Security Operations Analysis r8pndl
Security Operations Analysis r8pndl
Security Operations Analysis r8pndl
0
Table of Contents
1
Chapter 1: Information Security Incident Response
By Mohamed Marrouchi
Introduction
Identifying and responding to data security incidents is at the center of security activities.
The group appointed to security operations is relied upon to monitor the organization's
advantages inside extension and respond to security events and incidents, including the
identification and examination of what might be considered indicators of compromise (IOC).
Setting up a SOC to oversee incidents stretches out to cover people, processes, and
obviously, technology.
2
The correct arrangements of steps to pursue and the gatherings to include rely upon the idea
of the incident. A run of incident-handling taking care of process pursues the list of steps
exhibited in the incident response (IR) timeline in the Figure. We should take a gander at
detection, which is the second stop.
2.Incident Detection
Detection alludes to the stage in which an occurrence is watched and revealed by people or
technology, and the process that handles the reporting angles.
For the process to be powerful, the accompanying must be documented and formalized:
-Identify the sources, for example, technology and people, that are in charge of detection and
reporting incidents response team.
-Identify the channels through which incidents response team ought to be accounted for.
-Identify the means that ought to be taken to acknowledge and process incidents response
team reports.
-Identify the prerequisites on people and technology for the process to work.
3
3.Incident Triage
Incident triage appears to the underlying moves made on a detected e vent that is utilized to
decide the rest of the remaining as per the incident reaction plan. The triage phase
comprises of three sub-phases: verification, initial classification, and assignment.
The triage phase is worried about answering different question, for example, the
accompanying:
4.Incident Categories
1 Unauthorized access This represents when and individual gains logical or
physical access without permission to a client
network, system, application, data, or other resource.
4
2 Denial of Service(DoS) This is used when an attack successfully prevents or
impairs the normal authorized functionality of
networks, systems, or applications by exhausting
resources.
3 Malicious code This identifies when there is a successful installation
of malicious software, such as a virus, w
orm, Trojan
horse, or other code-based malicious entity, that
infects an OS or application.
4 Scans/Probes/Attempte This includes any activity that seeks to access or
d access identify a client computer, open ports, protocols,
service, or any combination for a future attack.
5 Investigation This includes unconfirmed incidents that are
potentially malicious or anomalous activity deemed by
the reporting entity to warrant further review
All security incidents ought to be belonged to a category. The category number distinguishes
the sort of the incident and its potential kind of impact. The table demonstrates an example
rundown of categories that we use for categorizing incidents.
The incidents may have beyond what one category and that the classification can change as
the incident advances or as the examination of the incident unfurls new discoveries.
5.Incident Severity
Severity levels depend on the normal or watched impact of an incident. This is utilized for the
prioritization of the incident, considering the measure of assets that ought to be appointed,
and decides the escalation procedure to pursue.
Level Description
5
High Incidents that have severe impact on operations
Medium Incidents that have a significant impact or the potential to have a
severe impact on operations
Low Incidents that have a minimal impact with the potential for significant
or severe impact on operations
6.Incident Resolution
The lifecycle of an occurrence ought to in the long run lead to some type of resolution. This
may incorporate information examination, resolution look into, a proposed or performed
activity, and recuperation. The goal of this phase is to find the underlying driver of the
incident, while chipping away at containing the incident at the earliest stage conceivable.
The investigation and analysis phase include the exercises attempted by SOC and by
different groups with the end goal of:
6
Identifying unapproved get to endeavors to private information
Understanding the chain of incident that have prompted the security incident
The containment stage includes the activities performed to rapidly stop a security incident
from raising or spreading to different networks or systems
The procedure appeared in the figure is an example containment process. We may, be that as
it may, continue with containment previously or amid the incident examination.
The correct strides to pursue to contain a security incident shift contingent upon the idea of
the incident and business criticality of the asset. Instances of containment activities
incorporate the accompanying:
7.Incident Closure
Closing a security incident alludes to the destruction phase in which vulnerabilities that lead
to the event or incident have been shut and all the occurrence follows have been washed
down. The closure procedure likewise incorporates testing systems to guarantee that the
7
annihilation steps and controls were compelling and that the vectors utilized by the attack
don’t exist any longer or is insufficient. Predefined activities to consider incorporate applying
any last data about the event, its last classification, any outside warnings, and documenting
information about the incident.
