Virtualization performance: VMware vSphere 5 vs. Red Hat Enterprise Virtualization 3
•
2 likes•5,001 views
Using a hypervisor that offers better resource management and scalability can deliver excellent virtual machine performance on your servers. In our testing, VMware vSphere 5 allowed our host’s virtual machines to outperform those running on RHEV 3 by over 28 percent in total OPM performance. Furthermore, VMware vSphere 5 performance continued to improve when going from 39 VMs to 42 VMs: Total performance for VMware vSphere 5 increased by 2.8 percent, whereas it decreased by 7.2 percent with RHEV 3.
With the capabilities and scalability that VMware vSphere 5 offers, you are able to utilize the full capacity of your servers with confidence and purchase fewer servers to handle workload spikes; this can translate to fewer racks in the data center, lower costs for your business, and more consistent overall application performance.
1 of 30
Download to read offline
More Related Content
Virtualization performance: VMware vSphere 5 vs. Red Hat Enterprise Virtualization 3
1. VIRTUALIZATION PERFORMANCE: VMWARE VSPHERE 5
VS. RED HAT ENTERPRISE VIRTUALIZATION 3
When you invest in a virtualization platform, you can maximize the performance of
your applications and the overall infrastructure by fully utilizing physical resources as much
as a hypervisor allows. Planning for scenarios with greater virtual machine (VM) densities,
such as during maintenance periods or during high availability (HA) events where a server
must host additional VMs from a failed server, is critical to an organization’s overall IT
strategy. A hypervisor that excels at resource management allows for greater virtual
machine (VM) density and better application performance.
We tested two hypervisors—VMware vSphere 5, and Red Hat Enterprise
Virtualization 3 (RHEV)—to compare their performance and ability to manage resources at
high levels of RAM utilization. When running a heavily loaded host with 39 virtual machines,
VMware vSphere 5 outperformed RHEV 3 by 16.2 percent; after adding a few more virtual
machines to the host, VMware’s advantage increased to 28.6 percent. Additionally, VMware
vSphere 5 continued to scale from 39 VMs up to 42 VMs with our workload: Overall server
performance increased by 2.8 percent with VMware vSphere 5, whereas performance
decreased by 7.2 percent with RHEV 3.
These results show that VMware vSphere 5 can deliver superior performance in a
densely virtualized environment over RHEV 3. When you fully utilize your servers, you need
fewer systems to perform the same amount of work during normal operations, and you do
not need excessive server capacity to handle workload peaks. This results in overall savings
for your organization.
MAY 2012
A PRINCIPLED TECHNOLOGIES TEST REPORT
Commissioned by VMware, Inc.
2. ADVANCED RESOURCE MANAGEMENT MAXIMIZES PERFORMANCE
A hypervisor with exceptional resource management lets you optimize virtual
machine performance across your entire infrastructure. The result? Greater density,
scalability, and performance. VMware offers several unique features that enable
vSphere 5 to utilize system resources better than competing hypervisors.
Direct driver model. The VMware approach is to install device drivers directly
onto the hypervisor, effectively making the hypervisor an intermediary between the
physical hardware and VMs that run on the server. The direct driver model improves
performance and scalability as the number of VMs on a host increases.
High-performance “gang” scheduler. This feature allows VMware vSphere 5 to
handle the challenging CPU and I/O needs of VMs. vSphere 5 is thus able to allocate
resources and processor time slices to the VMs that most need it.
How VMware manages memory
Additional VMware technologies allow vSphere 5 to optimize physical memory
allocation, dynamically shifting this critical resource from less active VMs to VMs that
are more active. This is accommodated in vSphere by the following features, working in
concert:
Transparent page sharing. Transparent page sharing (TPS) identifies common
pages across VMs and stores each in physical memory only once. This is somewhat
analogous to deduplication technologies used in storage implementations. All VMs then
share only that single copy of the memory page. VMware vSphere 5 determines
sharable pages by scanning the content of the virtual machines’ physical memory for
sharing opportunities. By eliminating these redundant copies, VMware vSphere 5 frees
up memory for use by workloads.
Memory ballooning. When the hypervisor needs to give more memory to VMs
that are powering on or experiencing a heavy workload, it “asks” the guest operating
systems in other VMs to provide memory to a balloon process that runs in the guest.
The hypervisor can then temporarily lend that “ballooned” memory to the busy VMs.
When the busy VMs no longer need the extra memory, the hypervisor “deflates” the
balloon, returning that memory to the original guest OS.
Memory compression. The innovative memory compression capability in
VMware vSphere 5 sets aside a small portion of physical RAM as a cache. Compressing
unused memory pages avoids hypervisor swapping and is orders of magnitude faster
than disk.
Hypervisor swap. If a system’s memory resources are experiencing intense
pressure, hypervisor swap acts as a safety valve, ensuring reliable operation of the host
and all workloads. While this may result in a short-term performance hit, it offers the
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 2
Red Hat Enterprise Virtualization 3
3. hypervisor another option to resolve memory issues. Furthermore, a new feature in
vSphere 5, called swap to host cache, can use solid-state disks for swap purposes,
reducing the impact on performance. However, we did not use this feature for this
testing.
DRS with resource pools. This feature is a safety net of sorts, largely because it
ensures that applications receive the resources they need when they need them. It
accomplishes this by dynamically load balancing resources throughout a cluster of VMs.
This does not apply to standalone hosts such as the one tested for this report, but to
vSphere clusters. Using a vSphere-clustered environment with DRS ensures optimization
of resources and the ability to accommodate shifting workloads.
PUTTING THE HYPERVISORS TO THE TEST
To compare these two hypervisors, we ran two scenarios on each hypervisor to
demonstrate different levels of memory usage. In the first scenario, we ran database
workloads against a heavily loaded host, and in the second scenario, we ran the same
database workloads and increased the VM count. We first ran these scenarios with our
server running VMware vSphere 5 and then ran the same scenarios with our server
running Red Hat Enterprise Virtualization 3. Our four-socket server had 80 logical CPUs
and 256 GB of RAM. We configured each VM with two virtual CPUs and 8 GB of RAM.
For the first scenario, we ran 39 VMs to mimic a highly utilized server scenario. In the
second scenario, we ran 42 VMs, taking the utilization higher. While RAM was fully
allocated and utilized, CPU utilization was 46 and 51 percent for the two density
scenarios on RHEV, and 39 and 41 percent for the two density scenarios on VMware
vSphere.
In both scenarios, two-thirds of our guests ran Microsoft® Windows® Server
2008 R2 SP1 with Microsoft SQL Server® 2008 R2 SP1, and one-third of the guests ran
Red Hat Enterprise Linux® 6.2 with PostgreSQL 9.1. Independent of platform, each guest
VM hosted an 8GB database created with the DVD Store Version 2 (DS2) benchmark
tool. We used clients to invoke the DS2 workload on each guest VM using identical
workload parameters. DS2 simulates an online store, and reports orders per minute
(OPM). We chose this database workload for our testing as it heavily utilizes processor,
memory, and I/O to create a broad and demanding load on the system.
