet.
Set-AzureRmVirtualNetworkSubnetConfig `
-Name 'mySubnet' `
-AddressPrefix "192.168.1.0/24" `
-NetworkSecurityGroup $nsg
Create a network interface with accelerated networking

Create a public IP address with New -AzureRmPublicIpAddress. A public IP address isn't required if you don't plan
to access the virtual machine from the Internet, but to complete the steps in this article, it is required.
$publicIp = New-AzureRmPublicIpAddress `
-Name 'myPublicIp' `
-location centralus `
-AllocationMethod Dynamic
Create a network interface with New -AzureRmNetworkInterface with accelerated networking enabled and assign
the public IP address to the network interface. The following example creates a network interface named myNic in
the mySubnet subnet of the myVnet virtual network and assigns the myPublicIp public IP address to it:
$nic = New-AzureRmNetworkInterface `
-Name "myNic" `
-Location "centralus" `
-SubnetId $vnet.Subnets[0].Id `
-PublicIpAddressId $publicIp.Id `
-EnableAcceleratedNetworking

Set your VM credentials to the $cred variable using Get-Credential:
First, define your VM with New -AzureRmVMConfig. The following example defines a VM named myVM with a
VM size that supports Accelerated Networking (Standard_DS4_v2):
$vmConfig = New-AzureRmVMConfig -VMName "myVm" -VMSize "Standard_DS4_v2"
For a list of all VM sizes and characteristics, see Windows VM sizes.

Create the rest of your VM configuration with Set-AzureRmVMOperatingSystem and Set-
AzureRmVMSourceImage. The following example creates a Windows Server 2016 VM:
$vmConfig = Set-AzureRmVMOperatingSystem -VM $vmConfig `

-Windows `
-ComputerName "myVM" `
-Credential $cred `
-ProvisionVMAgent `
-EnableAutoUpdate
$vmConfig = Set-AzureRmVMSourceImage -VM $vmConfig `
-PublisherName "MicrosoftWindowsServer" `
-Offer "WindowsServer" `
-Skus "2016-Datacenter" `
-Version "latest"
Attach the network interface that you previously created with Add-AzureRmVMNetworkInterface:
$vmConfig = Add-AzureRmVMNetworkInterface -VM $vmConfig -Id $nic.Id
Finally, create your VM with New -AzureRmVM:
New-AzureRmVM -VM $vmConfig -ResourceGroupName "myResourceGroup" -Location "centralus"

Confirm the driver is installed in the operating system
Once you create the VM in Azure, connect to the VM and confirm that the driver is installed in Windows.
1. From an Internet browser, open the Azure portal and sign in with your Azure account.
2. In the box that contains the text Search resources at the top of the Azure portal, type myVm. When myVm
appears in the search results, click it. If Creating is visible under the Connect button, Azure has not yet finished
creating the VM. Click Connect in the top left corner of the overview only after you no longer see Creating
under the Connect button.
3. Enter the username and password you entered in Create the virtual machine. If you've never connected to a
Windows VM in Azure, see Connect to virtual machine.
4. Right-click the Windows Start button and click Device Manager. Expand the Network adapters node.
Confirm that the Mellanox ConnectX-3 Virtual Function Ethernet Adapter appears, as shown in the
following picture:
Accelerated Networking is now enabled for your VM.
Enable Accelerated Networking on existing VMs

If you have created a VM without Accelerated Networking, it is possible to enable this feature on an existing VM.
The VM must support Accelerated Networking by meeting the following prerequisites that are also outlined above:
The VM must be a supported size for Accelerated Networking
The VM must be a supported Azure Gallery image (and kernel version for Linux)
All VMs in an availability set or VMSS must be stopped/deallocated before enabling Accelerated Networking
on any NIC
Individual VMs & VMs in an availability set
First stop/deallocate the VM or, if an Availability Set, all the VMs in the Set:
Stop-AzureRmVM -ResourceGroup "myResourceGroup" `

-Name "myVM"
Important, please note, if your VM was created individually, without an availability set, you only need to
stop/deallocate the individual VM to enable Accelerated Networking. If your VM was created with an availability
set, all VMs contained in the availability set will need to be stopped/deallocated before enabling Accelerated
Networking on any of the NICs.
Once stopped, enable Accelerated Networking on the NIC of your VM:
$nic = Get-AzureRmNetworkInterface -ResourceGroupName "myResourceGroup" `

-Name "myNic"
$nic.EnableAcceleratedNetworking = $true
$nic | Set-AzureRmNetworkInterface
Restart your VM or, if in an Availability Set, all the VMs in the Set and confirm that Accelerated Networking is
enabled:
Start-AzureRmVM -ResourceGroup "myResourceGroup" `

-Name "myVM"
VMSS
VMSS is slightly different but follows the same workflow. First, stop the VMs:
Stop-AzureRmVmss -ResourceGroupName "myResourceGroup" `

Once the VMs are stopped, update the Accelerated Networking property under the network interface:
$vmss = Get-AzureRmVmss -ResourceGroupName "myResourceGroup" `

$vmss.VirtualMachineProfile.NetworkProfile.NetworkInterfaceConfigurations[0].EnableAcceleratedNetworking =
$true
Update-AzureRmVmss -ResourceGroupName "myResourceGroup" `

Please note, a VMSS has VM upgrades that apply updates using three different settings, automatic, rolling and
manual. In these instructions the policy is set to automatic so that the VMSS will pick up the changes immediately
after restarting. To set it to automatic so that the changes are immediately picked up:
$vmss.UpgradePolicy.AutomaticOSUpgrade = $true
Update-AzureRmVmss -ResourceGroupName "myResourceGroup" `

Finally, restart the VMSS:
Start-AzureRmVmss -ResourceGroupName "myResourceGroup" `

Once you restart, wait for the upgrades to finish but once completed, the VF will appear inside the VM. (Please
make sure you are using a supported OS and VM size)
Resizing existing VMs with Accelerated Networking
VMs with Accelerated Networking enabled can only be resized to VMs that support Accelerated Networking.
A VM with Accelerated Networking enabled cannot be resized to a VM instance that does not support Accelerated
Networking using the resize operation. Instead, to resize one of these VMs:
Stop/Deallocate the VM or if in an availability set/VMSS, stop/deallocate all the VMs in the set/VMSS.
Accelerated Networking must be disabled on the NIC of the VM or if in an availability set/VMSS, all VMs in the
set/VMSS.
Once Accelerated Networking is disabled, the VM/availability set/VMSS can be moved to a new size that does
not support Accelerated Networking and restarted.
Create a fully qualified domain name in the Azure
portal for a Windows VM
When you create a virtual machine (VM ) in the Azure portal, a public IP resource for the virtual machine is
automatically created. You use this IP address to remotely access the VM. Although the portal does not create a
fully qualified domain name, or FQDN, you can create one once the VM is created. This article demonstrates the
steps to create a DNS name or FQDN.
Create a FQDN
This article assumes that you have already created a VM. If needed, you can create a VM in the portal or with
Azure PowerShell. Follow these steps once your VM is up and running:
1. Select your VM in the portal. Under DNS name, click Configure.
2. Enter the desired DNS name and then select Save.

3. To return to the VM overview blade, close the Public IP address blade. Verify that the DNS name is now
shown.
You can now connect remotely to the VM using this DNS name such as for Remote Desktop Protocol (RDP ).
Next steps
Now that your VM has a public IP and DNS name, you can deploy common application frameworks or services
such as IIS, SQL, or SharePoint.
You can also read more about using Resource Manager for tips on building your Azure deployments.
How Azure DNS works with other Azure services
Azure DNS is a hosted DNS management and name resolution service. This allows you to create public DNS
names for the other applications and services you have deployed in Azure. Creating a name for an Azure service in
your custom domain is as simple as adding a record of the correct type for your service.
For dynamically allocated IP addresses, you can create a DNS CNAME record that maps to the DNS name that
Azure created for your service. DNS standards prevent you from using a CNAME record for the zone apex, but
you can use an alias record instead. For more information, see Tutorial: Configure an alias record to refer to an
Azure Public IP address.
For statically allocated IP addresses, you can create a DNS A record using any name, including a naked domain
name at the zone apex.
The following table outlines the supported record types that can be used for various Azure services. As you can see
from this table, Azure DNS only supports DNS records for Internet-facing network resources. Azure DNS cannot
be used for name resolution of internal, private addresses.
AZURE SERVICE NETWORK INTERFACE DESCRIPTION
Application Gateway Front-end Public IP You can create a DNS A or CNAME

record.
Load Balancer Front-end Public IP You can create a DNS A or CNAME

record. Load Balancer can have an IPv6
Public IP address that is dynamically
assigned. Therefore, you must create a
CNAME record for an IPv6 address.
Traffic Manager Public name You can create an alias record that
maps to the trafficmanager.net name
assigned to your Traffic Manager profile.
For more information, see Tutorial:
Configure an alias record to support
apex domain names with Traffic
Manager.
Cloud Service Public IP For statically allocated IP addresses, you

can create a DNS A record. For
dynamically allocated IP addresses, you
must create a CNAME record that maps
to the cloudapp.net name.
App Service External IP For external IP addresses, you can

create a DNS A record. Otherwise, you
must create a CNAME record that maps
to the azurewebsites.net name. For
more information, see Map a custom
domain name to an Azure app
AZURE SERVICE NETWORK INTERFACE DESCRIPTION
Resource Manager VMs Public IP Resource Manager VMs can have Public
IP addresses. A VM with a Public IP
address may also be behind a load
balancer. You can create a DNS A,
CNAME, or alias record for the Public
address. This custom name can be used
to bypass the VIP on the load balancer.
Classic VMs Public IP Classic VMs created using PowerShell or

CLI can be configured with a dynamic or
static (reserved) virtual address. You can
create a DNS CNAME or A record,
respectively.
Azure virtual machine extensions and features
Azure virtual machine (VM ) extensions are small applications that provide post-deployment configuration and
automation tasks on Azure VMs, you can use existing images and then customize them as part of your
deployments, getting you out of the business of custom image building.
The Azure platform hosts many extensions that range from VM configuration, monitoring, security, and utility
applications. Publishers take an application, then wrap it into an extension, and simplify the installation, so all you
need to do is provide mandatory parameters.
There is a large choice of first and third party extensions, if the application in the extension repository does not
exist, then you can use the Custom Script extension and configure your VM with your own scripts and commands.
Examples of key scenarios that extensions are used for:
VM configuration, you can use Powershell DSC (Desired State Configuration), Chef, Puppet and Custom Script
Extensions to install VM configuration agents and configure your VM.
AV products, such as Symantec, ESET.
VM vulnerability tool, such as Qualys, Rapid7, HPE.
VM and App monitoring tooling, such as DynaTrace, Azure Network Watcher, Site24x7, and Stackify.
Extensions can be bundled with a new VM deployment. For example, they can be part of a larger deployment,
configuring applications on VM provision, or run against any supported extension operated systems post
deployment.
How can I find What extensions are available?

You can view available extensions in the VM blade in the Portal, under extensions, this represents just a small
amount, for the full list, you can use the CLI tools, see Discovering VM Extensions for Linux and Discovering VM
Extensions for Windows.
How can I install an extension?

Azure VM extensions can be managed using either the Azure CLI, Azure PowerShell, Azure Resource Manager
templates, and the Azure portal. To try an extension, you can go to the Azure portal, select the Custom Script
Extension, then pass in a command / script and run the extensions.
If you want to same extension you added in the portal by CLI or Resource Manager template, see different
extension documentation, such as Windows Custom Script Extension and Linux Custom Script Extension.
How do I manage extension application lifecycle?

You do not need to connect to a VM directly to install or delete the extension. As the Azure extension application
lifecycle is managed outside of the VM and integrated into the Azure platform, you also get integrated status of the
extension.
Anything else I should be thinking about for extensions?