8.Post-Incident
This is the "lessons-learned" stage in which us look to enhance the IR processes and ponder
other people, processes, and technology controls. Post-incident exercises will fluctuate
contingent upon the seriousness of the security incident. Important learning picked up from
security incident can be valuable to avert/moderate future incident as proactive
administrations, for example, improving security highlights of capacities inside protections.
9.SOC Generations
Our comprehension of SOC segments and expected services has changed after some time.
This is a reflection of the change in our view of the criticality of information assurance and
security activities. This change comes in light of the regularly changing security threat
landscape, notwithstanding our undeniably receiving formal information security standards,
requiring the foundation and management of a formal security activities model and audit
forms.
8
10.Conclusion
In outline, following an incident response course of events, the SOC group would deal with
various basic assignments, from occurrence identification to closure. Each phase of an
incident can have its very own process and include different gatherings inside the
association. It is significant that the arrangement, plan, and fabricate stages for occurrence
response be characterized, reported, and supported by the correct specialists inside the
association. incident response is tied in with timing, and the most noticeably terrible time to
make sense of duties and incident-handling care of process is during a functioning attack.
References:
● Gregory Jarpey and R. Scott McCoy (Auth.) - Security Operations Center Guidebook. A
Practical Guide for a Successful SOC (2017, Butterworth-Heinemann)
● Jeff Bollinger, Brandon Enright, Matthew Valites - Crafting the InfoSec Playbook_
Security Monitoring and Incident Response Master Plan-O_Reilly Media (2015)
9
● Joseph Muniz, Gary McIntyre, Nadhem AlFardan - Security Operations Center_ Building,
Operating, and Maintaining your SOC-Cisco Press (2015)
10
Chapter 2: An Overview of Knowledge Asset
Management for Cybersecurity
By Dana Winner
Collaboration is a major principle and a suite of behaviors that form the core of Knowledge
Assets Management (KAM) culture. That collaboration suite of behaviors include
dialog/conversation, teamwork, knowledge sharing, human capital development, respect for
intellectual property and other concepts that support knowledge development. These KM
principles and behaviors are critically important for any Learning Organization, such as a SOC.
11
in which “hacking” and being a “hacker” are cultural themes which elevate cyber criminals
(“hackers”) as a hero and discount the criminal character of that (“hacking”) behavior.
The growth of the Internet in the 70's and 80's, and even more so, the WWW in the 90's and
2000's, promised exponential growth of knowledge through greater access to explicit
knowledge through digitally documented information formatted as text, graphics, images and
voice. Hyperbole and hubris are rampant regarding the benefits that the Internet has and is
expected to provide. Mark Zuckerberg, CEO of Facebook, exemplifies that hyperbole and
hubris, claiming that “Connecting everyone to the Internet is...necessary for building an
informed community”. There is general acceptance of the idea that for the industrialized
world, and especially the liberal democracies, the economic growth accruing from Internet
use has been phenomenal. However, not everyone agrees that connecting to the Internet is a
guarantee of learning and development of knowledge. The matrix of the Internet, liberal
democracy, is threatened by abuse of its freedoms thanks to abuse of the Internet. The
benefits of the Internet are increasingly in question as the cost of cybercrime grows.
Nevertheless, for the largest economies, the Digital Economy contribution to economic
growth far outweighs the cost of cybercrime and smaller economies continue to compete for
their place in the Digital Economy. Consequently, there is currently little motivation to
fundamentally solve the cybercrime problem. We can expect the current trends and behavior
to continue until disrupted by extreme conditions, such as cyberwar or economic collapse.
Every analysis predicts increasing cybercrime; “the Rules of Engagement have broken”. As
we already know, cybercrime will continue to be unfairly externalized to individuals,
small-to-medium enterprises and smaller economies, especially as the "Internet of Things"
spreads its network of headless, unsecured devices. Every SOC is at the vanguard of this
struggle to make the Internet sufficiently safe and secure to fulfill the potential knowledge
growth of the Internet. This is a cultural revolution – for better or worse – globally and in
each and every SOC.
Looking back over the 70 years of Information Age evolution, contrary to popular opinion, it is
apparent that there are aspects of the IT culture that have severely deteriorated. The
earliest days of the Information Age were days of making computers do things that seemed
12
almost impossible – like landing humans on the moon 50 years ago when the Information
Age was merely 25 years old (Harvard Mark I, 1944). Those amazing deeds were
accomplished through humility, knowledge sharing, collaboration, teamwork and diversity.