For more details about the DS2 tool, see
http://www.delltechcenter.com/page/DVD+Store. For step-by-step instructions on how
we tested, see appendices to this report.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 3
Red Hat Enterprise Virtualization 3
4. VMWARE VSPHERE 5 DELIVERS
As Figure 1 shows, aggregate performance across all VMs in our 39-VM scenario
with VMware vSphere 5 was 161,040 OPM, 16 percent higher than RHEV 3, where the
total performance of all 39 VMs was 138,603 OPM.
VMware vSphere 5 delivered 16.2% better performance
with 39 VMs
180,000
161,040
160,000 RHEV 3
138,603
Figure 1. VMware 140,000
vSphere 5
120,000 VMware vSphere 5
delivered 16.2
OPM
percent better 100,000
overall
performance 80,000
with 39 VMs.
60,000
40,000
20,000
0
After increasing the VM density to 42 VMs on our single host, we ran the same
test again. As Figure 2 shows, the performance advantage of VMware vSphere 5
increases at the higher density. Aggregate performance across all VMs in our 42-VM
scenario with VMware vSphere 5 was 165,473 OPM, 28 percent higher than on the
server running RHEV 3, where the total performance of all 42 VMs was 128,684 OPM.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 4
Red Hat Enterprise Virtualization 3
5. VMware vSphere 5 delivered 28.6% better performance
with 42 VMs
180,000 165,473
160,000 RHEV 3
Figure 2. VMware 140,000 128,684
vSphere 5
120,000 VMware vSphere 5
delivered 28.6
OPM
percent better 100,000
overall
performance 80,000
with 42 VMs.
60,000
40,000
20,000
0
As the above results show, and as Figure 3 shows, VMware vSphere 5 better
handled the increase in density from 39 to 42 VMs, increasing in total output by almost
3 percent. RHEV 3, however, decreased in performance by over 7 percent at the higher
density.
Total database performance change when
scaling from 39 VMs to 42 VMs
4%
2.8%
Figure 3. VMware 2%
vSphere 5 RHEV 3
increases in
output with an 0%
increase in VMware vSphere 5
density, while -2%
RHEV 3
performance
degraded. -4%
-6%
-8% -7.2%
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 5
Red Hat Enterprise Virtualization 3
6. CONCLUSION
Using a hypervisor that offers better resource management and scalability can
deliver excellent virtual machine performance on your servers. In our testing, VMware
vSphere 5 allowed our host’s virtual machines to outperform those running on RHEV 3
by over 28 percent in total OPM performance. Furthermore, VMware vSphere 5
performance continued to improve when going from 39 VMs to 42 VMs: Total
performance for VMware vSphere 5 increased by 2.8 percent, whereas it decreased by
7.2 percent with RHEV 3.
With the capabilities and scalability that VMware vSphere 5 offers, you are able
to utilize the full capacity of your servers with confidence and purchase fewer servers to
handle workload spikes; this can translate to fewer racks in the data center, lower costs
for your business, and more consistent overall application performance.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 6
Red Hat Enterprise Virtualization 3
7. APPENDIX A – TEST CONFIGURATION OVERVIEW
We used 22 client machines for both test scenarios, with each client targeting one or two VMs on the server
under test depending on the specific scenario. We first tested the VMs on VMware vSphere 5, and then on Red Hat
Enterprise Virtualization Hypervisor 6.2. We ran both environments, one by one, on the same hardware.
We took advantage of the ability of DVD Store 2.1 to target multiple targets with each instance targeting one or
two VMs. Each client machine was a VM running Windows Server 2003 Enterprise R2 SP2 with the following:
A client share folder with the latest ds2postgresqlserverdriver.exe, ds2sqlserverdriver.exe,
Mono.Security.dll, Npgsql.dll, and the necessary run scripts
.NET framework 3.5 SP1, required for the DS2 SQL Server driver
.NET framework 4, required for the DS2 PostgreSQL driver and Npqsql .Net Data Provider.
We ran one physical workstation class client machine to ensure no performance differences between running virtual and
physical clients.
As Figure 4 shows, we cabled all four of the client host’s 1Gb NICs to a Dell™ PowerConnect™ 6248 switch. We
then created four virtual machine networks and dedicated each physical NIC to its own virtual machine network. We
used a round robin approach to distribute the virtual NICs amongst the VMs.
On the server under test, we attached all six NICs in the IBM® System x3850 X5 to the same PowerConnect 6248
switch, using the two onboard NICs for management traffic and a quad-port Intel® I340-T4 Ethernet server adapter for
VM traffic. We initiated and controlled all testing sessions from a separate controller machine.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 7
Red Hat Enterprise Virtualization 3
8. Figure 4. The test bed layout we used for both the VMware vSphere 5 and Red Hat Enterprise
Virtualization 3 testing.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 8
Red Hat Enterprise Virtualization 3
9. APPENDIX B – SETTING UP THE STORAGE
IBM System x3850 X5 and IBM System Storage DS3524 Express storage configuration overview
Our complete storage infrastructure consisted of four internal drives in the server, an IBM ServeRAID internal
RAID controller, two Emulex® LightPath LPe12000 single port 8Gb/s fibre cards attached directly to the IBM System
Storage DS3524, and one IBM System Storage DS3524 array with three IBM System Storage EXP3524 expansion trays.
Figure 4 on the previous page shows the complete layout of our test bed hardware.
We configured the internal drives on the server in two RAID 1 volumes, each volume containing two physical
disks. We dedicated one volume to the vSphere partition, and the other to the Red Hat Enterprise Virtualization
Hypervisor partition. To switch between environments during testing, we toggled the assigned boot volume in the RAID
controller LSI WebBIOS and enabled the Legacy Only boot option in the system BIOS (applicable to RHEV-H only).
For external storage, we used one IBM System Storage DS3524 Express array with three IBM System Storage
EXP3524 expansion trays, each containing 24 drives, for a total of 96 usable disk drives. We cabled each IBM tray
according to IBM best practices using SAS cables to interconnect the environmental service modules. We then cabled
the two controllers from IBM System Storage DS3524 directly to the two Emulex LightPath LPe12000 single port 8Gb/s
Fibre Channel cards installed in the IBM System x3850 X5 via two fibre cables. In both VMware and RHEV environments,
to ensure an even distribution of traffic across both controllers and fibre cards, we assigned half of our configured LUNs
to controller A and the other half to controller B using the IBM DS Storage Manager.
The entire IBM System Storage DS3524 Express array contained 96 drives. Due to the conceptual storage design
differences between VMware and RHEV, we used a different storage configuration for each, namely RHEV’s requirement
that VMs are owned by a single storage domain. When switching between each hypervisor, we reconfigured the IBM
System Storage DS3524 Express array using the parameters listed below using the IBM DS Storage Manager.
We created the following for VMware:
Note: VMware allows a virtual machine’s virtual disks to reside on separate datastores, allowing the storage of
OS, log, and data I/O across different sets of physical disks.
A 24-disk array in a RAID 5 configuration. We then used this array to create two Logical Drives (LUNs) for
our VMs’ virtual disks containing the operating system files. We assigned each LUN to one of two
controllers.