Extensions install applications, like any applications there are some requirements, for extensions there is a list of
supported Windows and Linux OSes, and you need to have the Azure VM agents installed. Some individual VM
extension applications may have their own environmental prerequisites, such as access to an endpoint.
Next steps
For more information about how the Linux Agent and Extensions work, see Azure VM extensions and features
for Linux.
For more information about how the Windows Guest Agent and Extensions work, see Azure VM extensions
and features for Windows.
To install the Windows Guest Agent, see Azure Windows Virtual Machine Agent Overview .
To install the Linux Agent, see Azure Linux Virtual Machine Agent Overview .
Migrate from Amazon Web Services (AWS) and other
platforms to Managed Disks in Azure
You can upload VHD files from AWS or on-premises virtualization solutions to Azure to create VMs that take
advantage of Managed Disks. Azure Managed Disks removes the need to manage storage accounts for Azure IaaS
VMs. You have to only specify the type (Premium or Standard) and size of disk you need, and Azure creates and
manages the disk for you.
You can upload either generalized and specialized VHDs.
Generalized VHD - has had all of your personal account information removed using Sysprep.
Specialized VHD - maintains the user accounts, applications, and other state data from your original VM.
IMPORTANT
Before uploading any VHD to Azure, you should follow Prepare a Windows VHD or VHDX to upload to Azure
SCENARIO DOCUMENTATION
You have existing AWS EC2 instances that you would like to Move a VM from Amazon Web Services (AWS) to Azure
migrate to Azure VMs using managed disks
You have a VM from another virtualization platform that you Upload a generalized VHD and use it to create a new VM in
would like to use as an image to create multiple Azure VMs. Azure
You have a uniquely customized VM that you would like to Upload a specialized VHD to Azure and create a new VM
recreate in Azure.
Overview of Managed Disks

Azure Managed Disks simplifies VM management by removing the need to manage storage accounts. Managed
Disks also benefit from better reliability of VMs in an Availability Set. It ensures that the disks of different VMs in
an Availability Set are sufficiently isolated from each other to avoid a single point of failure. It automatically places
disks of different VMs in an Availability Set in different Storage scale units (stamps) which limits the impact of
single Storage scale unit failures caused due to hardware and software failures. Based on your needs, you can
choose from two types of storage options:
Premium Managed Disks are Solid State Drive (SSD ) based storage media, which delivers high
performance, low -latency disk support for virtual machines running I/O -intensive workloads. You can take
advantage of the speed and performance of these disks by migrating to Premium Managed Disks.
Standard Managed Disks use Hard Disk Drive (HDD ) based storage media and are best suited for Dev/Test
and other infrequent access workloads that are less sensitive to performance variability.
Plan for the migration to Managed Disks

If you are planning on migrating from unmanaged disks to managed disks, you should be aware that users with
the Virtual Machine Contributor role will not be able to change the VM size (as they could pre-conversion). This is
because VMs with managed disks require the user to have the Microsoft.Compute/disks/write permission on the
OS disks.
Location
Pick a location where Azure Managed Disks are available. If you are migrating to Premium Managed Disks, also
ensure that Premium storage is available in the region where you are planning to migrate to. See Azure Services
by Region for up-to-date information on available locations.
VM sizes
Disk sizes
There are three types of Premium Managed disks that can be used with your VM and each has specific IOPs and
throughput limits. Consider these limits when choosing the Premium disk type for your VM based on the needs of
your application in terms of capacity, performance, scalability, and peak loads.
PREMIUM DISKS TYPE P10 P20 P30
Disk size 128 GB 512 GB 1024 GB (1 TB)
IOPS per disk 500 2300 5000
Throughput per disk 100 MB per second 150 MB per second 200 MB per second

There are five types of Standard Managed disks that can be used with your VM. Each of them have different
STANDARD DISK
TYPE S4 S6 S10 S20 S30
Disk size 30 GB 64 GB 128 GB 512 GB 1024 GB (1 TB)
IOPS per disk 500 500 500 500 500
Throughput per 60 MB per 60 MB per 60 MB per 60 MB per 60 MB per

disk second second second second second
Disk caching policy

Pricing
Next Steps
Before uploading any VHD to Azure, you should follow Prepare a Windows VHD or VHDX to upload to Azure
Upload a generalized VHD and use it to create new
VMs in Azure
This article walks you through using PowerShell to upload a VHD of a generalized VM to Azure, create an image
from the VHD, and create a new VM from that image. You can upload a VHD exported from an on-premises
virtualization tool or from another cloud. Using Managed Disks for the new VM simplifies the VM management
and provides better availability when the VM is placed in an availability set.
For a sample script, see Sample script to upload a VHD to Azure and create a new VM.
Before you begin

Before uploading any VHD to Azure, you should follow Prepare a Windows VHD or VHDX to upload to Azure.
Review Plan for the migration to Managed Disks before starting your migration to Managed Disks.
This article requires the AzureRM module, version 5.6 or later. Run Get-Module -ListAvailable AzureRM.Compute
to find your version. If you need to upgrade, see Install Azure PowerShell module.
Generalize the source VM by using Sysprep

Sysprep removes all your personal account information, among other things, and prepares the machine to be
used as an image. For details about Sysprep, see the Sysprep Overview.
Make sure the server roles running on the machine are supported by Sysprep. For more information, see Sysprep
Support for Server Roles.
IMPORTANT
If you plan to run Sysprep before uploading your VHD to Azure for the first time, make sure you have prepared your VM.
1. Sign in to the Windows virtual machine.

2. Open the Command Prompt window as an administrator. Change the directory to
%windir%\system32\sysprep, and then run sysprep.exe .
3. In the System Preparation Tool dialog box, select Enter System Out-of-Box Experience (OOBE ), and
make sure that the Generalize check box is enabled.
4. For Shutdown Options, select Shutdown.
5. Select OK.
6. When Sysprep finishes, it shuts down the virtual machine. Do not restart the VM.
Get a storage account

You'll need a storage account in Azure to store the uploaded VM image. You can either use an existing storage
account or create a new one.
If you'll be using the VHD to create a managed disk for a VM, the storage account location must be the same
location where you'll be creating the VM.
To show the available storage accounts, enter:
Get-AzureRmStorageAccount | Format-Table
Upload the VHD to your storage account

Use the Add-AzureRmVhd cmdlet to upload the VHD to a container in your storage account. This example
uploads the file myVHD.vhd from C:\Users\Public\Documents\Virtual hard disks\ to a storage account named
mystorageaccount in the myResourceGroup resource group. The file will be placed into the container named
mycontainer and the new file name will be myUploadedVHD.vhd.
$urlOfUploadedImageVhd = "https://mystorageaccount.blob.core.windows.net/mycontainer/myUploadedVHD.vhd"
Add-AzureRmVhd -ResourceGroupName $rgName -Destination $urlOfUploadedImageVhd `
-LocalFilePath "C:\Users\Public\Documents\Virtual hard disks\myVHD.vhd"
If successful, you get a response that looks similar to this:
MD5 hash is being calculated for the file C:\Users\Public\Documents\Virtual hard disks\myVHD.vhd.
MD5 hash calculation is completed.
Elapsed time for the operation: 00:03:35
Creating new page blob of size 53687091712...
Elapsed time for upload: 01:12:49
LocalFilePath DestinationUri
------------- --------------
C:\Users\Public\Doc... https://mystorageaccount.blob.core.windows.net/mycontainer/myUploadedVHD.vhd
Depending on your network connection and the size of your VHD file, this command may take a while to
complete.
Other options for uploading a VHD
You can also upload a VHD to your storage account using one of the following:
AzCopy
Azure Storage Copy Blob API
Azure Storage Explorer Uploading Blobs
Storage Import/Export Service REST API Reference
We recommend using the Import/Export Service if estimated uploading time is longer than seven days. You
can use DataTransferSpeedCalculator to estimate the time from data size and transfer unit. Import/Export can
be used to copy to a standard storage account. You'll need to copy from standard storage to premium storage
account by using a tool like AzCopy.
IMPORTANT
If you are using AzCopy to upload your VHD to Azure, make sure you have set /BlobType:page before running your
upload script. If the destination is a blob and this option is not specified, by default AzCopy creates a block blob.
Create a managed image from the uploaded VHD

Create a managed image from your generalized OS VHD. Replace the following values with your own
information.
First, set some parameters:

Create the image using your generalized OS VHD.
$imageConfig = New-AzureRmImageConfig `
-Location $location
$imageConfig = Set-AzureRmImageOsDisk `
-Image $imageConfig `
-OsType Windows `
-OsState Generalized `
-BlobUri $urlOfUploadedImageVhd `
-DiskSizeGB 20
New-AzureRmImage `
-ImageName $imageName `
-Image $imageConfig
Create the VM
Now that you have an image, you can create one or more new VMs from the image. This example creates a VM
named myVM from myImage, in myResourceGroup.
New-AzureRmVm `
-Name "myVM" `
-ImageName $imageName `
-SecurityGroupName "myNSG" `
-PublicIpAddressName "myPIP" `
-OpenPorts 3389
Next steps
Sign in to your new virtual machine. For more information, see How to connect and log on to an Azure virtual
machine running Windows.
Move a Windows VM from Amazon Web Services
(AWS) to Azure using PowerShell
If you are evaluating Azure virtual machines for hosting your workloads, you can export an existing Amazon Web
Services (AWS ) EC2 Windows VM instance then upload the virtual hard disk (VHD ) to Azure. Once the VHD is
uploaded, you can create a new VM in Azure from the VHD.
This article covers moving a single VM from AWS to Azure. If you want to move VMs from AWS to Azure at scale,
see Migrate virtual machines in Amazon Web Services (AWS ) to Azure with Azure Site Recovery.
Prepare the VM
You can upload both generalized and specialized VHDs to Azure. Each type requires that you prepare the VM
before exporting from AWS.
Generalized VHD - a generalized VHD has had all of your personal account information removed using
Sysprep. If you intend to use the VHD as an image to create new VMs from, you should:
Prepare a Windows VM.
Generalize the virtual machine using Sysprep.
Specialized VHD - a specialized VHD maintains the user accounts, applications and other state data from
your original VM. If you intend to use the VHD as-is to create a new VM, ensure the following steps are
completed.
Prepare a Windows VHD to upload to Azure. Do not generalize the VM using Sysprep.
Remove any guest virtualization tools and agents that are installed on the VM (i.e. VMware tools).
Ensure the VM is configured to pull its IP address and DNS settings via DHCP. This ensures that the
server obtains an IP address within the VNet when it starts up.
Export and download the VHD

Export the EC2 instance to a VHD in an Amazon S3 bucket. Follow the steps in the Amazon documentation article
Exporting an Instance as a VM Using VM Import/Export and run the create-instance-export-task command to
export the EC2 instance to a VHD file.
The exported VHD file is saved in the Amazon S3 bucket you specify. The basic syntax for exporting the VHD is
below, just replace the placeholder text in with your information.
aws ec2 create-instance-export-task --instance-id <instanceID> --target-environment Microsoft \

--export-to-s3-task DiskImageFormat=VHD,ContainerFormat=ova,S3Bucket=<bucket>,S3Prefix=<prefix>
Once the VHD has been exported, follow the instructions in How Do I Download an Object from an S3 Bucket? to
download the VHD file from the S3 bucket.
IMPORTANT
AWS charges data transfer fees for downloading the VHD. See Amazon S3 Pricing for more information.
Next steps
Now you can upload the VHD to Azure and create a new VM.
If you ran Sysprep on your source to generalize it before exporting, see Upload a generalized VHD and use it
to create a new VMs in Azure
If you did not run Sysprep before exporting, the VHD is considered specialized, see Upload a specialized VHD
to Azure and create a new VM
2 minutes to read
Platform-supported migration of IaaS resources from
classic to Azure Resource Manager
This article describes how to migrate infrastructure as a service (IaaS ) resources from the Classic to Resource
Manager deployment models and details how to connect resources from the two deployment models that coexist
in your subscription by using virtual network site-to-site gateways. You can read more about Azure Resource
Manager features and benefits.
Goal for migration

Resource Manager enables deploying complex applications through templates, configures virtual machines by
using VM extensions, and incorporates access management and tagging. Azure Resource Manager includes
scalable, parallel deployment for virtual machines into availability sets. The new deployment model also provides
lifecycle management of compute, network, and storage independently. Finally, there’s a focus on enabling
security by default with the enforcement of virtual machines in a virtual network.
Almost all the features from the classic deployment model are supported for compute, network, and storage
under Azure Resource Manager. To benefit from the new capabilities in Azure Resource Manager, you can
migrate existing deployments from the Classic deployment model.
Supported resources for migration

These classic IaaS resources are supported during migration
Virtual Machines
Availability Sets
Cloud Services with Virtual Machines
Storage Accounts
Virtual Networks
VPN Gateways
Express Route Gateways (in the same subscription as Virtual Network only )
Network Security Groups
Route Tables
Reserved IPs
Supported scopes of migration

There are four different ways to complete migration of compute, network, and storage resources:
Migration of virtual machines (NOT in a virtual network)
Migration of virtual machines (in a virtual network)
Migration of storage accounts
Migration of unattached resources
Migration of virtual machines (NOT in a virtual network)
In the Resource Manager deployment model, security is enforced for your applications by default. All VMs need
to be in a virtual network in the Resource Manager model. The Azure platform restarts ( Stop , Deallocate , and
Start ) the VMs as part of the migration. You have two options for the virtual networks that the Virtual Machines
will be migrated to:
You can request the platform to create a new virtual network and migrate the virtual machine into the new
virtual network.
You can migrate the virtual machine into an existing virtual network in Resource Manager.
NOTE
In this migration scope, both the management-plane operations and the data-plane operations may not be allowed for a
period of time during the migration.
Migration of virtual machines (in a virtual network)

For most VM configurations, only the metadata is migrating between the Classic and Resource Manager
deployment models. The underlying VMs are running on the same hardware, in the same network, and with the
same storage. The management-plane operations may not be allowed for a certain period of time during the
migration. However, the data plane continues to work. That is, your applications running on top of VMs (classic)
do not incur downtime during the migration.
The following configurations are not currently supported. If support is added in the future, some VMs in this
configuration might incur downtime (go through stop, deallocate, and restart VM operations).
You have more than one availability set in a single cloud service.
You have one or more availability sets and VMs that are not in an availability set in a single cloud service.
NOTE
In this migration scope, the management plane may not be allowed for a period of time during the migration. For certain
configurations as described earlier, data-plane downtime occurs.
Migration of storage accounts