The KM cultural values are often thought of as “feminine”, while the cybersecurity culture is
largely “masculine” (more about this later). In the formative era of the Information Age,
women invented software. Women were included in the IT workforce for their knowledge,
skills and abilities (KSA). Female engagement peaked in the '80's at 30-35% before
commercialization shifted the nexus of IT culture to Silicon Valley. These and other related
cultural values are in shorter supply in an era of information and cyber insecurity, or as Vint
Cerf has called it, a Digital Dark Age. If we want to win this struggle to fulfill the knowledge
development promise of the Internet, we need to regain the seminal cultural values of the
Information Age. Every SOC is well positioned to lead this cultural struggle, to rebuild the
cyberspace cultural (values, behaviors) superstructure to take full advantage of the
Information Age foundations built in the 40's-80's.
The cyberspace superstructure that we have built on that foundation is not sustainable.
Those people who are information and cyber security professionals know that we are in a “no
win” situation. We can only estimate risks and spend precious resources to mitigate those
risks, just as seismologists can warn and recommend regarding potential for earthquakes
and volcanic eruptions. By contrast to those forces of nature, the cyber risk that we are
struggling with is of our own making. Doing the same thing again and again and expecting a
different result is insane. We have created and continue to perpetrate the culture that
simultaneously builds and destroys our global information infrastructure.
The cyberspace culture is building with the dominant hand and destroying with the other
hand. When will the destruction become financially intolerable? We have to solve the
problem at the root; building on the culture and purpose that formed the Information Age
foundation. That KM culture is essential to an effective SOC. Developing cybersecurity
strategies, tactics and technologies does not solve the cyber risk problem. “Culture eats
strategy for breakfast”. We must change ourselves, individually; change our culture, our
13
values, our behavior in cyberspace. By changing ourselves we change the SOC. By changing
the SOC, we can drive the change in IT, the enterprise, the nation, the world.
Our first step in this cultural change is to remember the original purpose of our cyberspace
adventure. Purpose is an important driver of all principles, values and behavior. Let's step
back from our focus on technology, the data, the information. We need to remember why we
are managing technology, data and information. Our purpose is to support knowledge
creation and sustainability; know-how, competency, ability, skill and understanding that
leads to excellence in performance and productivity. This is why we manage information and
data using electronic digital technology. Clearly, we must drive information management, in
all of its aspects and components, to support our Knowledge Assets. The question is, “do we
have a framework within which to conduct a Knowledge Assets Management approach to
information systems and especially, information and cyber security?” The answer seems to
be “No.”
Beyond the anecdotal evidence in the twitter-sphere, as reflected in the opening paragraph,
we find further evidence in the formal information security frameworks, that the principles of
knowledge management and knowledge itself is, at best, assumed, and in reality, is mostly
an after-thought. For example, in the “ISO 27000:2018 ISO Information technology —
Security techniques — Information security management systems (ISMS) – Overview and
vocabulary”, the emphasis is not on knowledge or knowledge management. As might be
expected, the word count for “information” is high at 318. Another high word count in this
introductory document for the ISO27000 series, is “risk” at 174 repetitions. These high word
counts reflect the importance placed on explaining these concepts. The word count for
knowledge is 6.
Of course, there are many documents in the ISO27000 series and a thorough review might
produce a more positive reflection of the importance of knowledge and KM within the
context of ISMS, it is not unreasonable to suggest that the ISO27000 overview document
does not provide a satisfactory and useful explanation of a knowledge-driven approach to
building and sustaining the ISMS, which is a fundamental component of the SOC. So, what
framework is available to support knowledge-driven cyber security in the SOC?
14
The good news is that there is a relevant standard from the International Standards
Organization – ISO30401:2018 - Knowledge Management Systems Requirements – that can
support the application of a Knowledge Management System (KMS) to the SOC ISMS. What
are the primary KMS requirements that can be applied to the SOC ISMS?
ISO30401 describes in the introduction the problem posed by the common confusion
between knowledge and information and “for example the view that simply buying a
technology system will be enough for knowledge management to add value”. The ISO KMS
Requirements standard provides guidance regarding the following major components that are
unique to this document:
4.2-4.3 Knowledge Stakeholders and Inventory: The organization should identify knowledge
stakeholders who are best suited to defining the Knowledge Areas. Create a “list” of the
corporate Knowledge Areas, their values and priorities in relation to stakeholders.
4.4 KM System: The organization should formalize their approach to KM into a system of
people, policies, processes and tools. Sections 4.5-4.8 define the parts of the KMS
4.5 Knowledge Life Cycle: Knowledge change management supports the (SECI)
transformations of knowledge assets within the knowledge areas as it evolves through the
stages of the knowledge life cycle: acquiring, applying, retaining and managing knowledge.