An 18-disk array in a RAID 10 configuration. We then used this array to create two LUNs for our VMs’
virtual disks containing SQL Server and PostgreSQL transaction log files. We assigned each LUN to one of
two controllers.
A 54-disk array in a RAID 10 configuration. We then used this array to create two LUNs for our VMs’
virtual disks containing SQL Server and PostgreSQL database files. We also stored database backups on
these LUNs. We assigned each LUN to one of two controllers.
We created the following for RHEV:
Note: RHEV uses the concept of storage domains, a logical unit that can contain multiple LUNs. However, VMs’
virtual disks must all reside in a single storage domain. As such, we could not configure storage with detailed
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 9
Red Hat Enterprise Virtualization 3
10. granularity. Therefore, on each disk array we created, we used physical disks spread across all four IBM System
Storage trays.
A 96-disk array in a RAID 10 configuration. We then created four logical drives (LUNs), each with exactly
one-quarter of the overall capacity. We assigned two logical drives to controller A and the other two to
controller B.
Figure 5. The storage layout we used for both the VMware vSphere 5 and Red Hat Enterprise Virtualization 3 testing.
Setting up the external storage for VMware and RHEV
1. Log in to the IBM System Storage DS3524 storage array via the IBM DS Storage Manager on a client system.
2. Select the Logical tab.
3. Right-click the Total Unconfigured Capactiy, and click Create Array…
4. Click Next.
5. Enter a name for the drive array based on the current environment.
6. Select Manual, and click Next.
7. Select the proper amount of disks and RAID level for the current environment, and click Add.
8. Click Calculate Capacity, and click Finish.
9. At the Array Created prompt, click Yes to begin the creation of a logical drive.
10. Click Next.
11. Enter the proper capacity amount for the current logical drive and environment.
12. Enter a proper name for the drive, and click Next.
13. Select Map later using the Mappings View, and click Finish.
14. Repeat steps 3 through 13 until the storage is completely setup for the current environment.
15. Prior to testing, allow each logical drive to complete the initialization task.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 10
Red Hat Enterprise Virtualization 3
11. 16. Ensure that the logical drives are divided between the two controllers by select each logical drive, and checking
that the preferred and current owners are divided evenly between controller A and controller B.
17. Click the Mappings tab.
18. Right-click Storage Subsystem, and click DefineHost.
19. Enter a name for the host, and click Next.
20. Select FC for the host interface.
21. Select Add by selecting a known unassociated host port identifier.
22. Click Refresh to automatically identify attached Fibre Channel adapters.
23. Select the first host port identifier, enter an alias for the port, and click Add.
24. Repeat the last step for the second host port identifier.
25. Click Next.
26. Select the proper Host type (operating system) for the current hypervisor setup, and click Next.
27. Select No, and click Next.
28. Click Finish.
29. Once the host is created, click No.
30. In the left-hand pane, right-click the new host, and click DefineAdditional Mapping.
31. Select each logical drive listed, and click Add.
32. Each LUN will now be accessible from the host.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 11
Red Hat Enterprise Virtualization 3
12. APPENDIX C – SETTING UP THE VIRTUAL ENVIRONMENT – RED HAT
ENTERPRISE VIRTUALIZATION 3
Adjusting BIOS settings
We used the latest released BIOS update on the IBM System x3850 X5, version IBM 1.73, and adjusted the
default BIOS settings. We disabled CPU c-states and set the Operating Mode to Performance. For running RHEV-H, we
added the Legacy Only option to the BIOS boot options and promoted it to the first boot option. RHEV-H would not boot
properly if this option was not configured.
Setting up the Red Hat Enterprise Virtualization 3 environment
We used the following steps to install and configure the RHEV 3 environment. To manage our IBM System x3850
X5 running RHEV-H 6.2, we setup a separate management server running RHEV-M 3.0. We also configured a basic
Windows Active Directory domain controller running DNS to handle domain and name resolution. We created a user on
the domain to be used as the login for the RHEV-M web console. We also added the RHEV-M IP address to both the
forward and reverse lookup zones in DNS.
Installing and configuring Red Hat Enterprise Virtualization Manager 3.0
We performed the following steps to install RHEV Manager, starting from a basic RHEL 6.2 installation. We
assigned an IP address and hostname for the RHEV-M server during the installation of RHEL 6.2. The following steps
require an active Internet connection and a subscription to Red Hat Network’s service. We used a trial subscription for
this testing.
1. Log in to the RHEV Manager server as root.
2. Disable SELinux by modifying the /etc/selinux/config file, and changing the SELINUX line to
SELINUX=disabled
3. Synchronize the time source.
a. Edit /etc/ntp.conf, adding the IP address of the domain controller as a time source.
b. Run the command chkconfig ntpdate on.
c. Run the command chkconfig ntpd on.
d. Run the command date, and ensure that the time is synchronized with the domain controller.
4. Add the IP address and FQDN of the domain controller to /etc/hosts.
5. Type rhn_register to begin the registration on the RHN, and accept defaults, making sure to select the
appropriate region.
6. Follow the prompts to register the system with RHN, ensuring that the yum repository for RHEL 6.2 is added
during the process.
7. Add channels for RHEV by typing the following commands:
# rhn-channel --add --channel=rhel-x86_64-server-6-rhevm-3
# rhn-channel --add --channel=jbappplatform-5-x86_64-server-6-rpm
# rhn-channel --add --channel=rhel-x86_64-server-supplementary-6
8. Turn off RHEL firewalls by typing the following commands:
# service iptables save
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 12
Red Hat Enterprise Virtualization 3
13. # service iptables stop
# chkconfig iptables off
# service ip6tables save
# service ip6tables stop
# chkconfig ip6tables off
9. Update RHEL by typing yum upgrade
10. Start the RHEV-M install process by typing rhevm-setup
11. Accept the default HTTP port by pressing Enter.
12. Accept the default HTTPS port by pressing Enter.
13. Enter a password for the admin@internal username.
14. Enter a password for the internal database.
15. Choose your default storage type.
16. Allow the installer to configure an NFS share on the local server for an ISO Domain.
17. Choose a mount path for the ISO Domain share.
18. Choose a display name for the ISO Domain.
19. Allow the installer to automatically configure the firewalls.
20. After the install is complete run the following command:
# rhevm-manage-domains -action=add -domain=<Domian name> -user=<User
created for RHEV-M> -interactive
21. Enter the password for the domain user, and press Enter to allow the command to complete the addition of the
domain to RHEV-M.
22. Upload all necessary ISO images and drivers for the VM installs, including the RHEV tools ISO for Windows and
the latest VirtIO drivers.
Installing RHEV-H 6.2
We performed the following steps to install RHEV Hypervisor on our system under test, and after installation
configured approximately 120GB for swap. For further information, refer to the RHEV install documentation located on
the RHEV-Hypervisor disk.