To allow seamless migration, you can deploy Resource Manager VMs in a classic storage account. With this
capability, compute and network resources can and should be migrated independently of storage accounts. Once
you migrate over your Virtual Machines and Virtual Network, you need to migrate over your storage accounts to
complete the migration process.
If your storage account does not have any associated disks or Virtual Machines data and only has blobs, files,
tables, and queues then the migration to Azure Resource Manager can be done as a standalone migration
without dependencies.
NOTE
The Resource Manager deployment model doesn't have the concept of Classic images and disks. When the storage account
is migrated, Classic images and disks are not visible in the Resource Manager stack but the backing VHDs remain in the
storage account.
Migration of unattached resources

Storage Accounts with no associated disks or Virtual Machines data may be migrated independently.
Network Security Groups, Route Tables & Reserved IPs that are not attached to any Virtual Machines and Virtual
Networks can also be migrated independently.
Unsupported features and configurations
Some features and configurations are not currently supported; the following sections describe our
recommendations around them.
Unsupported features
The following features are not currently supported. You can optionally remove these settings, migrate the VMs,
and then re-enable the settings in the Resource Manager deployment model.
RESOURCE PROVIDER FEATURE RECOMMENDATION
Compute Unassociated virtual machine disks. The VHD blobs behind these disks will
get migrated when the Storage
Account is migrated
Compute Virtual machine images. The VHD blobs behind these disks will
get migrated when the Storage
Account is migrated
Network Endpoint ACLs. Remove Endpoint ACLs and retry

migration.
Network Application Gateway Remove the Application Gateway

before beginning migration and then
recreate the Application Gateway once
migration is complete.
Network Virtual networks using VNet Peering. Migrate Virtual Network to Resource
Manager, then peer. Learn more about
VNet Peering.
Unsupported configurations
The following configurations are not currently supported.
SERVICE CONFIGURATION RECOMMENDATION
Resource Manager Role Based Access Control (RBAC) for Because the URI of the resources is
classic resources modified after migration, it is
recommended that you plan the RBAC
policy updates that need to happen
after migration.
Compute Multiple subnets associated with a VM Update the subnet configuration to

reference only one subnet. This may
require you to remove a secondary NIC
(that is referring to another subnet)
from the VM and reattach it after
migration has completed.
Compute Virtual machines that belong to a You can optionally delete the VM.
virtual network but don't have an
explicit subnet assigned
Compute Virtual machines that have alerts, The migration goes through and these
Autoscale policies settings are dropped. It is highly
recommended that you evaluate your
environment before you do the
migration. Alternatively, you can
reconfigure the alert settings after
migration is complete.
Compute XML VM extensions (BGInfo 1.*, Visual This is not supported. It is

Studio Debugger, Web Deploy, and recommended that you remove these
Remote Debugging) extensions from the virtual machine to
continue migration or they will be
dropped automatically during the
migration process.
Compute Boot diagnostics with Premium storage Disable Boot Diagnostics feature for the
VMs before continuing with migration.
You can re-enable boot diagnostics in
the Resource Manager stack after the
migration is complete. Additionally,
blobs that are being used for
screenshot and serial logs should be
deleted so you are no longer charged
for those blobs.
Compute Cloud services that contain web/worker This is currently not supported.
roles
Compute Cloud services that contain more than This is currently not supported. Please
one availability set or multiple move the Virtual Machines to the same
availability sets. availability set before migrating.
Compute VM with Azure Security Center Azure Security Center automatically

extension installs extensions on your Virtual
Machines to monitor their security and
raise alerts. These extensions usually
get installed automatically if the Azure
Security Center policy is enabled on the
subscription. To migrate the Virtual
Machines, disable the security center
policy on the subscription, which will
remove the Security Center monitoring
extension from the Virtual Machines.
Compute VM with backup or snapshot extension These extensions are installed on a

Virtual Machine configured with the
Azure Backup service. While the
migration of these VMs is not
supported, follow the guidance here to
keep backups that were taken prior to
migration.
Network Virtual networks that contain virtual This is currently not supported. Please
machines and web/worker roles move the Web/Worker roles to their
own Virtual Network before migrating.
Once the classic Virtual Network is
migrated, the migrated Azure Resource
Manager Virtual Network can be
peered with the classic Virtual Network
to achieve similar configuration as
before.
Network Classic Express Route circuits This is currently not supported. These
circuits need to be migrated to Azure
Resource Manager before beginning
IaaS migration. To learn more, see
Moving ExpressRoute circuits from the
classic to the Resource Manager
deployment model.
Azure App Service Virtual networks that contain App This is currently not supported.
Service environments
Azure HDInsight Virtual networks that contain This is currently not supported.
HDInsight services
Microsoft Dynamics Lifecycle Services Virtual networks that contain virtual This is currently not supported.
machines that are managed by
Dynamics Lifecycle Services
Azure AD Domain Services Virtual networks that contain Azure AD This is currently not supported.
Domain services
Azure RemoteApp Virtual networks that contain Azure This is currently not supported.
RemoteApp deployments
Azure API Management Virtual networks that contain Azure API This is currently not supported. To
Management deployments migrate the IaaS VNET, change the
VNET of the API Management
deployment, which is a no downtime
operation.
Next steps
Technical deep dive on platform-supported migration from classic to Azure Resource Manager
Planning for migration of IaaS resources from classic to Azure Resource Manager
Use PowerShell to migrate IaaS resources from classic to Azure Resource Manager
Use CLI to migrate IaaS resources from classic to Azure Resource Manager
VPN Gateway classic to Resource Manager migration
Migrate ExpressRoute circuits and associated virtual networks from the classic to the Resource Manager
deployment model
Community tools for assisting with migration of IaaS resources from classic to Azure Resource Manager
Review most common migration errors
Review the most frequently asked questions about migrating IaaS resources from classic to Azure Resource
Manager
Technical deep dive on platform-supported
migration from classic to Azure Resource Manager
Let's take a deep-dive on migrating from the Azure classic deployment model to the Azure Resource Manager
deployment model. We look at resources at a resource and feature level to help you understand how the Azure
platform migrates resources between the two deployment models. For more information, please read the service
announcement article: Platform-supported migration of IaaS resources from classic to Azure Resource Manager.
Migrate IaaS resources from the classic deployment model to Azure

Resource Manager
First, it's important to understand the difference between data-plane and management-plane operations on the
infrastructure as a service (IaaS ) resources.
Management/control plane describes the calls that come into the management/control plane or the API for
modifying resources. For example, operations like creating a VM, restarting a VM, and updating a virtual
network with a new subnet manage the running resources. They don't directly affect connecting to the VMs.
Data plane (application) describes the runtime of the application itself, and involves interaction with instances
that don’t go through the Azure API. For example, accessing your website, or pulling data from a running SQL
Server instance or a MongoDB server, are data plane or application interactions. Other examples include
copying a blob from a storage account, and accessing a public IP address to use Remote Desktop Protocol
(RDP ) or Secure Shell (SSH) into the virtual machine. These operations keep the application running across
compute, networking, and storage.
The data plane is the same between the classic deployment model and Resource Manager stacks. The difference is
that during the migration process, Microsoft translates the representation of the resources from the classic
deployment model to that in the Resource Manager stack. As a result, you need to use new tools, APIs, and SDKs
to manage your resources in the Resource Manager stack.
NOTE
In some migration scenarios, the Azure platform stops, deallocates, and restarts your virtual machines. This causes a brief
data-plane downtime.
The migration experience

Before you start the migration:
Ensure that the resources that you want to migrate don't use any unsupported features or configurations.
Usually the platform detects these issues and generates an error.
If you have VMs that are not in a virtual network, they are stopped and deallocated as part of the prepare
operation. If you don't want to lose the public IP address, consider reserving the IP address before triggering
the prepare operation. If the VMs are in a virtual network, they are not stopped and deallocated.
Plan your migration during non-business hours to accommodate for any unexpected failures that might
happen during migration.
Download the current configuration of your VMs by using PowerShell, command-line interface (CLI)
commands, or REST APIs to make it easier for validation after the prepare step is complete.
Update your automation and operationalization scripts to handle the Resource Manager deployment model,
before you start the migration. You can optionally do GET operations when the resources are in the prepared
state.
Evaluate the Role-Based Access Control (RBAC ) policies that are configured on the IaaS resources in the
classic deployment model, and plan for after the migration is complete.
The migration workflow is as follows:
NOTE
The operations described in the following sections are all idempotent. If you have a problem other than an unsupported
feature or a configuration error, retry the prepare, abort, or commit operation. Azure tries the action again.
Validate
The validate operation is the first step in the migration process. The goal of this step is to analyze the state of the
resources you want to migrate in the classic deployment model. The operation evaluates whether the resources
are capable of migration (success or failure).
You select the virtual network or a cloud service (if it’s not in a virtual network) that you want to validate for
migration. If the resource is not capable of migration, Azure lists the reasons why.
Checks not done in the validate operation
The validate operation only analyzes the state of the resources in the classic deployment model. It can check for all
failures and unsupported scenarios due to various configurations in the classic deployment model. It is not
possible to check for all issues that the Azure Resource Manager stack might impose on the resources during
migration. These issues are only checked when the resources undergo transformation in the next step of
migration (the prepare operation). The following table lists all the issues not checked in the validate operation:
NETWORKING CHECKS NOT IN THE VALIDATE OPERATION
A virtual network having both ER and VPN gateways.
A virtual network gateway connection in a disconnected state.
All ER circuits are pre-migrated to Azure Resource Manager stack.
Azure Resource Manager quota checks for networking resources. For example: static public IP, dynamic public IPs, load balancer,
network security groups, route tables, and network interfaces.
All load balancer rules are valid across deployment and the virtual network.
Conflicting private IPs between stop-deallocated VMs in the same virtual network.
Prepare
The prepare operation is the second step in the migration process. The goal of this step is to simulate the
transformation of the IaaS resources from the classic deployment model to Resource Manager resources. Further,
the prepare operation presents this side-by-side for you to visualize.
NOTE
Your resources in the classic deployment model are not modified during this step. It's a safe step to run if you're trying out
migration.
You select the virtual network or the cloud service (if it’s not a virtual network) that you want to prepare for
migration.
If the resource is not capable of migration, Azure stops the migration process and lists the reason why the
prepare operation failed.
If the resource is capable of migration, Azure locks down the management-plane operations for the resources
under migration. For example, you are not able to add a data disk to a VM under migration.
Azure then starts the migration of metadata from the classic deployment model to Resource Manager for the
migrating resources.
After the prepare operation is complete, you have the option of visualizing the resources in both the classic
deployment model and Resource Manager. For every cloud service in the classic deployment model, the Azure
platform creates a resource group name that has the pattern cloud-service-name>-Migrated .
NOTE
It is not possible to select the name of a resource group created for migrated resources (that is, "-Migrated"). After
migration is complete, however, you can use the move feature of Azure Resource Manager to move resources to any
resource group you want. For more information, see Move resources to new resource group or subscription.
The following two screenshots show the result after a succesful prepare operation. The first one shows a resource
group that contains the original cloud service. The second one shows the new "-Migrated" resource group that
contains the equivalent Azure Resource Manager resources.
Here is a behind-the-scenes look at your resources after the completion of the prepare phase. Note that the
resource in the data plane is the same. It's represented in both the management plane (classic deployment model)
and the control plane (Resource Manager).
NOTE
VMs that are not in a virtual network in the classic deployment model are stopped and deallocated in this phase of
migration.
Check (manual or scripted)