4.7 KM Facilitators: The knowledge life cycle does not happen by nature. Some people are
natural in connecting, communicating and knowledge-sharing which are fundamental
aspects of collaboration, while other people are averse to all aspects of collaboration. Even
those who are adept at collaboration require guidance to ensure that their collaboration
15
supports the corporate strategic goals. For those people who are not natural collaborators,
talented coaches can teach attitudes and skills that facilitate most of the less talented
collaborators. Dialog and conversation are underestimated skills that require work to
develop, especially amongst people who are typically attracted to IT and cybersecurity.
4.8 Learning Organization Culture: create the values that promote knowledge creation and
sustainability. This is primarily the responsibility of KMS Leadership (Section 5). The leaders
establish Policies, Roles and Responsibilities (Sections 5.2-5.3).
Section 6.1 of the ISO30401 KMS Requirements defines the importance of planning and risk
management of the KMS in order to achieve the knowledge objectives. This includes the
development of needed resources, especially competency (Section 7) of responsible staff.
Sections 8 and 9 define operations and performance measurements. Throughout the
document is it clear that “documented information” is subsidiary to and supports
knowledge..
SOC Analysis – the subject of this book – is a Knowledge Asset Management Process that
creates and sustains the ISMS, i.e. the cybersecurity risk management body of knowledge of
an organization. The framework for a Knowledge Assets Management driven SOC, is to apply
the ISO30401 KMIS Requirements guidance to the SOC operations, and specifically to the
SOC analysis process. It is hoped that the justification for applying ISO 30401:2018 has
been so convincing that the reader is how eager for some practical implementation. That
response calls for a complete book, not just a chapter. However, there are three KAM
strategies that can be briefly described, which might be immediately helpful and effective.
The first of those three KAM strategies is to ensure that Knowledge Risk Management is the
foundation for Information and Cyber Risk Management. By placing the emphasis on
knowledge rather than information, data and technical infrastructure, it becomes possible to
understand which Knowledge Assets (of which the great majority are human capital: tacit
knowledge) can be valued and thereby prioritized for protection. Only then can the supporting
assets, including the information, data and technical assets, be valued and prioritized. In
16
other words, the value of information and data cannot be estimated except in the context of
the knowledge that they support.
By insisting that the leadership estimate the value of Knowledge Assets and conduct
Knowledge Risk Assessment, the CISCO will not engage in meaningless estimations of the
value of information and data assets, basing Information and Cyber Risk Evaluations on false
assumptions. Another benefit of this approach is to place the Human Factor and Human
Capital squarely at the center of all Information and Cyber Security strategies, as we all
know it should be. By emphasizing the Human Factor, funding of the most effective security
strategies and tactics will be more convincingly justified.
The second of those three KAM strategies that could be immediately helpful and effective is
to improve the conversation by changing the rhetoric. Words are powerful. Let's begin with
the words “hacker” and “hacked”. The original of the word “hacker” in modern slang means
someone who is “amateur”, a “hobbyist”, i.e. someone who is not professional and not
criminal. This word was first used regarding PC and Internet hobbyists and then stuck with
them as some of them evolved into criminal behavior. Now, we need to change our culture to
recognize that cybercrime is perpetrated by cyber criminals, not “hackers”. We need to
change our rhetoric to reflect the reality that what cyber criminals engage is in “crime” not
“hacking”. We need to stop referring to the people who fight the criminals as “hackers” or
“ethical hackers” or “White Hat Hackers” and give them the serious respect of calling then
cybersecurity professionals. The SOC is not a place for individualistic “hackers”. The SOC is a
team of cyber security professionals.
The third suggested KAM strategy is to increase diversity - which is proven to create new
knowledge - by attracting more women to the SOC. The great majority of cyber crime is
perpetrated by males. Insiders are the greatest cybersecurity vulnerability and IT insiders are
the biggest vulnerability. Women are by nature less likely to engage in risky behavior, and
therefore engage in cybercrime at a very low rate. As well as being more risk averse than
males, as a group, women are more socially-oriented, more prone to communication,
networking and information sharing. This is a useful generalization, but should not be the
basis for ruling out exceptions. Certainly, many men share these natural talents with women.
17
Nevertheless, the talents of women are a great addition to IT, and especially to the SOC
team. Make a qualified woman responsible for SOC KAM and increase the potential for the
SOC to become a Learning Organization.