1. Boot to the Red Hat Enterprise Virtualization Hypervisor standalone installation disk.
2. Once the installer loads, select Install Hypervisor.
3. Select the proper internal logical disk to install RHEV on, and click Continue.
4. Select the disk again, and click Continue.
5. Enter and confirm a new password, and click Install.
6. Once the installation is completed, click Reboot.
Adding the RHEV-H 6.2 host to RHEV-M
1. Once the server has rebooted from the RHEV-H install, log in using the username “admin”. This will log in to the
RHEV-H graphical setup interface.
2. Tab to Network, and press Enter.
3. Type in a proper host name and DNS server IP address for the RHEV-H server.
4. Tab down to Apply, and press Enter.
5. On the same screen, select the proper Ethernet device for management traffic, and press Enter.
6. Select Static, and assign the proper IP address, subnet mask, and gateway.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 13
Red Hat Enterprise Virtualization 3
14. 7. Tab to Apply, and press Enter.
8. Allow the network service to reboot, and press Enter when it completes.
9. Tab to Security on the left-hand side of the screen, and press Enter.
10. Select Enable ssh password authentication, and enter the desired password.
11. Tab to Apply, and press Enter.
12. Tab to RHEVM on the left-hand side of the screen, and press Enter.
13. Enter the IP address for the RHEV-M server.
14. Select Connect to RHEV Manager and validate certificate.
15. Enter the password for the RHEV-M admin, and click Apply.
16. Once the RHEV-H server finishes connecting to the RHEV-M server, press Enter.
Connecting to the RHEV-M server from the client
1. Connect to RHEV-M by typing the fully qualified domain name of the RHEV-M server in a Web browser from a
client machine.
2. Click Administrator Portal (no SSL).
3. At the login screen, enter the name of the domain user created for RHEV-M, the user password, and select the
proper domain.
4. Press Enter to log into the administrator console.
5. Change to the Clusters tab, select the default cluster, then move to the Logical Networks sub-tab, and select Add
Network.
6. Add five additional virtual networks, one for Internet access, and four for the test network.
7. Navigate to the Default host cluster, and click Edit.
8. Click Memory Optimization, and select Optimized for Server Load to enable memory page sharing at a threshold
of 150%.
9. Click OK.
10. Navigate to the Hosts tab.
11. The RHEV-H sever should be listed. Click the Install button to finalize adding the RHEV-H server to RHEV-M.
12. Once the RHEV-H server’s status changes to Up, select the host and click Edit.
13. Select the host, and move to the Network Interfaces sub-tab.
14. In the Network Interfaces sub-tab, assign the logical networks to the appropriate ports by selecting a network
interface, selecting Add/Edit, and assigning the appropriate logical network you created earlier. Ensure that the
four test networks are assigned to the Intel quad-port NIC.
15. Change to the Storage tab and select New Domain. Add the ISO domain that was created automatically during
the RHEV-M install.
16. Still in the Storage tab, add the remaining four data domains, selecting DataFibre Channel as the storage type.
Use a naming scheme that will allow for easy identification when deploying the VMs via templates.
17. Ensure that each domain is attached to the Default cluster.
Creating the first Windows or RHEL VM
1. Log into the RHEV Manager.
2. In the Hosts tab, select New Server.
3. In the New Server Virtual Machine window, name your guest, select the Default cluster, make it based on the
Blank template, select 8 GB of RAM, and select the appropriate operating system.
4. Click Resource Allocation, and enter 1 MB for Physical Memory Guaranteed.
5. Click OK.
6. A New Virtual Machine – Guide Me window appears. Click Configure Network Interfaces.
7. In the New Network Interface window, name the NIC, select the proper guest network, select the Red Hat VirtIO
type, and click OK.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 14
Red Hat Enterprise Virtualization 3
15. 8. Back at the New Virtual Machine – Guide Me window, select Configure Virtual Disks.
9. In the New Virtual Disk window, make your disk 20 GB and select Thin Provision, and click OK. This virtual disk
will be used to install the guest operating system.
10. Back at the New Virtual Machine – Guide Me window, click Configure Later.
11. Select the newly created VM, and click the Virtual Disks tab.
12. Click New.
13. Enter 22 GB for the size, leave the default settings, and click OK. This virtual disk will be used for SQL Server and
PostgreSQL data files.
14. Click New.
15. Enter 13 GB for the size, leave the default settings, and click OK. This virtual disk will be used for SQL Server and
PostgreSQL log files.
16. Click New.
17. Enter 13 GB for the size for RHEL guests or 3GB for the size for Windows guests, leave the default settings, and
click OK. This virtual disk will be used for backup data.
See Appendix E for VM related setup.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 15
Red Hat Enterprise Virtualization 3
16. APPENDIX D – SETTING UP THE VIRTUAL ENVIRONMENT – VMWARE
VSPHERE 5
Adjusting BIOS settings
We used the latest released BIOS update on the IBM System x3850 X5, version IBM 1.73, and adjusted the
default BIOS settings. We disabled CPU c-states and set the Operating Mode to Performance.
Setting up the VMware vSphere 5 environment
We used the following steps to install and configure the VMware vSphere 5 environment. To manage our IBM
System x3850 X5 running VMware vSphere 5 (ESXi), we setup a separate management server running Windows Server
2008 R2 with the latest build of VMware vCenter™ Server and the vSphere client software.
Installing VMware vSphere 5 (ESXi) on the IBM System x3850 X5
1. Insert the disk, and select Boot from disk.
2. On the Welcome screen, press Enter.
3. On the End User License Agreement (EULA) screen, press F11.
4. On the Select a Disk to Install or Upgrade Screen, select the relevant volume to install ESXi on and press Enter.
5. On the Please Select a Keyboard Layout screen, press Enter.
6. On the Enter a Root Password Screen, assign a root password and confirm it by entering it again, and press Enter
to continue.
7. On the Confirm Install Screen, press F11 to install.
8. On the Installation complete screen, press Enter to reboot.
Configuring ESXi after Installation
1. On the 5.0.0 ESXi screen, press F2, enter the root password, and press Enter.
2. On the System Customization screen, select troubleshooting options, and press Enter.
3. On the Troubleshooting Mode Options screen, select enable ESXi Shell, and press Enter.
4. Select Enable SSH, press Enter, and press ESC.
5. On the System Customization screen, select Configure Management Network.
6. On the Configure Management Network screen, select IP Configuration.
7. On the IP Configuration screen, select set static IP, enter an IP address, subnet mask, and default gateway, and
press Enter.
8. On the Configure Management Network screen, press Esc. When asked if you want to apply the changes, press
Y.