In the check step, you have the option to use the configuration that you downloaded earlier to validate that the
migration looks correct. Alternatively, you can sign in to the portal, and spot check the properties and resources to
validate that metadata migration looks good.
If you are migrating a virtual network, most configuration of virtual machines is not restarted. For applications on
those VMs, you can validate that the application is still running.
You can test your monitoring and operational scripts to see if the VMs are working as expected, and if your
updated scripts work correctly. Only GET operations are supported when the resources are in the prepared state.
There is no set window of time before which you need to commit the migration. You can take as much time as you
want in this state. However, the management plane is locked for these resources until you either abort or commit.
If you see any issues, you can always abort the migration and go back to the classic deployment model. After you
go back, Azure opens the management-plane operations on the resources, so that you can resume normal
operations on those VMs in the classic deployment model.
Abort
This is an optional step if you want to revert your changes to the classic deployment model and stop the
migration. This operation deletes the Resource Manager metadata (created in the prepare step) for your
resources.
NOTE
This operation can't be done after you have triggered the commit operation.
Commit
After you finish the validation, you can commit the migration. Resources do not appear anymore in the classic
deployment model, and are available only in the Resource Manager deployment model. The migrated resources
can be managed only in the new portal.
NOTE
This is an idempotent operation. If it fails, retry the operation. If it continues to fail, create a support ticket or create a forum
post with a "ClassicIaaSMigration" tag on our VM forum.
Migration flowchart
Here is a flowchart that shows how to proceed with migration:
Translation of the classic deployment model to Resource Manager

resources
You can find the classic deployment model and Resource Manager representations of the resources in the
following table. Other features and resources are not currently supported.
CLASSIC REPRESENTATION RESOURCE MANAGER REPRESENTATION NOTES
Cloud service name DNS name During migration, a new resource

group is created for every cloud service
with the naming pattern
<cloudservicename>-migrated . This
resource group contains all your
resources. The cloud service name
becomes a DNS name that is associated
with the public IP address.
Virtual machine Virtual machine VM-specific properties are migrated

unchanged. Certain osProfile
information, like computer name, is not
stored in the classic deployment model,
and remains empty after migration.
Disk resources attached to VM Implicit disks attached to VM Disks are not modeled as top-level
resources in the Resource Manager
deployment model. They are migrated
as implicit disks under the VM. Only
disks that are attached to a VM are
currently supported. Resource Manager
VMs can now use storage accounts in
the classic deployment model, which
allows the disks to be easily migrated
without any updates.
VM extensions VM extensions All the resource extensions, except XML

extensions, are migrated from the
classic deployment model.
Virtual machine certificates Certificates in Azure Key Vault If a cloud service contains service
certificates, the migration creates a new
Azure key vault per cloud service, and
moves the certificates into the key
vault. The VMs are updated to
reference the certificates from the key
vault.
Do not delete the key vault. This can

cause the VM to go into a failed state.
WinRM configuration WinRM configuration under osProfile Windows Remote Management

configuration is moved unchanged, as
part of the migration.
Availability-set property Availability-set resource Availability-set specification is a

property on the VM in the classic
deployment model. Availability sets
become a top-level resource as part of
the migration. The following
configurations are not supported:
multiple availability sets per cloud
service, or one or more availability sets
along with VMs that are not in any
availability set in a cloud service.
Network configuration on a VM Primary network interface Network configuration on a VM is

represented as the primary network
interface resource after migration. For
VMs that are not in a virtual network,
the internal IP address changes during
migration.
Multiple network interfaces on a VM Network interfaces If a VM has multiple network interfaces

associated with it, each network
interface becomes a top-level resource
as part of the migration, along with all
the properties.
Load-balanced endpoint set Load balancer In the classic deployment model, the
platform assigned an implicit load
balancer for every cloud service. During
migration, a new load-balancer
resource is created, and the load-
balancing endpoint set becomes load-
balancer rules.
Inbound NAT rules Inbound NAT rules Input endpoints defined on the VM are
converted to inbound network address
translation rules under the load
balancer during the migration.
VIP address Public IP address with DNS name The virtual IP address becomes a public
IP address, and is associated with the
load balancer. A virtual IP can only be
migrated if there is an input endpoint
assigned to it.
Virtual network Virtual network The virtual network is migrated, with all
its properties, to the Resource Manager
deployment model. A new resource
group is created with the name
-migrated .
Reserved IPs Public IP address with static allocation Reserved IPs associated with the load
method balancer are migrated, along with the
migration of the cloud service or the
virtual machine. Unassociated reserved
IP migration is not currently supported.
Public IP address per VM Public IP address with dynamic The public IP address associated with
allocation method the VM is converted as a public IP
address resource, with the allocation
method set to static.
NSGs NSGs Network security groups associated

with a subnet are cloned as part of the
migration to the Resource Manager
deployment model. The NSG in the
classic deployment model is not
removed during the migration.
However, the management-plane
operations for the NSG are blocked
when the migration is in progress.
DNS servers DNS servers DNS servers associated with a virtual

network or the VM are migrated as
part of the corresponding resource
migration, along with all the properties.
UDRs UDRs User-defined routes associated with a

subnet are cloned as part of the
migration to the Resource Manager
deployment model. The UDR in the
classic deployment model is not
removed during the migration. The
management-plane operations for the
UDR are blocked when the migration is
in progress.
IP forwarding property on a VM's IP forwarding property on the NIC The IP forwarding property on a VM is
network configuration converted to a property on the
network interface during the migration.
Load balancer with multiple IPs Load balancer with multiple public IP Every public IP associated with the load
resources balancer is converted to a public IP
resource, and associated with the load
balancer after migration.
Internal DNS names on the VM Internal DNS names on the NIC During migration, the internal DNS
suffixes for the VMs are migrated to a
read-only property named
“InternalDomainNameSuffix” on the
NIC. The suffix remains unchanged after
migration, and VM resolution should
continue to work as previously.
Virtual network gateway Virtual network gateway Virtual network gateway properties are
migrated unchanged. The VIP
associated with the gateway does not
change either.
Local network site Local network gateway Local network site properties are
migrated unchanged to a new resource
called a local network gateway. This
represents on-premises address
prefixes and the remote gateway IP.
Connections references Connection Connectivity references between the

gateway and the local network site in
network configuration is represented by
a new resource called Connection. All
properties of connectivity reference in
network configuration files are copied
unchanged to the Connection resource.
Connectivity between virtual networks
in the classic deployment model is
achieved by creating two IPsec tunnels
to local network sites representing the
virtual networks. This is transformed to
the virtual-network-to-virtual-network
connection type in the Resource
Manager model, without requiring local
network gateways.
Changes to your automation and tooling after migration

As part of migrating your resources from the classic deployment model to the Resource Manager deployment
model, you must update your existing automation or tooling to ensure that it continues to work after the
migration.
Next steps
Overview of platform-supported migration of IaaS resources from classic to Azure Resource Manager
deployment model
Manager
Planning for migration of IaaS resources from classic
to Azure Resource Manager
While Azure Resource Manager offers many amazing features, it is critical to plan out your migration journey to
make sure things go smoothly. Spending time on planning will ensure that you do not encounter issues while
executing migration activities.
NOTE
The following guidance was heavily contributed to by the Azure Customer Advisory team and Cloud Solution architects
working with customers on migrating large environments. As such this document will continue to get updated as new
patterns of success emerge, so check back from time to time to see if there are any new recommendations.
There are four general phases of the migration journey:
Plan
Technical considerations and tradeoffs
Depending on your technical requirements size, geographies and operational practices, you might want to
consider:
1. Why is Azure Resource Manager desired for your organization? What are the business reasons for a
migration?
2. What are the technical reasons for Azure Resource Manager? What (if any) additional Azure services would
you like to leverage?
3. Which application (or sets of virtual machines) is included in the migration?
4. Which scenarios are supported with the migration API? Review the unsupported features and configurations.
5. Will your operational teams now support applications/VMs in both Classic and Azure Resource Manager?
6. How (if at all) does Azure Resource Manager change your VM deployment, management, monitoring, and
reporting processes? Do your deployment scripts need to be updated?
7. What is the communications plan to alert stakeholders (end users, application owners, and infrastructure
owners)?
8. Depending on the complexity of the environment, should there be a maintenance period where the application
is unavailable to end users and to application owners? If so, for how long?
9. What is the training plan to ensure stakeholders are knowledgeable and proficient in Azure Resource
Manager?
10. What is the program management or project management plan for the migration?
11. What are the timelines for the Azure Resource Manager migration and other related technology road maps?
Are they optimally aligned?
Patterns of success
Successful customers have detailed plans where the preceding questions are discussed, documented and
governed. Ensure the migration plans are broadly communicated to sponsors and stakeholders. Equip yourself
with knowledge about your migration options; reading through this migration document set below is highly
recommended.
Manager
Pitfalls to avoid
Failure to plan. The technology steps of this migration are proven and the outcome is predictable.
Assumption that the platform supported migration API will account for all scenarios. Read the unsupported
features and configurations to understand what scenarios are supported.
Not planning potential application outage for end users. Plan enough buffer to adequately warn end users of
potentially unavailable application time.
Lab Test
Replicate your environment and do a test migration
NOTE
Exact replication of your existing environment is executed by using a community-contributed tool which is not officially
supported by Microsoft Support. Therefore, it is an optional step but it is the best way to find out issues without touching
your production environments. If using a community-contributed tool is not an option, then read about the
Validate/Prepare/Abort Dry Run recommendation below.
Conducting a lab test of your exact scenario (compute, networking, and storage) is the best way to ensure a
smooth migration. This will help ensure:
A wholly separate lab or an existing non-production environment to test. We recommend a wholly separate lab
that can be migrated repeatedly and can be destructively modified. Scripts to collect/hydrate metadata from
the real subscriptions are listed below.
It's a good idea to create the lab in a separate subscription. The reason is that the lab will be torn down
repeatedly, and having a separate, isolated subscription will reduce the chance that something real will get
accidentally deleted.
This can be accomplished by using the AsmMetadataParser tool. Read more about this tool here
Patterns of success
The following were issues discovered in many of the larger migrations. This is not an exhaustive list and you
should refer to the unsupported features and configurations for more detail. You may or may not encounter these
technical issues but if you do solving these before attempting migration will ensure a smoother experience.
Do a Validate/Prepare/Abort Dry Run - This is perhaps the most important step to ensure Classic to
Azure Resource Manager migration success. The migration API has three main steps: Validate, Prepare and
Commit. Validate will read the state of your classic environment and return a result of all issues. However,
because some issues might exist in the Azure Resource Manager stack, Validate will not catch everything.
The next step in migration process, Prepare will help expose those issues. Prepare will move the metadata
from Classic to Azure Resource Manager, but will not commit the move, and will not remove or change
anything on the Classic side. The dry run involves preparing the migration, then aborting (not
committing) the migration prepare. The goal of validate/prepare/abort dry run is to see all of the
metadata in the Azure Resource Manager stack, examine it (programmatically or in Portal), and verify that
everything migrates correctly, and work through technical issues. It will also give you a sense of migration
duration so you can plan for downtime accordingly. A validate/prepare/abort does not cause any user
downtime; therefore, it is non-disruptive to application usage.
The items below will need to be solved before the dry run, but a dry run test will also safely flush out
these preparation steps if they are missed. During enterprise migration, we've found the dry run to be a
safe and invaluable way to ensure migration readiness.
When prepare is running, the control plane (Azure management operations) will be locked for the whole
virtual network, so no changes can be made to VM metadata during validate/prepare/abort. But
otherwise any application function (RD, VM usage, etc.) will be unaffected. Users of the VMs will not
know that the dry run is being executed.
Express Route Circuits and VPN. Currently Express Route Gateways with authorization links cannot be
migrated without downtime. For the workaround, see Migrate ExpressRoute circuits and associated virtual
networks from the classic to the Resource Manager deployment model.
VM Extensions - Virtual Machine extensions are potentially one of the biggest roadblocks to migrating
running VMs. Remediation of VM Extensions could take upwards of 1-2 days, so plan accordingly. A
working Azure agent is needed to report back VM Extension status of running VMs. If the status comes
back as bad for a running VM, this will halt migration. The agent itself does not need to be in working order
to enable migration, but if extensions exist on the VM, then both a working agent AND outbound internet
connectivity (with DNS ) will be needed for migration to move forward.
If connectivity to a DNS server is lost during migration, all VM Extensions except BGInfo version 1.*
need to first be removed from every VM before migration prepare, and subsequently re-added back to
the VM after Azure Resource Manager migration. This is only for VMs that are running. If the VMs
are stopped deallocated, VM Extensions do not need to be removed.
NOTE
Many extensions like Azure diagnostics and security center monitoring will reinstall themselves after migration, so
removing them is not a problem.
In addition, make sure Network Security Groups are not restricting outbound internet access. This can
happen with some Network Security Groups configurations. Outbound internet access (and DNS ) is
needed for VM Extensions to be migrated to Azure Resource Manager.
Two versions of the BGInfo extension exist and are called versions 1 and 2.
If the VM is using the BGInfo version 1 extension, you can leave this extension as is. The
migration API skips this extension. The BGInfo extension can be added after migration.
If the VM is using the JSON -based BGInfo version 2 extension, the VM was created using the
Azure portal. The migration API includes this extension in the migration to Azure Resource
Manager, provided the agent is working and has outbound internet access (and DNS ).
Remediation Option 1. If you know your VMs will not have outbound internet access, a working
DNS service, and working Azure agents on the VMs, then uninstall all VM extensions as part of the
migration before Prepare, then reinstall the VM Extensions after migration.
Remediation Option 2. If VM extensions are too big of a hurdle, another option is to
shutdown/deallocate all VMs before migration. Migrate the deallocated VMs, then restart them on
the Azure Resource Manager side. The benefit here is that VM extensions will migrate. The
downside is that all public facing Virtual IPs will be lost (this may be a non-starter), and obviously
the VMs will shut down causing a much greater impact on working applications.
NOTE
If an Azure Security Center policy is configured against the running VMs being migrated, the security policy
needs to be stopped before removing extensions, otherwise the security monitoring extension will be
reinstalled automatically on the VM after removing it.
Availability Sets - For a virtual network (vNet) to be migrated to Azure Resource Manager, the Classic
deployment (i.e. cloud service) contained VMs must all be in one availability set, or the VMs must all not be
in any availability set. Having more than one availability set in the cloud service is not compatible with
Azure Resource Manager and will halt migration. Additionally, there cannot be some VMs in an availability
set, and some VMs not in an availability set. To resolve this, you will need to remediate or reshuffle your
cloud service. Plan accordingly as this might be time consuming.
Web/Worker Role Deployments - Cloud Services containing web and worker roles cannot migrate to
Azure Resource Manager. The web/worker roles must first be removed from the virtual network before
migration can start. A typical solution is to just move web/worker role instances to a separate Classic
virtual network that is also linked to an ExpressRoute circuit, or to migrate the code to newer PaaS App
Services (this discussion is beyond the scope of this document). In the former redeploy case, create a new
Classic virtual network, move/redeploy the web/worker roles to that new virtual network, then delete the
deployments from the virtual network being moved. No code changes required. The new Virtual Network
Peering capability can be used to peer together the classic virtual network containing the web/worker roles
and other virtual networks in the same Azure region such as the virtual network being migrated (after
virtual network migration is completed as peered virtual networks cannot be migrated), hence
providing the same capabilities with no performance loss and no latency/bandwidth penalties. Given the
addition of Virtual Network Peering, web/worker role deployments can now easily be mitigated and not
block the migration to Azure Resource Manager.
Azure Resource Manager Quotas - Azure regions have separate quotas/limits for both Classic and Azure
Resource Manager. Even though in a migration scenario new hardware isn't being consumed (we're
swapping existing VMs from Classic to Azure Resource Manager ), Azure Resource Manager quotas still
need to be in place with enough capacity before migration can start. Listed below are the major limits we've
seen cause problems. Open a quota support ticket to raise the limits.
NOTE
These limits need to be raised in the same region as your current environment to be migrated.
Network Interfaces
Load Balancers
Public IPs
Static Public IPs
Cores
Network Security Groups
Route Tables
You can check your current Azure Resource Manager quotas using the following commands with the
latest version of Azure PowerShell.
Compute (Cores, Availability Sets)
Get-AzureRmVMUsage -Location <azure-region>
Network (Virtual Networks, Static Public IPs, Public IPs, Network Security Groups, Network
Interfaces, Load Balancers, Route Tables)
Get-AzureRmUsage /subscriptions/<subscription-id>/providers/Microsoft.Network/locations/<azure-
region> -ApiVersion 2016-03-30 | Format-Table
Storage (Storage Account)
Get-AzureRmStorageUsage
Azure Resource Manager API throttling limits - If you have a large enough environment (eg. > 400
VMs in a VNET), you might hit the default API throttling limits for writes (currently 1200 writes/hour ) in
Azure Resource Manager. Before starting migration, you should raise a support ticket to increase this limit
for your subscription.
Provisioning Timed Out VM Status - If any VM has the status of provisioning timed out , this needs to
be resolved pre-migration. The only way to do this is with downtime by deprovisioning/reprovisioning the
VM (delete it, keep the disk, and recreate the VM ).
RoleStateUnknown VM Status - If migration halts due to a role state unknown error message, inspect
the VM using the portal and ensure it is running. This error will typically go away on its own (no
remediation required) after a few minutes and is often a transient type often seen during a Virtual Machine
start , stop , restart operations. Recommended practice: re-try migration again after a few minutes.
Fabric Cluster does not exist - In some cases, certain VMs cannot be migrated for various odd reasons.
One of these known cases is if the VM was recently created (within the last week or so) and happened to
land an Azure cluster that is not yet equipped for Azure Resource Manager workloads. You will get an error
that says fabric cluster does not exist and the VM cannot be migrated. Waiting a couple of days will
usually resolve this particular problem as the cluster will soon get Azure Resource Manager enabled.
However, one immediate workaround is to stop-deallocate the VM, then continue forward with
migration, and start the VM back up in Azure Resource Manager after migrating.
Pitfalls to avoid
Do not take shortcuts and omit the validate/prepare/abort dry run migrations.
Most, if not all, of your potential issues will surface during the validate/prepare/abort steps.
Migration
Now you are ready because you have worked through the known issues with your environment.
For the real migrations, you might want to consider:
1. Plan and schedule the virtual network (smallest unit of migration) with increasing priority. Do the simple
virtual networks first, and progress with the more complicated virtual networks.
2. Most customers will have non-production and production environments. Schedule production last.
3. (OPTIONAL ) Schedule a maintenance downtime with plenty of buffer in case unexpected issues arise.
4. Communicate with and align with your support teams in case issues arise.
Patterns of success
The technical guidance from the Lab Test section should be considered and mitigated prior to a real migration.
With adequate testing, the migration is actually a non-event. For production environments, it might be helpful to
have additional support, such as a trusted Microsoft partner or Microsoft Premier services.
Pitfalls to avoid
Not fully testing may cause issues and delay in the migration.
Beyond Migration
Now that you are in Azure Resource Manager, maximize the platform. Read the overview of Azure Resource
Manager to find out about additional benefits.
Things to consider:
Bundling the migration with other activities. Most customers opt for an application maintenance window. If so,
you might want to use this downtime to enable other Azure Resource Manager capabilities like encryption and
migration to Managed Disks.
Revisit the technical and business reasons for Azure Resource Manager; enable the additional services
available only on Azure Resource Manager that apply to your environment.
Modernize your environment with PaaS services.
Patterns of success
Be purposeful on what services you now want to enable in Azure Resource Manager. Many customers find the
below compelling for their Azure environments:
Role Based Access Control.
Azure Resource Manager templates for easier and more controlled deployment.
Tags.
Activity Control
Azure Policies
Pitfalls to avoid
Remember why you started this Classic to Azure Resource Manager migration journey. What were the original
business reasons? Did you achieve the business reason?
Next steps
deployment model
Manager
Migrate IaaS resources from classic to Azure
Resource Manager by using Azure PowerShell
These steps show you how to use Azure PowerShell commands to migrate infrastructure as a service (IaaS )
resources from the classic deployment model to the Azure Resource Manager deployment model.
If you want, you can also migrate resources by using the Azure Command Line Interface (Azure CLI).
For background on supported migration scenarios, see Platform-supported migration of IaaS resources from
classic to Azure Resource Manager.
For detailed guidance and a migration walkthrough, see Technical deep dive on platform-supported migration
from classic to Azure Resource Manager.
Here is a flowchart to identify the order in which steps need to be executed during a migration process
Step 1: Plan for migration