These three strategic changes are merely a suggested start for how the SOC can become a
Learning Organization that is driven by and supportive of Knowledge Assets Management.
The SOC should start the cultural change internally, so that each member is recognized as a
Knowledge Worker (not a “hacker”) and the SOC becomes a Learning Organization providing
professional cybersecurity protection to the enterprise. The SOC should insist on making
know-how the supreme purpose of the SOC and then spread this change from the SOC to the
Enterprise, and then to the Community, the Nation and the World.
18
Chapter 3 Introduction to Security Operation Centers
By Prasannakumar B Mundas and Elyes Chemengui
Before building SOC for the company, it is important to understand the needs to have the
requirements and plans. Successful SOC is not so easy to implement without the support of
all people from all levels. SOC can be implemented with the basic rules and processes but it
should include a framework to have a continuous progress.
There are many obvious things which has to be followed by questioning ourselves. By
answering these questions we will give the overall idea which helps to cover the gaps using
SOC.
It is very important to understand known and unknown security issues from your organization
are facing business impacts. These can be affecting your company CIA (Confidentiality,
Integrity and Availability). If you have already experienced issues, then list them out or else
make standard use cases which will help you to detect and prevent the attacks.
List out the existing problems and impacts so that your implementation of SOC can resolve
it. It is good to think about future problems as well, but it is mandating to cover the existing
one.
19
Define your goals and visions which will support your business and compare your visions with
the Risks which might your organization faces. Set the new goals for limited time to
evaluate the progress.
Obvious thing is, it needs your precious time, people, process and product (Technology) and
governance. There are some things which your organization can handle, and some cannot. In
this case, needs to understand and list out those tasks/resources which can be outsourced.
What do we need?
Understanding the environment (Network), employees and resources which are needs to be
monitored. Then, Calculate the budget which can be allocated to SOC. Then plan, evaluate
and divide the budget for technologies and people to run the SOC without any diversions.
From the beginning, it is good to have a framework which will help to implement the process.
It is recommended to use ITIL framework (Life Cycle) which will help to define, design,
transiting and the execution of the process including continuous improvement. Now we will
see how the basic rule of SOC implementation starting with People.
The significant components of a SOC , appeared in the figure, are people, processes , and
technology.
20
People:
SOC requires people who work in all the levels of management operations. It is good to have
or adopt the governance model which will help to construct the team which will run the
operation smooth and efficient manner.
People are required with a different kind of knowledge and experience in the various security
domains including as below:
● Business.
● Compliance.
● Legal.
● Human Resource.
● Internal Audit.
● IT.
● Physical Security.
● Communication.
● Dedicated specialized security analyst team (Offensive and Defensive).
21
Below is the Organizational hierarchy for handling security which will cover all the levels of
management to track the security in all aspects. From the ITIL framework, we can make use
of the RACI model to have the design of Responsible, Accountable, Consulted and Informed.
The SANS Institute (www.sans.org) classifies the roles people play in a SOC into four job
titles:
● Tier 1 Alert Analyst – These professionals monitor incoming alerts, verify that a true
incident has occurred, and forward tickets to Tier 2, if necessary.
● Tier 2 Incident Responder - These professionals are responsible for deep investigation
of incidents and advise remediation or action to be taken.
22
● Tier 3 Subject Matter Expert (SME)/Hunter – These professionals have expert-level
skills in network, endpoint, threat intelligence, and malware reverse engineering. They
are experts at tracing the processes of the malware to determine its impact and how
it can be removed. They are also deeply involved in hunting for potential threats and
implementing threat detection tools.
● SOC Manager – This professional manages all the resources of the SOC and serves as
the point of contact for the larger organization or customer.
The core SOC team should be talented having deep technical knowledge in all the security
domains to detect, investigate and prevent security incidents. They should be capable of
analyzing a large volume of data and Core team should have escalation Matrix with defined
levels based on the responsibilities, capabilities, knowledge, and experience. Designations
can L1 (Junior), L2 (Senior), L3 (Lead) and Consultant and Manager. Additionally, it would be
great if there is a team formed and aligned properly. For example, the Red Team (Offensive
security), blue team (Defensive Security), Compliance Team and Risk Identity management
Team soon. These all teams work together to have the proper security to the organization
and the business.
The most important part of people in SOC is their roles and responsibilities. To understand
these, you just need to answer three simple questions.
Some people, they work for Mission, vision, and values, those are top-level executive officers
such as CEO and COO. Some people they work for Governance and Operations, those are all
CISO and Director of Security. Some people they work for an organization, roles, and other
responsibilities and they are SOC core teams such as Managers, Leads, Seniors and Juniors.