Configuring VM networking on ESXi
1. Using the vSphere client from another machine, connect to the vCenter Server which manages the ESXi server.
2. Add the necessary vSwitches for the network that DVD Store traffic will use:
a. Click the host, click the Configuration tab, and click Networking.
b. Click Add Networking.
c. Choose Virtual Machine, and click Next.
d. Choose create a vSphere standard switch.
e. Choose the NIC associated with VM traffic.
f. Assign the network label and assign IP settings.
g. Click Finish.
h. Repeat steps a through g for the remaining three NICs from the Intel quad-port adapter.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 16
Red Hat Enterprise Virtualization 3
17. Configuring the external volumes in VMware vSphere 5
1. In the vSphere client, select the host.
1. Click the Configuration tab.
2. Click Storage, and click Add Storage…
3. Choose Disk/Lun.
4. Select the disk, and click Next.
5. Accept the default of VMFS-5 for the file system.
6. Review the disk layout, and click Next.
7. Enter the datastore name, and click Next.
8. Accept the default of using maximum capacity, and click Next.
9. Click Finish.
10. Repeat steps 3 through 10 for the remaining LUNs.
Creating the first Windows or RHEL VM
1. In the vSphere client, connect to the vCenter Server, and browse to the ESXi host.
2. Click the Virtual Machines tab.
3. Right-click, and choose New Virtual Machine.
4. Choose Custom, and click Next.
5. Assign a name to the virtual machine, and click Next.
6. Select the first assigned OS Datastore on the external storage, and click Next.
7. Choose Virtual Machine Version 8, and click Next.
8. Choose the appropriate operating system, and click Next.
9. Choose two virtual sockets, and click Next.
10. Choose 8GB RAM, and click Next.
11. Click 1 for the number of NICs, select vmxnet3, and click Next.
12. Leave the default virtual storage controller, and click Next.
13. Choose to create a new virtual disk, and click Next.
14. Make the OS virtual disk size 20 GB, choose thin-provisioned, specify the first OS datastore on the external
storage, and click Next.
15. Keep the default virtual device node (0:0), and click Next.
16. Click Finish.
17. Right-click the VM, and choose Edit Settings.
18. On the Hardware tab, click Add…
19. Click Hard Disk, and click Next.
20. Click Create a new virtual disk, and click Next.
21. Specify 22GB for the virtual disk size, choose thick-provisioned lazy zeroed, and specify the first datastore for
backup and SQL Server/PostgreSQL data usage.
22. Choose SCSI(1:0) for the device node, and click Next.
23. On the Hardware tab, click Add…
24. Click Hard Disk, and click Next.
25. Click Create a new virtual disk, and click Next.
26. Specify 13GB for the virtual disk size, choose thick-provisioned lazy zeroed, and specify the first datastore for
SQL Server/PostgreSQL log usage.
27. Choose SCSI(1:1) for the device node, and click Next.
28. On the Hardware tab, click Add…
29. Click Hard Disk, and click Next.
30. Click Create a new virtual disk, and click Next.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 17
Red Hat Enterprise Virtualization 3
18. 31. Specify 13GB for the virtual disk size for RHEL guests or 3GB for the virtual disk size for Windows guests, choose
thick-provisioned lazy zeroed, and specify the datastore for usage.
32. Choose SCSI(1:2) for the device node, and click Next.
33. Click SCSI Controller 1, and choose Change Type.
34. Choose VMware Paravirtual, and click OK.
35. Click Finish, and click OK.
36. Start the VM.
37. Attach the Windows Server 2008 R2 SP1 ISO or RHEL 6.2 ISO image to the VM and install the operating system
on your VM.
See Appendix E for VM-related setup.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 18
Red Hat Enterprise Virtualization 3
19. APPENDIX E – CONFIGURING THE VMS ON EACH HYPERVISOR
See the above sections regarding the initial creation of the virtual machines on each hypervisor. We provide
steps below for installing the operating system, Microsoft SQL Server or PostgreSQL, and configuration of the VMs.
Installing the operating system on the Windows VM
1. Insert the installation DVD for Windows Server 2008 R2 SP1 Enterprise into the DVD drive, and attach the
physical DVD drive to the VM. Alternatively, use an ISO image and connect to the ISO image from the VM
console. For RHEV, right-click the VM, and click Run Once to open up a menu to attach both the VirtIO driver
virtual floppy and the Windows installation ISO.
2. Open the VM console on vCenter Server or RHEV-M.
3. At the Language Selection Screen, click Next.
4. Click Install Now.
5. Select Windows Server 2008 R2 Enterprise (Full Installation), and click Next.
6. Click the I accept the license terms check box, and click Next.
7. Click Custom.
8. On the installation hard drive selection, select the virtual hard drive. For RHEL, the VirtIO drivers need to be
installed before the disk will appear in the list.
9. Click Next.
10. At the User’s password must be changed before logging on warning screen, click OK.
11. Enter the desired password for the administrator in both fields, and click the arrow to continue.
12. At the Your password has been changed screen, click OK.
13. Install the latest VMware tools package on the VM if using vSphere; install the latest RHEV tools on the VM if
using RHEV. Restart as necessary.
14. Connect the machine to the Internet, and install all available Windows updates. Restart as necessary.
15. Enable remote desktop access.
16. Change the hostname and reboot when prompted.
17. Create a shared folder to store test script files. Set permissions as needed.
18. Adjust Internet Time synchronization.
a. Click the clock in the lower right-hand corner.
b. Select Change date and time settings.
c. Click the Internet Time tab.
d. Click Change settings…
e. Enter the IP address for the domain controller in the Server field.
f. Click Update now, and click OK.
g. Click OK to close the properties window.
19. Set up networking:
a. Click StartControl Panel, right-click Network Connections, and choose Open.
b. Right-click the VM traffic NIC, and choose Properties.
c. Select TCP/IP (v4), and choose Properties.
d. Set the IP address and subnet for the virtual NIC, which will handle outgoing server traffic. Click OK, and click
Close.
20. In the VM, configure the VM storage:
a. Click the Server Manager icon in the taskbar.
b. In the left pane, expand Storage and click Disk Management.
c. Right-click the first volume, and choose Initialize Disk.
d. In the right pane, right-click the volume and choose New Simple VoIume…
e. At the welcome window, click Next.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 19
Red Hat Enterprise Virtualization 3
20. f. At the Specify Volume Size window, leave the default selection, and click Next.
g. At the Assign Drive Letter or Path window, choose a drive letter, and click Next.
h. At the Format Partition window, choose NTFS and 64K allocation unit size, and click Next.
i. At the Completing the New Simple Volume Wizard window, click Finish.
j. Repeat steps c through i for the remaining VM volumes.
21. Copy the pre-created DVD Store backup file to the backup virtual disk inside the first VM.
Installing SQL Server 2008 R2 SP1 on the Windows VM
1. Open the console for the VM.
2. Log into the virtual machine.
3. Insert the installation DVD for SQL Server 2008 R2 into the host server’s DVD drive.
4. Attach the physical DVD drive to the VM.
5. Click Run SETUP.EXE. If Autoplay does not begin the installation, navigate to the SQL Server 2008 R2 DVD, and
double-click.
6. If the installer prompts you with a .NET installation prompt, click Yes to enable the .NET Framework Core role.
7. In the left pane, click Installation.
8. Click New installation or add features to an existing installation.
9. At the Setup Support Rules screen, wait for the check to complete. If there are no failures or relevant warnings,
click OK.
10. Select the Enter the product key radio button, and enter the product key. Click Next.
11. Click the checkbox to accept the license terms, and click Next.
12. Click Install to install the setup support files.
13. If there are no failures displayed, click Next. You may see a Computer domain controller warning and a Windows
Firewall warning. For now, ignore these.