Here are a few best practices that we recommend as you evaluate migrating IaaS resources from classic to
Resource Manager:
Read through the supported and unsupported features and configurations. If you have virtual machines that
use unsupported configurations or features, we recommend that you wait for the configuration/feature
support to be announced. Alternatively, if it suits your needs, remove that feature or move out of that
configuration to enable migration.
If you have automated scripts that deploy your infrastructure and applications today, try to create a similar test
setup by using those scripts for migration. Alternatively, you can set up sample environments by using the
Azure portal.
IMPORTANT
Application Gateways are not currently supported for migration from classic to Resource Manager. To migrate a classic
virtual network with an Application gateway, remove the gateway before running a Prepare operation to move the
network. After you complete the migration, reconnect the gateway in Azure Resource Manager.
ExpressRoute gateways connecting to ExpressRoute circuits in another subscription cannot be migrated automatically. In
such cases, remove the ExpressRoute gateway, migrate the virtual network and recreate the gateway. Please see Migrate
ExpressRoute circuits and associated virtual networks from the classic to the Resource Manager deployment model for
more information.
Step 2: Install the latest version of Azure PowerShell

There are two main options to install Azure PowerShell: PowerShell Gallery or Web Platform Installer (WebPI).
WebPI receives monthly updates. PowerShell Gallery receives updates on a continuous basis. This article is based
on Azure PowerShell version 2.1.0.
For installation instructions, see How to install and configure Azure PowerShell.
Step 3: Ensure that you are an administrator for the subscription in

Azure portal
To perform this migration, you must be added as a co-administrator for the subscription in the Azure portal.
1. Sign into the Azure portal.
2. On the Hub menu, select Subscription. If you don't see it, select All services.
3. Find the appropriate subscription entry, then look at the MY ROLE field. For a co-administrator, the value
should be Account admin.
If you are not able to add a co-administrator, then contact a service administrator or co-administrator for the
subscription to get yourself added.
Step 4: Set your subscription and sign up for migration

First, start a PowerShell prompt. For migration, you need to set up your environment for both classic and
Resource Manager.
Sign in to your account for the Resource Manager model.
Get the available subscriptions by using the following command:
Get-AzureRMSubscription | Sort Name | Select Name
Set your Azure subscription for the current session. This example sets the default subscription name to My
Azure Subscription. Replace the example subscription name with your own.
Select-AzureRmSubscription –SubscriptionName "My Azure Subscription"

NOTE
Registration is a one-time step, but you must do it once before attempting migration. Without registering, you see the
following error message:
BadRequest : Subscription is not registered for migration.
Register with the migration resource provider by using the following command:
Register-AzureRmResourceProvider -ProviderNamespace Microsoft.ClassicInfrastructureMigrate
Please wait five minutes for the registration to finish. You can check the status of the approval by using the
following command:
Get-AzureRmResourceProvider -ProviderNamespace Microsoft.ClassicInfrastructureMigrate
Make sure that RegistrationState is Registered before you proceed.

Now sign in to your account for the classic model.
Add-AzureAccount
Get the available subscriptions by using the following command:
Get-AzureSubscription | Sort SubscriptionName | Select SubscriptionName
Set your Azure subscription for the current session. This example sets the default subscription to My Azure
Subscription. Replace the example subscription name with your own.
Select-AzureSubscription –SubscriptionName "My Azure Subscription"
Step 5: Make sure you have enough Azure Resource Manager Virtual
Machine vCPUs in the Azure region of your current deployment or
VNET
You can use the following PowerShell command to check the current number of vCPUs you have in Azure
Resource Manager. To learn more about vCPU quotas, see Limits and the Azure Resource Manager.
This example checks the availability in the West US region. Replace the example region name with your own.
Get-AzureRmVMUsage -Location "West US"
Step 6: Run commands to migrate your IaaS resources

Migrate VMs in a cloud service (not in virtual network)
Migrate VMs in a virtual network
Migrate Storage Account
NOTE
All the operations described here are idempotent. If you have a problem other than an unsupported feature or a
configuration error, we recommend that you retry the prepare, abort, or commit operation. The platform then tries the
action again.
Step 6.1: Option 1 - Migrate virtual machines in a cloud service (not in a virtual network)
Get the list of cloud services by using the following command, and then pick the cloud service that you want to
migrate. If the VMs in the cloud service are in a virtual network or if they have web or worker roles, the command
returns an error message.
Get-AzureService | ft Servicename
Get the deployment name for the cloud service. In this example, the service name is My Service. Replace the
example service name with your own service name.
$serviceName = "My Service"

$deployment = Get-AzureDeployment -ServiceName $serviceName
$deploymentName = $deployment.DeploymentName
Prepare the virtual machines in the cloud service for migration. You have two options to choose from.
Option 1. Migrate the VMs to a platform -created virtual network
First, validate if you can migrate the cloud service using the following commands:
$validate = Move-AzureService -Validate -ServiceName $serviceName `

-DeploymentName $deploymentName -CreateNewVirtualNetwork
$validate.ValidationMessages
The following command displays any warnings and errors that block migration. If validation is successful,
then you can move on to the Prepare step:
Move-AzureService -Prepare -ServiceName $serviceName `

-DeploymentName $deploymentName -CreateNewVirtualNetwork
Option 2. Migrate to an existing virtual network in the Resource Manager deployment model
This example sets the resource group name to myResourceGroup, the virtual network name to
myVirtualNetwork and the subnet name to mySubNet. Replace the names in the example with the
names of your own resources.
$existingVnetRGName = "myResourceGroup"
$vnetName = "myVirtualNetwork"
$subnetName = "mySubNet"
First, validate if you can migrate the virtual network using the following command:
$validate = Move-AzureService -Validate -ServiceName $serviceName `

-DeploymentName $deploymentName -UseExistingVirtualNetwork -VirtualNetworkResourceGroupName
$existingVnetRGName -VirtualNetworkName $vnetName -SubnetName $subnetName
$validate.ValidationMessages
The following command displays any warnings and errors that block migration. If validation is successful,
then you can proceed with the following Prepare step:
Move-AzureService -Prepare -ServiceName $serviceName -DeploymentName $deploymentName `

-UseExistingVirtualNetwork -VirtualNetworkResourceGroupName $existingVnetRGName `
-VirtualNetworkName $vnetName -SubnetName $subnetName
After the Prepare operation succeeds with either of the preceding options, query the migration state of the VMs.
Ensure that they are in the Prepared state.
This example sets the VM name to myVM. Replace the example name with your own VM name.
$vmName = "myVM"
$vm = Get-AzureVM -ServiceName $serviceName -Name $vmName
$vm.VM.MigrationState
Check the configuration for the prepared resources by using either PowerShell or the Azure portal. If you are not
ready for migration and you want to go back to the old state, use the following command:
Move-AzureService -Abort -ServiceName $serviceName -DeploymentName $deploymentName
If the prepared configuration looks good, you can move forward and commit the resources by using the
following command:
Move-AzureService -Commit -ServiceName $serviceName -DeploymentName $deploymentName
Step 6.1: Option 2 - Migrate virtual machines in a virtual network