All people will have their own uniqueness and talents. These talents will be identified and
placed in required positions as mentioned in the Organization hierarchy. All people will work
using their skills, capabilities, communications, processes, technology, and tools. Every
23
person or the team should be working by maintaining the quality, efficiency, consistency and
team spirits.
Every person will have his defined set of tasks to do. SOC team is basically working majorly
on two categories and those are Operations and Service.
It is important to hire the required people based on the tool and work experience. It is better
to match technology which you are using and their experience. It might cut down your
training cost.
The day of a Tier 1 Analyst starts with observing security ready lines. A ticketing framework
is every now and again used to enable investigators to choose alarms from a line to
research. Since the product that creates cautions can trigger false alerts, one employment
of the Tier 1 Analyst may be to check that an alarm speaks to a genuine security episode. At
the point when check is set up, the episode can be sent to specialists or other security work
force to be followed up on, or settled as a false caution.
In the event that a ticket can't be settled, the Tier 1 Analyst will forward the ticket to a Tier
2 Analyst for more profound examination and remediation. On the off chance that the Tier 2
Analyst can't resolve the ticket, she will forward it to a Tier 3 Analyst with inside and out
learning and risk chasing aptitudes.
Process:
24
Most of the time SOC works on various processes which need to be followed by the team to
keep the operations smooth and tracked. Instead of having various processes, can have few
processes which are well defined and can be followed in a good way to make the operations
easy and efficient manner. Let’s see what all processes can be used to implement and make
the SOC successful.
● Incident response plans: Every SOC should have the process of how to detect,
investigate, remediation and prevention of threat.
● Playbooks for all types of Cyber Attacks: incident response plan should also contain
the process/steps/procedures to respond to a specific type of threat.
● Incident Escalation Matrix: Every response plan should be there in place and should be
escalated to the next level based on the criticality.
● Security policies and guidelines: Every process/task/procedure should have security
policies and guidelines to avoid future legal actions.
● Incident investigation report: Every occurred incident should be investigated and well
documented for further reference. It can be for the compliance or the investigation
reference.
● Change management: Keeping the change management process in place can avoid
confusion of incident which is monitored by the Blue team and incident response
team.
● Roles and responsibilities of a team and team members: It is best practice to define
all the roles and responsibilities of the team so that those can be useful when it is
required.
● Threat/Asset/Risk management: As an on-going process it is recommended to
maintain the Threat management (Defense in Depth)/Asset management and Risk
management process which helps blue to understand the monitoring environment.
● Continuous improvement processes: It is good to have other continuous improvement
processes such as regular review of security controls, awareness/Training programs
etc.
● The process to connect SOC with other teams such as DevOps, IT, Patch management
etc.
25
Technology:
It is the most confusing part of understanding the needs and choosing the proper technology
which suits for feature requirement and cost.
For choosing technology and products or tools, there will be experience required to measure
the product from all the perspective. There are few important measurement points need to
be noted before choosing the product and technology.
The requirement for your network: Decide what are all your network loopholes and then
decide whether you require controls or detectors. Better to use defense in depth model to
figure out the present challenges.
● Features: the present market will provide various products for each solution, it is
good to have more feature which will satisfy your basic needs and more can be
chosen based on the recommendations.
● Cost: Calculate the cost of the required device, technology which provides more
features and less cost. Adjust the budget or the device but choice should excellent.
There are some cases where you might require an Intrusion detection system but if
you get the intrusion prevention system in the same cost then go for it. Choosing
options among many will be confusion.
● Support: Excellent support for the technology or the product can help to use the
technology in an efficient manner for the best results. Getting support from the
vendor will give you overall details about products and ideas to use it as per the
requirements.
● Ease of use: Getting more featured product is good until the user know how to use
and it can be configured or handled in a good manner if it is easy to use or
self-guiding.
26
● Review: There are many ways you can review the product for choosing a good one.
Compare all the above key points to each device and definitely, you might get a good
product. The offline and online review can be considered.
What are the most devices/technology which is used by SOC for securing your network?
● System Level: Anti-Virus, Anti-Malware, DLP (Data Loss Prevention) and Disk locker
etc.
● Network level: SIEM, firewall, I DS, IPS, VPN, WAF, VAS, HoneyNet, Spam Filters, Web
filters etc.
● Physical level: Biometrics, Passcode tokens, and Security cameras etc.