14. At the Setup Role screen, choose SQL Server Feature Installation.
15. At the Feature Selection screen, select Database Engine Services, Full-Text Search, Client Tools Connectivity,
Client Tools Backwards Compatibility, Management Tools –Basic, and Management Tools – Complete. Click Next.
16. At the Installation Rules screen, click Next once the check completes.
17. At the Instance configuration screen, leave the default selection of default instance, and click Next.
18. At the Disk space requirements screen, click Next.
19. At the Server configuration screen, choose NT AUTHORITYSYSTEM for SQL Server Agent, and choose NT
AUTHORITYSYSTEM for SQL Server Database Engine. Click Next.
20. At the Database Engine Configuration screen, select Mixed Mode.
21. Enter and confirm a password for the system administrator account.
22. Click Add Current user. This may take several seconds.
23. Click Next.
24. At the Error and usage reporting screen, click Next.
25. At the Installation Configuration rules screen, check that there are no failures or relevant warnings, and click
Next.
26. At the Ready to Install screen, click Install.
27. After installation completes, click Next.
28. Click Close.
29. Create a SQL Server login for the ds2user (see the Configuring the database server (DVD Store) section for the
specific script to use).
30. Copy the pre-created DVD Store backup to the specified backup VHD volume.
31. Click StartAll ProgramsMicrosoft SQL Server 2008 R2Configuration Tools, and click SQL Server
Configuration Manager.
32. Expand SQL Server Network Configuration, and click Protocols for MSSQLSERVER.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 20
Red Hat Enterprise Virtualization 3
21. 33. Right-click TCP/IP, and select Enable.
34. Download and install Microsoft SQL Server 2008 R2 SP1.
Installing the operating system on the RHEL VM
1. Mount the RHEL 6.2 DVD or ISO to the VM.
2. Boot the VM to the Red Hat Enterprise Linux 6.2 installation DVD.
3. Press Enter to install using graphical mode.
4. At the media test screen, select Skip, and click Next.
5. At the Red Hat Enterprise Linux 6.2 title screen, click Next.
6. At the Choose a Language screen, select English, and click Next.
7. At the Keyboard Type screen, select U.S. English, and click Next.
8. At the Installation Devices screen, select Basic Storage Devices, and click Next.
9. Enter the VM’s hostname, and click Next.
10. At the Time zone selection screen, select the appropriate time zone, and click Next.
11. Enter the root password in the Root Password and Confirm fields, and click Next.
12. At the type of installation screen, select Use All Space, check Review and modify partitioning layout, and click
Next.
13. At the partitioning layout screen, select lv_home, and click Delete. At the Confirm Delete warning, click Delete.
14. Select lv_root, and click Edit. In the Size field, enter the Max size displayed. Click OK.
15. Click Next. If a warning appears, click Write changes to disk.
16. At the Boot loader screen, click Next.
17. At the default installation screen, select Basic Server, Customize now, and click Next.
18. Installation will begin automatically.
19. At the Congratulations screen, click Reboot.
20. Log into the server as root to complete the OS configuration.
21. Type rhn_register to begin the registration on the RHN, and accept defaults, making sure to select your
appropriate region.
22. Follow the prompts to register the system with RHN, ensuring that the yum repository for RHEL 6.2 is added
during the process.
23. Run the command yum update and follow the prompts to install all available system updates.
24. Edit the file /etc/selinux/config and change the line SELINUX=enforcing to SELINUX=disabled
25. Install additional software:
# yum install tuned
26. Disable these unused daemons with the following script:
# for s in auditd autofs avahi-daemon bluetooth cpuspeed crond cups dnsmasq
fcoe firstboot ip6tables iptables irqbalance kdump libvirt-guests lldpad
mdmonitor netconsole netfs nfs nfslock ntpdate portserve postfix qpidd
restorecond rhnsd rhsmcertd rpcbind rpcgssd rpcidmapd rpcsvcgssd; do
chkconfig $s off
service $s stop
done
27. Ensure these services are enabled using the following script:
# for s in cpuspeed sshd sysstat multipathd tuned; do
chkconfig $s on
service $s start
done
28. Synchronize the time source.
a. Edit /etc/ntp.conf, adding the IP address of the domain controller as a time source.
b. Run the command chkconfig ntpdate on
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 21
Red Hat Enterprise Virtualization 3
22. c. Run the command chkconfig ntpd on
d. Run the command date, and ensure that the time is synchronized with the domain controller.
29. Configure the VM’s IP address on interface eth0.
30. Type the following command to restart networking to effect these changes:
# service network restart
31. Install guest tools or agents.
a. (VMware only) Install VMware Tools on the guest:
Open a console window. Select VM on the window title-barGuestInstall/Upgrade VMware
Tools.
Log onto the guest as root
Mount the CDROM device:
# mount –o ro /dev/cdrom /mnt
Untar VMware Tools into a temporary directory:
# tar –C /tmp –zxf /mnt/VMwareTools-8.6.0-425873.tar.gz
Run the install script and accept the defaults:
# /tmp/vmware-tools-distrib/vmware-install.pl
The installer will automatically load the NIC drivers, create a new initrd, and unmount the CD.
b. (RHEV only) Install the RHEV Agent via the latest RHEV Agent RPM, which in our case was rhev-agent-
2.3.16-3.el6.x86_64.rpm).
32. Create one pair of file systems for the DS2 databases, and another for backup purposes:
# fdisk /dev/<device ID> ; fdisk /dev/<device ID> ; fdisk /dev/<device ID>
# mkfs.ext4 -L log01 /dev/<device ID>1
# mkfs.ext4 -L data01 /dev/<device ID>1
# mkfs.ext4 -L backup /dev/<device ID>1
# mkdir –p /vol/pglog1
# mkdir -p /vol/ds2data1
# mkdir –p /backup
33. Edit /etc/ fstab to mount these file systems on boot:
/dev/<device ID>1 /vol/pglog1 ext4 defaults 0 0
/dev/<device ID>1 /vol/ds2data1 ext4 defaults 0 0
/dev/<device ID>1 /backup ext4 defaults 0 0
34. Mount these file systems, and reset the tuned profile:
# mount -a
# tuned-adm profile default
35. Reboot the system:
# shutdown –r now
Installing PostgreSQL on the RHEL VM
We used the PostgreSQL database server, version 9.1, as the database software on the RHEL VMs.
1. Download the rpm for PostgreSQL 9.1 from http://yum.postgresql.org/repopackages.php and upload the rpm
file to the VM.