To migrate virtual machines in a virtual network, you migrate the virtual network. The virtual machines
automatically migrate with the virtual network. Pick the virtual network that you want to migrate.
NOTE
Migrate single classic virtual machine by creating a new Resource Manager virtual machine with Managed Disks using the
VHD (OS and data) files of the virtual machine.
NOTE
The virtual network name might be different from what is shown in the new Portal. The new Azure Portal displays the name
as [vnet-name] but the actual virtual network name is of type Group [resource-group-name] [vnet-name] . Before
migrating, lookup the actual virtual network name using the command Get-AzureVnetSite | Select -Property Name or
view it in the old Azure Portal.
This example sets the virtual network name to myVnet. Replace the example virtual network name with your
own.
$vnetName = "myVnet"
NOTE
If the virtual network contains web or worker roles, or VMs with unsupported configurations, you get a validation error
message.
First, validate if you can migrate the virtual network by using the following command:
Move-AzureVirtualNetwork -Validate -VirtualNetworkName $vnetName
The following command displays any warnings and errors that block migration. If validation is successful, then
you can proceed with the following Prepare step:
Move-AzureVirtualNetwork -Prepare -VirtualNetworkName $vnetName
Check the configuration for the prepared virtual machines by using either Azure PowerShell or the Azure portal.
If you are not ready for migration and you want to go back to the old state, use the following command:
Move-AzureVirtualNetwork -Abort -VirtualNetworkName $vnetName
following command:
Move-AzureVirtualNetwork -Commit -VirtualNetworkName $vnetName
Step 6.2 Migrate a storage account

Once you're done migrating the virtual machines, we recommend you perform the following prerequisite checks
before you migrate the storage accounts.
NOTE
If your storage account had no associated disks or VM data, you can skip directly to the Validate Storage Account and
Start Migration section.
Prerequisite checks if you migrated any VMs or your storage account has Disk resources
Migrate classic virtual machines whose disks are stored in the storage account
The following command returns RoleName and DiskName properties of all the classic VM disks in
the storage account. RoleName is the name of the virtual machine to which a disk is attached. If this
command returns disks then ensure that virtual machines to which these disks are attached are
migrated before migrating the storage account.
$storageAccountName = 'yourStorageAccountName'
Get-AzureDisk | where-Object {$_.MediaLink.Host.Contains($storageAccountName)} | Select-Object
-ExpandProperty AttachedTo -Property `
DiskName | Format-List -Property RoleName, DiskName
Delete unattached classic VM disks stored in the storage account

Find unattached classic VM disks in the storage account using following command:
$storageAccountName = 'yourStorageAccountName'
Get-AzureDisk | where-Object {$_.MediaLink.Host.Contains($storageAccountName)} | Where-
Object -Property AttachedTo -EQ $null | Format-List -Property DiskName
If above command returns disks then delete these disks using following command:
Remove-AzureDisk -DiskName 'yourDiskName'
Delete VM images stored in the storage account

The following command returns all the VM images with OS disk stored in the storage account.
Get-AzureVmImage | Where-Object { $_.OSDiskConfiguration.MediaLink -ne $null -and

$_.OSDiskConfiguration.MediaLink.Host.Contains($storageAccountName)`
} | Select-Object -Property ImageName, ImageLabel
The following command returns all the VM images with data disks stored in the storage account.
Get-AzureVmImage | Where-Object {$_.DataDiskConfigurations -ne $null `

-and ($_.DataDiskConfigurations | Where-Object {$_.MediaLink
-ne $null -and $_.MediaLink.Host.Contains($storageAccountName)}).Count -gt 0 `
} | Select-Object -Property ImageName, ImageLabel
Delete all the VM images returned by above commands using this command:
Remove-AzureVMImage -ImageName 'yourImageName'
Validate Storage Account and Start Migration

Validate each storage account for migration by using the following command. In this example, the storage
account name is myStorageAccount. Replace the example name with the name of your own storage
account.
$storageAccountName = "myStorageAccount"
Move-AzureStorageAccount -Validate -StorageAccountName $storageAccountName
Next step is to prepare the storage account for migration
$storageAccountName = "myStorageAccount"
Move-AzureStorageAccount -Prepare -StorageAccountName $storageAccountName
Check the configuration for the prepared storage account by using either Azure PowerShell or the Azure
portal. If you are not ready for migration and you want to go back to the old state, use the following
command:
Move-AzureStorageAccount -Abort -StorageAccountName $storageAccountName
following command:
Move-AzureStorageAccount -Commit -StorageAccountName $storageAccountName
Next steps
Manager
Common errors during Classic to Azure Resource
Manager migration
This article catalogs the most common errors and mitigations during the migration of IaaS resources from Azure
classic deployment model to the Azure Resource Manager stack.
List of errors
ERROR STRING MITIGATION
Internal server error In some cases, this is a transient error that goes away with a
retry. If it continues to persist, contact Azure support as it
needs investigation of platform logs.
NOTE: Once the incident is tracked by the support team,

please do not attempt any self-mitigation as this might have
unintended consequences on your environment.
Migration is not supported for Deployment {deployment- This happens when a deployment contains a web/worker role.
name} in HostedService {hosted-service-name} because it is a Since migration is only supported for Virtual Machines, please
PaaS deployment (Web/Worker). remove the web/worker role from the deployment and try
migration again.
Template {template-name} deployment failed. CorrelationId= In the backend of migration service, we use Azure Resource
{guid} Manager templates to create resources in the Azure Resource
Manager stack. Since templates are idempotent, usually you
can safely retry the migration operation to get past this error.
If this error continues to persist, please contact Azure
support and give them the CorrelationId.
NOTE: Once the incident is tracked by the support team,

please do not attempt any self-mitigation as this might have
unintended consequences on your environment.
The virtual network {virtual-network-name} does not exist. This can happen if you created the Virtual Network in the
new Azure portal. The actual Virtual Network name follows
the pattern "Group * "
VM {vm-name} in HostedService {hosted-service-name} XML extensions such as BGInfo 1.* are not supported in
contains Extension {extension-name} which is not supported Azure Resource Manager. Therefore, these extensions cannot
in Azure Resource Manager. It is recommended to uninstall it be migrated. If these extensions are left installed on the
from the VM before continuing with migration. virtual machine, they are automatically uninstalled before
completing the migration.
VM {vm-name} in HostedService {hosted-service-name} This is the scenario where the virtual machine is configured
contains Extension VMSnapshot/VMSnapshotLinux, which is for Azure Backup. Since this is currently an unsupported
currently not supported for Migration. Uninstall it from the scenario, please follow the workaround at
VM and add it back using Azure Resource Manager after the https://aka.ms/vmbackupmigration
Migration is Complete
VM {vm-name} in HostedService {hosted-service-name} Azure guest agent & VM Extensions need outbound internet
contains Extension {extension-name} whose Status is not access to the VM storage account to populate their status.
being reported from the VM. Hence, this VM cannot be Common causes of status failure include
migrated. Ensure that the Extension status is being reported a Network Security Group that blocks outbound access to
or uninstall the extension from the VM and retry migration. the internet
If the VNET has on-prem DNS servers and DNS
VM {vm-name} in HostedService {hosted-service-name} connectivity is lost
contains Extension {extension-name} reporting Handler
Status: {handler-status}. Hence, the VM cannot be migrated. If you continue to see an unsupported status, you can
Ensure that the Extension handler status being reported is uninstall the extensions to skip this check and move forward
{handler-status} or uninstall it from the VM and retry with migration.
migration.
VM Agent for VM {vm-name} in HostedService {hosted-

service-name} is reporting the overall agent status as Not
Ready. Hence, the VM may not be migrated, if it has a
migratable extension. Ensure that the VM Agent is reporting
overall agent status as Ready. Refer to
https://aka.ms/classiciaasmigrationfaqs.
Migration is not supported for Deployment {deployment- Currently, only hosted services that have 1 or less Availability
name} in HostedService {hosted-service-name} because it has sets can be migrated. To work around this problem, please
multiple Availability Sets. move the additional Availability sets and Virtual machines in
those Availability sets to a different hosted service.
Migration is not supported for Deployment {deployment- The workaround for this scenario is to either move all the
name} in HostedService {hosted-service-name because it has virtual machines into a single Availability set or remove all
VMs that are not part of the Availability Set even though the Virtual machines from the Availability set in the hosted
HostedService contains one. service.
Storage account/HostedService/Virtual Network {virtual- This error happens when the "Prepare" migration operation
network-name} is in the process of being migrated and hence has been completed on the resource and an operation that
cannot be changed would make a change to the resource is triggered. Because of
the lock on the management plane after "Prepare" operation,
any changes to the resource are blocked. To unlock the
management plane, you can run the "Commit" migration
operation to complete migration or the "Abort" migration
operation to roll back the "Prepare" operation.
Migration is not allowed for HostedService {hosted-service- The VM might be undergoing through a state transition,
name} because it has VM {vm-name} in State: which usually happens when during an update operation on
RoleStateUnknown. Migration is allowed only when the VM is the HostedService such as a reboot, extension installation etc.
in one of the following states - Running, Stopped, Stopped It is recommended for the update operation to complete on
Deallocated. the HostedService before trying migration.
Deployment {deployment-name} in HostedService {hosted- This error happens if you've resized the VHD blob without
service-name} contains a VM {vm-name} with Data Disk updating the size in the VM API model. Detailed mitigation
{data-disk-name} whose physical blob size {size-of-the-vhd- steps are outlined below.
blob-backing-the-data-disk} bytes does not match the VM
Data Disk logical size {size-of-the-data-disk-specified-in-the-
vm-api} bytes. Migration will proceed without specifying a
size for the data disk for the Azure Resource Manager VM.
A storage exception occurred while validating data disk {data This error can happen if the disks of the VM have been
disk name} with media link {data disk Uri} for VM {VM name} deleted or are not accessible anymore. Please make sure the
in Cloud Service {Cloud Service name}. Please ensure that the disks for the VM exist.
VHD media link is accessible for this virtual machine
VM {vm-name} in HostedService {cloud-service-name} This error occurs when the name of the blob has a "/" in it
contains Disk with MediaLink {vhd-uri} which has blob name which is not supported in Compute Resource Provider
{vhd-blob-name} that is not supported in Azure Resource currently.
Manager.
Migration is not allowed for Deployment {deployment-name} In 2014, Azure announced that networking resources will
in HostedService {cloud-service-name} as it is not in the move from a cluster level scope to regional scope. See
regional scope. Please refer to http://aka.ms/regionalscope for http://aka.ms/regionalscope for more details. This error
moving this deployment to regional scope. happens when the deployment being migrated has not had
an update operation, which automatically moves it to a
regional scope. Best workaround is to either add an endpoint
to a VM or a data disk to the VM and then retry migration.
See How to set up endpoints on a classic Windows virtual
machine in Azure or Attach a data disk to a Windows virtual
machine created with the classic deployment model
Migration is not supported for Virtual Network {vnet-name} This error occurs when you have non-gateway PaaS
because it has non-gateway PaaS deployments. deployments such as Application Gateway or API
Management services that are connected to the Virtual
Network.
Detailed mitigations
VM with Data Disk whose physical blob size bytes does not match the VM Data Disk logical size bytes.
This happens when the Data disk logical size can get out of sync with the actual VHD blob size. This can be easily
verified using the following commands:
Verifying the issue
# Store the VM details in the VM object
$vm = Get-AzureVM -ServiceName $servicename -Name $vmname
# Display the data disk properties

# NOTE the data disk LogicalDiskSizeInGB below which is 11GB. Also note the MediaLink Uri of the VHD blob as
we'll use this in the next step
$vm.VM.DataVirtualHardDisks
HostCaching : None
DiskLabel :
DiskName : coreosvm-coreosvm-0-201611230636240687
Lun : 0
LogicalDiskSizeInGB : 11
MediaLink : https://contosostorage.blob.core.windows.net/vhds/coreosvm-dd1.vhd
SourceMediaLink :
IOType : Standard
ExtensionData :
# Now get the properties of the blob backing the data disk above
# NOTE the size of the blob is about 15 GB which is different from LogicalDiskSizeInGB above
$blob = Get-AzureStorageblob -Blob "coreosvm-dd1.vhd" -Container vhds
$blob
ICloudBlob : Microsoft.WindowsAzure.Storage.Blob.CloudPageBlob
BlobType : PageBlob
Length : 16106127872
ContentType : application/octet-stream
LastModified : 11/23/2016 7:16:22 AM +00:00
SnapshotTime :
ContinuationToken :
Context : Microsoft.WindowsAzure.Commands.Common.Storage.AzureStorageContext
Name : coreosvm-dd1.vhd
Mitigating the issue
# Convert the blob size in bytes to GB into a variable which we'll use later
$newSize = [int]($blob.Length / 1GB)
# See the calculated size in GB

$newSize
15
# Store the disk name of the data disk as we'll use this to identify the disk to be updated
$diskName = $vm.VM.DataVirtualHardDisks[0].DiskName
# Identify the LUN of the data disk to remove

$lunToRemove = $vm.VM.DataVirtualHardDisks[0].Lun
# Now remove the data disk from the VM so that the disk isn't leased by the VM and it's size can be updated
Remove-AzureDataDisk -LUN $lunToRemove -VM $vm | Update-AzureVm -Name $vmname -ServiceName $servicename
OperationDescription OperationId OperationStatus

-------------------- ----------- ---------------
Update-AzureVM 213xx1-b44b-1v6n-23gg-591f2a13cd16 Succeeded
# Verify we have the right disk that's going to be updated

Get-AzureDisk -DiskName $diskName
AffinityGroup :
AttachedTo :
IsCorrupted : False
Label :
Location : East US
DiskSizeInGB : 11
DiskSizeInGB : 11
SourceImageName :
OS :
IOType : Standard
OperationDescription : Get-AzureDisk
OperationId : 0c56a2b7-a325-123b-7043-74c27d5a61fd
OperationStatus : Succeeded
# Now update the disk to the new size