After having all in place, just following the ITIL lifecycle can maintain continuous
improvements on the operations and security.
Summary
Security Operations Centers (SOC) are responsible for preventing, detecting, and responding
to cybercrime. SOCs consist of people following processes to use technologies to respond to
threats. There are four main roles in the SOC. Tier 1 analysts verify security alerts using
network data. Tier 2 responders investigate verified incidents and decide on how to act. Tier
3 SME/Hunters are experts and are able to investigate threats at the highest level. The
fourth role is the SOC managers. They manage the resources of the center and communicate
with customers. Customers can be internal or external. A SOC may be operated by a single
company or may provide services to many companies. Finally, although network security is
extremely important, it cannot interfere with the ability of the company and its employees to
fulfill the mission of an organization.
27
Chapter 4 Open source tools for Security Operations
By Prasannakumar B Mundas
As we know there are many things included for building SOC. From a technology standpoint,
it is very important to have open source tools for identifying threats as well as for cost
reduction. From the DID (Defense in-depth) standpoint there are many devices and
technologies that need to be used to build the SOC. As per the industry experience below is
the technologies can be used for building proper SOC to monitor the threats to detect the
anomaly to safeguard the company.
Mainly since most of the attacks come from external sources, it is very important to use
proper controls at the perimeter of the network. By using the open source products we can
reduce the cost of the product and support is not mandatory.
IDS/IPS: Intrusion detection system is very important which is required to monitor the traffic
for identifying or detecting the anomaly and attacks. Snort is one of the open sources
network-based intrusion detection/prevention systems which can perform real-time traffic
28
analysis with packet logging on internet protocol networks. Snort has 5 important
components which help to detect the attacks.
● Packet Decoder
● Preprocessors
● Detection Engine
● Logging and Alerting System
● Output Modules
By using the above components, Snort can detect the network-based attacks or probes
including operating system fingerprinting attempts, semantic URL attacks, buffer overflows,
SMB (Server Message Blocks) and stealth port scans. And, it can also detect web
application attacks such as SQL Injections,
Since Snort is just an engine it requires GUI for ease of use if you are not very familiar with
the command line, so it is good to configure Snorby as well requires normal web server
application such as Apache.
Snorby will be helpful in analyzing the alerts which are triggered by snort. It helps to see
alerts, alert matching criteria.
Tip: Make sure all signatures are updated for detecting and preventing emerging threats.
These can be open-source or paid signatures (Depends on a budget).
Vulnerability Scanner (OpenVAS): For being proactive security guy it is most important to
have vulnerability scanner so that scan and confirm whether any assets are running with
critical vulnerabilities which can lead to any security breach or attack. A vulnerability
scanner is a product which has various updated scripts which are useful to identify the
vulnerabilities in a system or applications. Performing regular scans on the systems
especially the external facing systems or systems which are connected to the internet and
patching those regularly.
29
Tip: For every update or deployment it is mandatory to make sure all the systems or
applications patched for existing vulnerabilities.
There are various tools which are open-source with limited licensing such as OpenVAS.
Regular update of NVT is useful to detect the emerging vulnerabilities.
An OpenVAS engine can use with GUI Greenbone and Barnyard database for populating results
in UI. It can scan all the system in the network and it is good to have authenticated scan
using domain credentials. Greenbone provides options for creating credentials, hosts, tasks,
and schedules in the user interface. For more details refer to http://openvas.org/
For effectiveness, it is better to use emerging tools such as caldera intelligent tool which
can be used to emulate the adversary behavior and this is developed by MITRE. For more
details please refer to
https://www.mitre.org/research/technology-transfer/open-source-software/caldera.
Vega (https://subgraph.com/vega/): Vega is a free and open source web security scanner
and web security testing platform to test the security of web applications. Vega can help you
find and validate SQL Injection, Cross-Site Scripting (XSS), inadvertently disclosed sensitive
information and other vulnerabilities. It is written in Java, GUI based and runs on Linux, OS X,
and Windows.
30
HoneyNet: Nowadays, attackers are getting smarter every day, so it is good to have honeynet
to see and analyze the attack patterns which are tried by the attackers to know and
defending. It is a very important technology which is mandatory to trick the attacker and
safeguarding the assets. You can use Honeynet as internal honeynet or external honeynet as
per the requirement. Just mimic the used services to avoid the actual attacks. HoneyNet
has mainly 4 components such as mentioned below.