2. Log onto the system as root.
3. Install the PostgreSQL database server and client:
# rpm -ivh <rpm name>
# yum install <name of installed postgresql package>
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 22
Red Hat Enterprise Virtualization 3
23. 4. Edit these two lines /etc/init.d/postgresql-9.1 as follows:
PGDATA=/vol/ds2data1
PGLOG=/vol/ds2data1/pgstartup.log
5. Run the command initdb to initialize the PostgreSQL database cluster for the first time.
6. Ensure database files now appear in /vol/ds2data1.
7. To move the database log files to the /vol/pglog1 mounted to the separate virtual hard disk use the following
steps.
a. Run service postgresql-9.1 stop
b. Remove the directory pg_xlog from /vol/ds2data1.
c. Run ln –s /vol/pglog1 /vol/ds2data1/pg_xlog
d. Run service postgresql-9.1 start
e. Ensure that the PostgreSQL log files appear in /vol/pglog1.
8. Tune the database engine. Modify the following lines of the postgresql.conf file:
checkpoint_segments = 10
wal_buffers = 16MB
shared_buffers = 2GB
effective_cache_size = 6GB
checkpoint_timeout = 300s
checkpoint_completion_target=0.9
9. Add one line to the PostgreSQL configuration file pg_hba.conf to permit SQL queries from the client network:
host all all <test bed IP subnet>.0/24 trust
10. Modify the listen_addresses line in the PostgreSQL configuration file postgresql.conf:
listen_addresses = '*'
11. Run service postgresql-9.1 restart to restart PostgreSQL with the new settings.
Configuring additional VMs on Red Hat Enterprise Virtualization Manager 3
1. In the RHEV-M console, right-click the Windows VM, and click Make Template.
2. Enter a name, description, select the Default cluster, and select the proper storage domain.
3. Uncheck Make Private, and click OK.
4. Repeat steps 1 through 3 for the RHEL VM.
5. Select VMs, and click New Server.
6. Select the proper template, with every third VMs being based on the RHEL template. For example, VM-001 and
VM-002 are based on the Windows template, while VM-003 is based on the RHEL template.
7. Type a name for the new VM.
8. Click Resource Allocation, and ensure that the Physical Memory Guaranteed parameter is set to 1 MB.
9. Select the proper storage domain. For RHEV, we spread the VMs virtual disk storage using a round robin
approach across the four storage domains.
10. Click OK.
11. Continue deploying each VM up to 42 total, modifying the setting outlined above. For every third VM, use the
RHEL VM for the clone image.
12. Ensure in each VM that the necessary virtual disks are all online.
13. Assign the proper IP addresses and hostnames to each VM.
14. For Windows/SQL Server VMs, modify the SQL Server hostname of each VM using the instructions provided by
Microsoft (http://msdn.microsoft.com/en-us/library/ms143799.aspx).
Configuring additional VMs on VMware vSphere 5
1. Log into the vCenter Server, which manages the host.
2. Right-click the first Windows VM, and choose Clone.
3. Name the new VM.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 23
Red Hat Enterprise Virtualization 3
24. 4. Choose the cluster, and select the host.
5. For the storage screen, choose advanced and direct the new virtual disks to the applicable datastores. Again, we
used a round-robin approach to spread the OS, database, log VMDKs and configuration files across our separate
datastores. For instance, VM1’s configuration files and OS VMDK were placed on datastore OS1, the log VMDK
on Log1, and the database and backup VMDKs on DB1. VM2 files and VMDKs were placed on OS2, Log2, and
DB2.
6. Choose to customize using the customization wizard. Save the clone details as a new customization
specification.
7. Continue cloning each VM up to 42 total, modifying the customization specification as necessary for IP
addressing and so on. For every third VM, use the RHEL VM for the clone image.
8. Ensure in each VM that the necessary virtual disks are all online, the hostname is renamed, and the IP
addressing was properly assigned by the customization wizard.
9. For Windows/SQL Server VMs, modify the SQL Server hostname of each VM using the instructions provided by
Microsoft (http://msdn.microsoft.com/en-us/library/ms143799.aspx).
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 24
Red Hat Enterprise Virtualization 3
25. APPENDIX F - CONFIGURING THE DATABASE (DVD STORE)
Data generation overview
We generated the data using the Install.pl script included with DVD Store version 2.1 (DS2), providing the
parameters for our 8GB database size and the database platforms on which we ran: Microsoft SQL Server and
PostgreSQL. We ran the Install.pl script on a utility system running Linux.
For our Windows VMs, we transferred the data files and schema creation files to a Windows-based system
running SQL Server 2008 R2 SP1. We built the 8GB database in SQL Server 2008 R2 SP1, and then performed a full
backup. We transferred that backup file to each Windows VM and used that backup file to restore on all Windows VMs
between test runs.
For our RHEL VMs, we used the available scripts from the DVD Store distribution to load the generated data into
PostgreSQL. We then stopped the PostgreSQL server, created a backup copy of the data files, and used those backup
files on each RHEL VM to restore the PostgreSQL environment between test runs. Specifically, we followed the steps
below:
1. We generated the data and created the database and file structure using database creation scripts in the DS2
download.
2. We transferred the files from our Linux data generation system to a Windows system running SQL Server and a
Linux system running PostgreSQL .
3. We created database tables, stored procedures, and objects using the provided DVD Store scripts.
4. We loaded the data we generated into the database. For data loading in SQL Server, we used the import wizard
in SQL Server Management Studio. Where necessary, we retained options from the original scripts, such as
Enable Identity Insert. For data loading in PostgreSQL, we used provided scripts from DVD Store.
5. We created indices, full-text catalogs, primary keys, and foreign keys using the database-creation scripts.
6. We updated statistics on each table according to database-creation scripts, which sample 18 percent of the
table data.
7. On each database platform, we created the necessary users.
8. For SQL Server, we created the necessary full text index using SQL Server Management Studio. For PostgreSQL,
we used the provided scripts.
For RHEL guests specifically, we made these additional adjustments.
1. Create the DS2 user on the OS on the Red Hat system
# useradd ds2user
2. Download the latest DVD Store distribution for PostgreSQL from http://linux.dell.com/dvdstore/ and upload the
file to the VM.
3. Unpack the DS2 distribution into /ds2.
4. Change directory to /ds2/postgresqlds2.
5. Run the shell script postgresqlds2_create_all.sh
6. Copy the database directories to /vol/ds2data/.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 25
Red Hat Enterprise Virtualization 3
26. Running the DVD Store tests
We created a series of batch files, SQL scripts, and shell scripts to automate the complete test cycle. DVD Store
outputs an orders-per-minute metric, which is a running average calculated through the test. In this report, we report
the last OPM reported by each client/target pair.
Each complete test cycle consisted of the general steps listed below. For each scenario, we ran three test cycles,
and chose the median outcome.
1. Clean up prior outputs from the host system and all client driver systems.
2. Drop all databases from all target VMs.
3. Restore all databases on all target VMs.
4. Shut down all VMs.
5. Reboot the host system and all client systems.
6. Wait for a ping response from the server under test (the hypervisor system), all client systems, and all VMs.
7. Let the test server idle for one hour.
8. Start the DVD Store driver on all respective clients.
We used the following DVD Store parameters for testing the virtual machines in this study:
<driver exe name> --target=<target_IP> --ramp_rate=10 --run_time=45 --
n_threads=10 --db_size=8GB --think_time=0.1 --detailed_view=Y
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 26
Red Hat Enterprise Virtualization 3
27. APPENDIX G – SERVER, AND STORAGE CONFIGURATION INFORMATION
Figure 6 provides detailed configuration information for the IBM System x3850 X5 server, and Figure 7 provides
configuration information for the IBM System Storage DS3524 Express storage array.