Update-AzureDisk -DiskName $diskName -ResizedSizeInGB $newSize -Label $diskName

-------------------- ----------- ---------------
Update-AzureDisk cv134b65-1b6n-8908-abuo-ce9e395ac3e7 Succeeded
# Now verify that the "DiskSizeInGB" property of the disk matches the size of the blob
Get-AzureDisk -DiskName $diskName
AffinityGroup :
AttachedTo :
IsCorrupted : False
Label : coreosvm-coreosvm-0-201611230636240687
Location : East US
DiskSizeInGB : 15
SourceImageName :
OS :
IOType : Standard
OperationDescription : Get-AzureDisk
OperationId : 1v53bde5-cv56-5621-9078-16b9c8a0bad2
OperationStatus : Succeeded
# Now we'll add the disk back to the VM as a data disk. First we need to get an updated VM object
$vm = Get-AzureVM -ServiceName $servicename -Name $vmname
Add-AzureDataDisk -Import -DiskName $diskName -LUN 0 -VM $vm -HostCaching ReadWrite | Update-AzureVm -Name
$vmname -ServiceName $servicename

-------------------- ----------- ---------------
Update-AzureVM b0ad3d4c-4v68-45vb-xxc1-134fd010d0f8 Succeeded
Moving a VM to a different subscription after completing migration

After you complete the migration process, you may want to move the VM to another subscription. However, if
you have a secret/certificate on the VM that references a Key Vault resource, the move is currently not supported.
The below instructions will allow you to workaround the issue.
PowerShell
$vm = Get-AzureRmVM -ResourceGroupName "MyRG" -Name "MyVM"

Remove-AzureRmVMSecret -VM $vm
Update-AzureRmVM -ResourceGroupName "MyRG" -VM $vm
Azure CLI
az vm update -g "myrg" -n "myvm" --set osProfile.Secrets=[]
Next steps
Manager
Community tools to migrate IaaS resources from
classic to Azure Resource Manager
This article catalogs the tools that have been provided by the community to assist with migration of IaaS
resources from classic to the Azure Resource Manager deployment model.
NOTE
These tools are not officially supported by Microsoft Support. Therefore they are open sourced on GitHub and we're happy
to accept PRs for fixes or additional scenarios. To report an issue, use the GitHub issues feature.
Migrating with these tools will cause downtime for your classic Virtual Machine. If you're looking for platform supported
migration, visit
Platform supported migration of IaaS resources from Classic to Azure Resource Manager stack
Technical Deep Dive on Platform supported migration from Classic to Azure Resource Manager
Migrate IaaS resources from Classic to Azure Resource Manager using Azure PowerShell
AsmMetadataParser
This is a collection of helper tools created as part of enterprise migrations from Azure Service Management to
Azure Resource Manager. This tool allows you to replicate your infrastructure into another subscription which
can be used for testing migration and iron out any issues before running the migration on your Production
subscription.
Link to the tool documentation
migAz
migAz is an additional option to migrate a complete set of classic IaaS resources to Azure Resource Manager
IaaS resources. The migration can occur within the same subscription or between different subscriptions and
subscription types (ex: CSP subscriptions).
Link to the tool documentation
Next Steps
Manager
Frequently asked questions about classic to Azure
Resource Manager migration
Does this migration plan affect any of my existing services or

applications that run on Azure virtual machines?
No. The VMs (classic) are fully supported services in general availability. You can continue to use these resources
to expand your footprint on Microsoft Azure.
What happens to my VMs if I don’t plan on migrating in the near

future?
We are not deprecating the existing classic APIs and resource model. We want to make migration easy,
considering the advanced features that are available in the Resource Manager deployment model. We highly
recommend that you review some of the advancements that are part of IaaS under Resource Manager.
What does this migration plan mean for my existing tooling?

Updating your tooling to the Resource Manager deployment model is one of the most important changes that
you have to account for in your migration plans.
How long will the management-plane downtime be?

It depends on the number of resources that are being migrated. For smaller deployments (a few tens of VMs), the
whole migration should take less than an hour. For large-scale deployments (hundreds of VMs), the migration
can take a few hours.
Can I roll back after my migrating resources are committed in

Resource Manager?
You can abort your migration as long as the resources are in the prepared state. Rollback is not supported after
the resources have been successfully migrated through the commit operation.
Can I roll back my migration if the commit operation fails?

You cannot abort migration if the commit operation fails. All migration operations, including the commit
operation, are idempotent. So we recommend that you retry the operation after a short time. If you still face an
error, create a support ticket or create a forum post with the ClassicIaaSMigration tag on our VM forum.
Do I have to buy another express route circuit if I have to use IaaS

under Resource Manager?
No. We recently enabled moving ExpressRoute circuits from the classic to the Resource Manager deployment
model. You don’t have to buy a new ExpressRoute circuit if you already have one.
What if I had configured Role-Based Access Control policies for my

classic IaaS resources?
During migration, the resources transform from classic to Resource Manager. So we recommend that you plan
the RBAC policy updates that need to happen after migration.
I backed up my classic VMs in a vault. Can I migrate my VMs from

classic mode to Resource Manager mode and protect them in a
Recovery Services vault?
you move a VM from classic to Resource Manager mode, backups taken prior to migration will not migrate to
When
newly migrated Resource Manager VM. However, if you wish to keep your backups of classic VMs, follow these
steps before the migration.
1. In the Recovery Services vault, go to the Protected Items tab and select the VM.
2. Click Stop Protection. Leave Delete associated backup data option unchecked.
NOTE
You will be charged backup instance cost till you retain data. Backup copies will be pruned as per retention range. However,
last backup copy is always kept until you explicitly delete backup data. It is advised to check your retention range of the
Virtual machine and trigger "Delete Backup Data" on the protected item in the vault once the retention range is over.
To migrate the virtual machine to Resource Manager mode,

1. Delete the backup/snapshot extension from the VM.
2. Migrate the virtual machine from classic mode to Resource Manager mode. Make sure the storage and
network information corresponding to the virtual machine is also migrated to Resource Manager mode.
Additionally, if you want to back up the migrated VM, go to Virtual Machine management blade to enable
backup.
Can I validate my subscription or resources to see if they're capable of

migration?
Yes. In the platform-supported migration option, the first step in preparing for migration is to validate that the
resources are capable of migration. In case the validate operation fails, you receive messages for all the reasons
the migration cannot be completed.
What happens if I run into a quota error while preparing the IaaS
resources for migration?
We recommend that you abort your migration and then log a support request to increase the quotas in the
region where you are migrating the VMs. After the quota request is approved, you can start executing the
migration steps again.
How do I report an issue?

Post your issues and questions about migration to our VM forum, with the keyword ClassicIaaSMigration. We
recommend posting all your questions on this forum. If you have a support contract, you're welcome to log a
support ticket as well.
What if I don't like the names of the resources that the platform chose
during migration?
All the resources that you explicitly provide names for in the classic deployment model are retained during
migration. In some cases, new resources are created. For example: a network interface is created for every VM.
We currently don't support the ability to control the names of these new resources created during migration. Log
your votes for this feature on the Azure feedback forum.
Can I migrate ExpressRoute circuits used across subscriptions with

authorization links?
ExpressRoute circuits which use cross-subscription authorization links cannot be migrated automatically without
downtime. We have guidance on how these can be migrated using manual steps. See Migrate ExpressRoute
circuits and associated virtual networks from the classic to the Resource Manager deployment model for steps
and more information.
I got the message "VM is reporting the overall agent status as Not
Ready. Hence, the VM cannot be migrated. Ensure that the VM Agent is
reporting overall agent status as Ready" or "VM contains Extension
whose Status is not being reported from the VM. Hence, this VM cannot
be migrated."
This message is received when the VM does not have outbound connectivity to the internet. The VM agent uses
outbound connectivity to reach the Azure storage account for updating the agent status every five minutes.
Next steps
Allocation failures
Allocation failures
Allocation failures for classic deployments
Boot diagnostics
RDP
Reset RDP
RDP troubleshooting
Detailed RDP troubleshooting
Troubleshoot specific errors
SSH
SSH troubleshooting
Detailed SSH troubleshooting
Common error messages
Performance issues with Windows VMs
How to use PerfInsights
Performance diagnostics extension
Install Windows VM agent offline
Redeploy a VM
Linux
Windows
Reset VM password
Windows
Linux
Reset NIC
Restarting or resizing a VM
Use the serial console
Linux VM
Serial Console GRUB/Single user mode
Serial Console NMI/SysRq
Windows VM
CMD and PowerShell commands
Errors when deleting storage resources
Unexpected reboots of VMs with attached VHDs
Windows activation problems
Application access issues
Troubleshoot deployments
Linux
Windows
Device names are changed
VM recovery access
Windows
PowerShell
Azure portal
Linux
CLI
Azure portal
Boot errors
BitLocker errors
Checking file system errors
Blue screen errors
Throttling errors
Use nested virtualization
Understand a system reboot
Understand the structure and syntax of Azure
Resource Manager Templates
This article describes the structure of an Azure Resource Manager template. It presents the different sections of a
template and the properties that are available in those sections. The template consists of JSON and expressions
that you can use to construct values for your deployment. For a step-by-step tutorial on creating a template, see
Create your first Azure Resource Manager template.
Quickstarts and tutorials

Use the following quickstarts and tutorials to learn how to develop resource manager templates:
Quickstarts
TITLE DESCRIPTION
Use the Azure portal Generate a template using the portal, and the process of
editing and deploying the template.
Use Visual Studio Code Use Visual Studio Code to create and edit templates, and
how to use the Azure Cloud shell to deploy templates.
Use Visual Studio Use Visual Studio to create, edit, and deploy templates.
Tutorials
TITLE DESCRIPTION
Utilize template reference Utilize the template reference documentation to develop

templates. In the tutorial, you find the storage account
schema, and use the information to create an encrypted
storage account.
Create multiple instances Create multiple instances of Azure resources. In the

tutorial, you create multiple instances of storage account.
Set resource deployment order Define resource dependencies. In the tutorial, you create a
virtual network, a virtual machine, and the dependent
Azure resources. You learn how the dependencies are
defined.
Use conditions Deploy resources based on some parameter values. In the

tutorial, you define a template to create a new storage
account or use an existing storage account based on the
value of a parameter.
Integrate key vault Retrieve secrets/passwords from Azure Key Vault. In the
tutorial, you create a virtual machine. The virtual machine
administrator password is retrieved from a Key Vault.
TITLE DESCRIPTION
Create linked templates Modularize templates, and call other templates from a
template. In the tutorial, you create a virtual network, a
virtual machine, and the dependent resources. The
dependent storage account is defined in a linked template.
Use safe deployment practices Use Azure Deployment manager.
Template format
In its simplest structure, a template has the following elements:
{
"$schema": "http://schema.management.azure.com/schemas/2015-01-01/deploymentTemplate.json#",
"contentVersion": "",
"parameters": { },
"variables": { },
"functions": [ ],
"resources": [ ],
"outputs": { }
}
ELEMENT NAME REQUIRED DESCRIPTION
$schema Yes Location of the JSON schema file that

describes the version of the template
language. Use the URL shown in the
preceding example.
contentVersion Yes Version of the template (such as