System IBM System x3850 X5 server
Power supplies
Total number 2
Vendor and model number Emerson™ Network Power 70071524-J000
Wattage of each (W) 875
Cooling fans
Total number 4
Vendor and model number Nidec® V60E12BS1C3-07 Panasonic® PBT-GF30-FR
Dimensions (h x w) of each 1-1/2” x 2-1/5” / 1-1/2” x 4-1/2”
Volts 12 / 12
Amps 1.04 / 1.65
General
Number of processor packages 4
Number of cores per processor 10
Number of hardware threads per core 2
CPU
Vendor Intel
Name Xeon®
Model number E7-4870
Stepping A2
Socket type Socket LGA1567
Core frequency (GHz) 2.4
Bus frequency 6.4 GT/s
L1 cache 32 KB + 32 KB (per core)
L2 cache 256 KB (per core)
L3 cache 30 MB (shared)
Platform
Vendor and model number IBM System x3850 X5
Motherboard model number IBM 88Y5351 Microprocessor Board (Type 7145)
BIOS name and version IBM 1.73
Operating Mode set to Performance Mode; CPU C-states:
BIOS settings
Disabled
Memory module(s)
Total RAM in system (GB) 256
Vendor and model number Samsung® M393B1K73DH0-YF8
Type PC3L-8500R
Speed (MHz) 1,066
Speed running in the system (MHz) 1,066
Size (GB) 8
Number of RAM module(s) 32
Chip organization Double-sided
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 27
Red Hat Enterprise Virtualization 3
28. System IBM System x3850 X5 server
Rank Quad
RHEV-H hypervisor
Name Red Hat Enterprise Virtualization Hypervisor 6.2
File system ext4
Kernel 2.6.32-220.4.2.el6.x86_64
Language English
VMware hypervisor
Name VMware vSphere 5 Enterprise Update 1
Build number 623860
File system VMFS
Kernel 5.0.0
Language English
Graphics
Vendor and model number ATI ES1000
Graphics memory (MB) 16
RAID controller
Vendor and model number IBM ServeRAID M5015 6.0 Gb/s SAS/SATA Controller
Firmware version 2.120.223-1467
Cache size (MB) 512
Hard Drives
Vendor and model number IBM 42D0677
Number of drives 4
Size (GB) 146 GB
RPM 15,000
Type 6.0 Gb/s SAS
Onboard Ethernet adapter
Vendor and model number Broadcom® 5709S
Type Integrated
Discrete quad-port Ethernet adapters
Vendor and model number Intel Ethernet Server Adapter I340-T4
Quantity 1
Type Discrete
Optical drive(s)
Vendor and model number IBM UltraSlim Enhanced SATA DVD ROM
Type DVD ROM
USB ports
Number 6
Type 2.0
Figure 6. Detailed configuration information for the server under test.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 28
Red Hat Enterprise Virtualization 3
29. Storage array IBM System Storage DS3524 Express
Arrays 1 x DS3524 tray with 3 x EXP3524 expansion trays
Number of active storage controllers 2
Number of active storage ports 2
Firmware revision 07.77.20.00
Disk vendor and model number IBM ST973452SS
Disk quantity 96
Disk size (GB) 73
Disk buffer size (MB) 16
Disk RPM 15,000
Disk type SAS 6.0 Gb/s
Figure 7. Detailed configuration information for the storage array.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 29
Red Hat Enterprise Virtualization 3
30. ABOUT PRINCIPLED TECHNOLOGIES
We provide industry-leading technology assessment and fact-based marketing
services. We bring to every assignment extensive experience with and expertise
in all aspects of technology testing and analysis, from researching new
technologies, to developing new methodologies, to testing with existing and new
tools.
Principled Technologies, Inc.
1007 Slater Road, Suite 300 When the assessment is complete, we know how to present the results to a
Durham, NC, 27703
www.principledtechnologies.com broad range of target audiences. We provide our clients with the materials they
need, from market-focused data to use in their own collateral to custom sales
aids, such as test reports, performance assessments, and white papers. Every
document reflects the results of our trusted independent analysis.
We provide customized services that focus on our clients’ individual
requirements. Whether the technology involves hardware, software, Web sites,
or services, we offer the experience, expertise, and tools to help our clients
assess how it will fare against its competition, its performance, its market
readiness, and its quality and reliability.
Our founders, Mark L. Van Name and Bill Catchings, have worked together in
technology assessment for over 20 years. As journalists, they published over a
thousand articles on a wide array of technology subjects. They created and led
the Ziff-Davis Benchmark Operation, which developed such industry-standard
benchmarks as Ziff Davis Media’s Winstone and WebBench. They founded and
led eTesting Labs, and after the acquisition of that company by Lionbridge
Technologies were the head and CTO of VeriTest.
Principled Technologies is a registered trademark of Principled Technologies, Inc.
All other product names are the trademarks of their respective owners.
Disclaimer of Warranties; Limitation of Liability:
PRINCIPLED TECHNOLOGIES, INC. HAS MADE REASONABLE EFFORTS TO ENSURE THE ACCURACY AND VALIDITY OF ITS TESTING, HOWEVER, PRINCIPLED
TECHNOLOGIES, INC. SPECIFICALLY DISCLAIMS ANY WARRANTY, EXPRESSED OR IMPLIED, RELATING TO THE TEST RESULTS AND ANALYSIS, THEIR ACCURACY,
COMPLETENESS OR QUALITY, INCLUDING ANY IMPLIED WARRANTY OF FITNESS FOR ANY PARTICULAR PURPOSE. ALL PERSONS OR ENTITIES RELYING ON THE
RESULTS OF ANY TESTING DO SO AT THEIR OWN RISK, AND AGREE THAT PRINCIPLED TECHNOLOGIES, INC., ITS EMPLOYEES AND ITS SUBCONTRACTORS SHALL
HAVE NO LIABILITY WHATSOEVER FROM ANY CLAIM OF LOSS OR DAMAGE ON ACCOUNT OF ANY ALLEGED ERROR OR DEFECT IN ANY TESTING PROCEDURE OR
RESULT.
IN NO EVENT SHALL PRINCIPLED TECHNOLOGIES, INC. BE LIABLE FOR INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES IN CONNECTION WITH ITS
TESTING, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. IN NO EVENT SHALL PRINCIPLED TECHNOLOGIES, INC.’S LIABILITY, INCLUDING FOR DIRECT
DAMAGES, EXCEED THE AMOUNTS PAID IN CONNECTION WITH PRINCIPLED TECHNOLOGIES, INC.’S TESTING. CUSTOMER’S SOLE AND EXCLUSIVE REMEDIES ARE
AS SET FORTH HEREIN.
Virtualization performance: VMware vSphere 5 vs. A Principled Technologies test report 30
Red Hat Enterprise Virtualization 3