1.0.0.0). You can provide any value for
this element. Use this value to
document significant changes in your
template. When deploying resources
using the template, this value can be
used to make sure that the right
template is being used.
parameters No Values that are provided when

deployment is executed to customize
resource deployment.
variables No Values that are used as JSON fragments

in the template to simplify template
language expressions.
functions No User-defined functions that are

available within the template.
resources Yes Resource types that are deployed or

updated in a resource group.
outputs No Values that are returned after

deployment.
Each element has properties you can set. The following example shows the full syntax for a template:
{
"$schema": "http://schema.management.azure.com/schemas/2015-01-01/deploymentTemplate.json#",
"contentVersion": "",
"parameters": {
"<parameter-name>" : {
"type" : "<type-of-parameter-value>",
"defaultValue": "<default-value-of-parameter>",
"allowedValues": [ "<array-of-allowed-values>" ],
"minValue": <minimum-value-for-int>,
"maxValue": <maximum-value-for-int>,
"minLength": <minimum-length-for-string-or-array>,
"maxLength": <maximum-length-for-string-or-array-parameters>,
"metadata": {
"description": "<description-of-the parameter>"
}
}
},
"variables": {
"<variable-name>": "<variable-value>",
"<variable-object-name>": {
<variable-complex-type-value>
},
"<variable-object-name>": {
"copy": [
{
"name": "<name-of-array-property>",
"count": <number-of-iterations>,
"input": {
<properties-to-repeat>
}
}
]
},
"copy": [
{
"name": "<variable-array-name>",
"count": <number-of-iterations>,
"input": {
<properties-to-repeat>
}
}
]
},
"functions": [
{
"namespace": "<namespace-for-your-function>",
"members": {
"<function-name>": {
"parameters": [
{
"name": "<parameter-name>",
"type": "<type-of-parameter-value>"
}
],
"output": {
"type": "<type-of-output-value>",
"value": "<function-expression>"
}
}
}
}
],
"resources": [
{
"condition": "<boolean-value-whether-to-deploy>",
"apiVersion": "<api-version-of-resource>",
"type": "<resource-provider-namespace/resource-type-name>",
"name": "<name-of-the-resource>",
"name": "<name-of-the-resource>",
"location": "<location-of-resource>",
"tags": {
"<tag-name1>": "<tag-value1>",
"<tag-name2>": "<tag-value2>"
},
"comments": "<your-reference-notes>",
"copy": {
"name": "<name-of-copy-loop>",
"count": "<number-of-iterations>",
"mode": "<serial-or-parallel>",
"batchSize": "<number-to-deploy-serially>"
},
"dependsOn": [
"<array-of-related-resource-names>"
],
"properties": {
"<settings-for-the-resource>",
"copy": [
{
"name": ,
"count": ,
"input": {}
}
]
},
"resources": [
"<array-of-child-resources>"
]
}
],
"outputs": {
"<outputName>" : {
"type" : "<type-of-output-value>",
"value": "<output-value-expression>"
}
}
}
This article describes the sections of the template in greater detail.
Syntax
The basic syntax of the template is JSON. However, expressions and functions extend the JSON values available
within the template. Expressions are written within JSON string literals whose first and last characters are the
brackets: [ and ] , respectively. The value of the expression is evaluated when the template is deployed. While
written as a string literal, the result of evaluating the expression can be of a different JSON type, such as an array
or integer, depending on the actual expression. To have a literal string start with a bracket [ , but not have it
interpreted as an expression, add an extra bracket to start the string with [[ .
Typically, you use expressions with functions to perform operations for configuring the deployment. Just like in
JavaScript, function calls are formatted as functionName(arg1,arg2,arg3) . You reference properties by using the dot
and [index] operators.
The following example shows how to use several functions when constructing a value:
"variables": {
"storageName": "[concat(toLower(parameters('storageNamePrefix')), uniqueString(resourceGroup().id))]"
}
For the full list of template functions, see Azure Resource Manager template functions.
Parameters
In the parameters section of the template, you specify which values you can input when deploying the resources.
These parameter values enable you to customize the deployment by providing values that are tailored for a
particular environment (such as dev, test, and production). You don't have to provide parameters in your template,
but without parameters your template would always deploy the same resources with the same names, locations,
and properties.
The following example shows a simple parameter definition:
"parameters": {
"siteNamePrefix": {
"type": "string",
"metadata": {
"description": "The name prefix of the web app that you wish to create."
}
},
},
For information about defining parameters, see Parameters section of Azure Resource Manager templates.
Variables
In the variables section, you construct values that can be used throughout your template. You don't need to define
variables, but they often simplify your template by reducing complex expressions.
The following example shows a simple variable definition:
"variables": {
"webSiteName": "[concat(parameters('siteNamePrefix'), uniqueString(resourceGroup().id))]",
},
For information about defining variables, see Variables section of Azure Resource Manager templates.
Functions
Within your template, you can create your own functions. These functions are available for use in your template.
Typically, you define complicated expression that you don't want to repeat throughout your template. You create
the user-defined functions from expressions and functions that are supported in templates.
When defining a user function, there are some restrictions:
The function can't access variables.
The function can't access template parameters. That is, the parameters function is restricted to function
parameters.
The function can't call other user-defined functions.
The function can't use the reference function.
Parameters for the function can't have default values.
Your functions require a namespace value to avoid naming conflicts with template functions. The following
example shows a function that returns a storage account name:
"functions": [
{
"namespace": "contoso",
"members": {
"uniqueName": {
"parameters": [
{
"name": "namePrefix",
"type": "string"
}
],
"output": {
"type": "string",
"value": "[concat(toLower(parameters('namePrefix')), uniqueString(resourceGroup().id))]"
}
}
}
}
],
You call the function with:
"resources": [
{
"name": "[contoso.uniqueName(parameters('storageNamePrefix'))]",
"apiVersion": "2016-01-01",
"sku": {
},
"kind": "Storage",
"location": "South Central US",
"tags": {},
"properties": {}
}
]
Resources
In the resources section, you define the resources that are deployed or updated. This section can get complicated
because you must understand the types you're deploying to provide the right values.
"resources": [
{
"apiVersion": "2016-08-01",
"name": "[variables('webSiteName')]",
"type": "Microsoft.Web/sites",
"properties": {
"serverFarmId": "/subscriptions/<subscription-id>/resourcegroups/<resource-group-
name>/providers/Microsoft.Web/serverFarms/<plan-name>"
}
}
],
For more information, see Resources section of Azure Resource Manager templates.
Outputs
In the Outputs section, you specify values that are returned from deployment. For example, you could return the
URI to access a deployed resource.
"outputs": {
"newHostName": {
"type": "string",
"value": "[reference(variables('webSiteName')).defaultHostName]"
}
}
For more information, see Outputs section of Azure Resource Manager templates.
Template limits
Limit the size of your template to 1 MB, and each parameter file to 64 KB. The 1-MB limit applies to the final state
of the template after it has been expanded with iterative resource definitions, and values for variables and
parameters.
You're also limited to:
256 parameters
256 variables
800 resources (including copy count)
64 output values
24,576 characters in a template expression
You can exceed some template limits by using a nested template. For more information, see Using linked templates
when deploying Azure resources. To reduce the number of parameters, variables, or outputs, you can combine
several values into an object. For more information, see Objects as parameters.
Next steps
To view complete templates for many different types of solutions, see the Azure Quickstart Templates.
For details about the functions you can use from within a template, see Azure Resource Manager Template
Functions.
To combine multiple templates during deployment, see Using linked templates with Azure Resource Manager.
For recommendations on creating Resource Manager templates that you can use across global Azure, Azure
sovereign clouds, and Azure Stack, see Develop Azure Resource Manager templates for cloud consistency.
Frequently asked question about Windows Virtual
Machines
This article addresses some common questions about Windows virtual machines created in Azure using the
Resource Manager deployment model. For the Linux version of this topic, see Frequently asked question about
Linux Virtual Machines
What can I run on an Azure VM?

All subscribers can run server software on an Azure virtual machine. For information about the support policy for
running Microsoft server software in Azure, see Microsoft server software support for Azure Virtual Machines
Certain versions of Windows 7, Windows 8.1, and Windows 10 are available to MSDN Azure benefit subscribers
and MSDN Dev and Test Pay-As-You-Go subscribers, for development and test tasks. For details, including
instructions and limitations, see Windows Client images for MSDN subscribers.
How much storage can I use with a virtual machine?

Each data disk can be up to 4 TB (4,095 GB ). The number of data disks you can use depends on the size of the
virtual machine. For details, see Sizes for Virtual Machines.
Azure Managed Disks are the recommended disk storage offerings for use with Azure Virtual Machines for
persistent storage of data. You can use multiple Managed Disks with each Virtual Machine. Managed Disks offer
two types of durable storage options: Premium and Standard Managed Disks. For pricing information, see
Managed Disks Pricing.
Azure storage accounts can also provide storage for the operating system disk and any data disks. Each disk is a
.vhd file stored as a page blob. For pricing details, see Storage Pricing Details.
How can I access my virtual machine?

Establish a remote connection using Remote Desktop Connection (RDP ) for a Windows VM. For instructions, see
How to connect and log on to an Azure virtual machine running Windows. A maximum of two concurrent
connections are supported, unless the server is configured as a Remote Desktop Services session host.
If you’re having problems with Remote Desktop, see Troubleshoot Remote Desktop connections to a Windows-
based Azure Virtual Machine.
If you’re familiar with Hyper-V, you might be looking for a tool similar to VMConnect. Azure doesn’t offer a similar
tool because console access to a virtual machine isn’t supported.
Can I use the temporary disk (the D: drive by default) to store data?
Don’t use the temporary disk to store data. It is only temporary storage, so you would risk losing data that can’t be
recovered. Data loss can occur when the virtual machine moves to a different host. Resizing a virtual machine,
updating the host, or a hardware failure on the host are some of the reasons a virtual machine might move.
If you have an application that needs to use the D: drive letter, you can reassign drive letters so that the temporary
disk uses something other than D:. For instructions, see Change the drive letter of the Windows temporary disk.
How can I change the drive letter of the temporary disk?
You can change the drive letter by moving the page file and reassigning drive letters, but you need to make sure
you do the steps in a specific order. For instructions, see Change the drive letter of the Windows temporary disk.
Can I add an existing VM to an availability set?

No. If you want your VM to be part of an availability set, you need to create the VM within the set. There currently
isn't a way to add a VM to an availability set after it has been created.
Can I upload a virtual machine to Azure?

Yes. For instructions, see Migrating on-premises VMs to Azure.
Can I resize the OS disk?

Yes. For instructions, see How to expand the OS drive of a Virtual Machine in an Azure Resource Group.
Can I copy or clone an existing Azure VM?

Yes. Using managed images, you can create an image of a virtual machine and then use the image to build multiple
new VMs. For instructions, see Create a custom image of a VM.
Why am I not seeing Canada Central and Canada East regions through
Azure Resource Manager?
The two new regions of Canada Central and Canada East are not automatically registered for virtual machine
creation for existing Azure subscriptions. This registration is done automatically when a virtual machine is
deployed through the Azure portal to any other region using Azure Resource Manager. After a virtual machine is
deployed to any other Azure region, the new regions should be available for subsequent virtual machines.
Does Azure support Linux VMs?

Yes. To quickly create a Linux VM to try out, see Create a Linux VM on Azure using the Portal.
Can I add a NIC to my VM after it's created?

Yes, this is now possible. The VM first needs to be stopped deallocated. Then you can add or remove a NIC (unless
it's the last NIC on the VM ).
Are there any computer name requirements?

Yes. The computer name can be a maximum of 15 characters in length. See Naming conventions rules and
restrictions for more information around naming your resources.
Are there any resource group name requirements?

Yes. The resource group name can be a maximum of 90 characters in length. See Naming conventions rules and
restrictions for more information about resource groups.
What are the username requirements when creating a VM?

Usernames can be a maximum of 20 characters in length and cannot end in a period (".").
The following usernames are not allowed:
1 123 a actuser
adm admin admin1 admin2
administrator aspnet backup console
david guest john owner
root server sql support
support_388945a0 sys test test1
test2 test3 user user1
user2 user3 user4 user5
What are the password requirements when creating a VM?

Passwords must be 12 - 123 characters in length and meet 3 out of the following 4 complexity requirements:
Have lower characters
Have upper characters
Have a digit
Have a special character (Regex match [\W_])
The following passwords are not allowed:
abc@123 iloveyou! P@$$w0rd P@ssw0rd P@ssword123
Pa$$word pass@word1 Password! Password1 Password22

Azure MS Docs PDF

Uploaded by

Copyright:

Available Formats

Azure MS Docs PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Azure MS Docs PDF

Uploaded by

Copyright:

Available Formats

Contents

Windows VMs Documentation

What do I need to think about before creating a VM?

Azure PowerShell Use the Get-AzureRmLocation command.

REST API Use the List locations operation.

Azure CLI Use the az account list-locations operation.

Azure PowerShell Get-AzureRMVMImagePublisher -Location location

REST APIs List image publishers

Azure CLI az vm image list-publishers --location location

RESOURCE REQUIRED DESCRIPTION

Resource group Yes The VM must be contained in a

Storage account Yes The VM needs the storage account to

Virtual network Yes The VM must be a member of a virtual

Public IP address No The VM can have a public IP address

Network interface Yes The VM needs the network interface to

Data disks No The VM can include data disks to

How do I create my first VM?

Templates Create a Windows virtual machine with a Resource Manager

Azure PowerShell Create a Windows VM using PowerShell

Client SDKs Deploy Azure Resources using C#

REST APIs Create or update a VM

Azure CLI Create a VM with the Azure CLI

How do I manage the VM that I created?

REST API Use the Get VM information operation to get information

Create virtual machine

Connect to virtual machine

Install web server

Install-WindowsFeature -name Web-Server -IncludeManagementTools

When done, close the RDP connection to the VM.

View the IIS welcome page

Launch Azure Cloud Shell

Create resource group

New-AzureRmResourceGroup -Name myResourceGroup -Location EastUS

Create virtual machine

Get-AzureRmPublicIpAddress -ResourceGroupName "myResourceGroup" | Select "IpAddress"

Install web server

Install-WindowsFeature -name Web-Server -IncludeManagementTools

When done, close the RDP connection to the VM.

View the web server in action

Remove-AzureRmResourceGroup -Name myResourceGroup

Open Azure Cloud Shell

Select Try It in the upper-right corner of a code block.

Open Cloud Shell in your browser.

Select the Cloud Shell button on the menu in the upper-right

Create a resource group

az group create --name myResourceGroup --location eastus

Create virtual machine

Open port 80 for web traffic

az vm open-port --port 80 --resource-group myResourceGroup --name myVM

Connect to virtual machine

Install web server

Install-WindowsFeature -name Web-Server -IncludeManagementTools

When done, close the RDP connection to the VM.

View the web server in action

az group delete --name myResourceGroup

Launch Azure Cloud Shell

Click Try It in the upper right corner of a code block.

Open Cloud Shell in your browser.

Create resource group