HIDS (OSSEC): OSSEC: Open Source HIDS SECurity is a free, open-source host-based
intrusion detection system (HIDS). It performs log analysis, integrity checking, Windows
registry monitoring, rootkit detection, time-based alerting, and active response. It provides
intrusion detection for most operating systems, including Linux, OpenBSD, FreeBSD, OS X,
Solaris, and Windows. OSSEC has a centralized, cross-platform architecture allowing
multiple systems to be easily monitored and managed. OSSEC has a log analysis engine that
is able to correlate and analyze logs from multiple devices and formats.
Network Monitoring tool: Nagios Core, is a free and open source computer-software
application that monitors systems, networks, and infrastructure. Nagios offers to monitor
and alerting services for servers, switches, applications and services. For more details
please refer https://www.nagios.com/products/nagios-core/attachment/visibility-2/
Red team activities: It is good to use Kali Linux or Backtrack operating systems which are
having all the tools which are required for vulnerability and penetration testing.
31
Kali (https://www.kali.org/downloads/): Kali Linux is an advanced Penetration Testing Linux
distribution used for Penetration Testing, Ethical Hacking and network security
assessments. It is the net version of Backtrack.
Commando VM
(https://www.fireeye.com/blog/threat-research/2019/03/commando-vm-windows-offensiv
e-distribution.html). Commando VM is the new Penetration testing Opensource Virtual
Machine build on Windows Operating system with full of penetration testing tools inbuilt and
it was built by FireEye.
Forensic: For being more in forensic like malware analysis, and recovery, there are various
Microsoft tools and other open source frameworks as mentioned below:
Threat intelligence sharing platforms: Threat intelligence sharing platform play an important
role in detecting attacks and infections based on the indicators of compromises. It can be
integrated into your log management and network monitoring tools to have proper prevention
and detection in place.
32
MISP - Open Source Threat Intelligence Platform & Open Standards For Threat Information
Sharing and abbreviated as Malware Information Sharing Platform which built using various
tools technologies such as MySQL, PostgreSQL, Shell script and Python, etc.
Reference links would be more helpful for going further such as hardware and software
requirements. There are some of the open source tools which can be used based on the
requirement and SOC capabilities. These are shortlisted based on ease of use and industry
experiences. To run the SOC smooth and cost-effectively, these tools play important roles.
33
Chapter 5 Threat Intelligence
By Digit Oktavianto
Latest trend of threats in information security landscape become more sophisticated since
attacker now prefer and focus on targeted attack rather than random attack. The degree of
threat sophistication is expanding very rapidly and giving serious impact on individuals,
businesses and governments. As a result, cyber defense technique is becoming increasingly
significant to anticipate the evolving threat. Traditional defense mechanism such as
monitoring, detection and preventive program is not sufficient enough and most of the time
fails to detect new kind of attack using unknown malware variants and attack that perform
evasive maneuver. Cyber threat intelligence concept offer deep insight, provide the ability to
recognize and act upon indicators of attack and compromise scenarios in a timely manner.
Threat intelligence also described as evidence-based knowledge, including context,
mechanisms, indicators, implications and actionable advice, about an existing or emerging
menace or hazard to assets that can be used to inform decisions regarding the subject's
response to that menace or hazard.
Several questions that came up after the incident occur will be : what kind of attack, when
did it happen, where was it found, how can the attacker breach into our system, and who is
the attacker. Suddenly the entire organization become more reactive and seems to be
concerned about security issue after the incident occurred. But it is too late, the attacker
may have been hiding in our system for several days, weeks, months and collects all
sensitive information or steal confidential document from your company. The sooner you can
adapt to the latest type of attack, the more you can learn how to protect your system. The
threats to our national security and economic interests in the cyber arena vary in identity,
objectives, assets, and capabilities. Their range can stretch from disruption, to simple theft,
to taking down critical infrastructure, to disrupt government functions. The advantage
34
almost always lies with the threat. Ability and intent of these actors become important
distinctions to the defender’s action [1].
To develop a serious plan for combating cyber criminal, company needs to understand the
basic capability and readiness of their infrastructure. Old approaches such as intrusion
detection systems, anti-virus, is no longer sufficient to combat the modern attack.
Organizations need to take a more proactive approach on their cyber defense mechanism
than reactive mechanism. Cyber threat intelligence has an important role to help
organization combating the cyber security threat. Cyber threat intelligence (CTI) is an
advanced process that enables the organization to gather valuable insights based on the
analysis of contextual and situational risks and can be tailored to the organization’s specific
threat landscape, its industry and markets[2]. Threat intelligence is not intrusion detection
system, not a firewall that can block attacks in real time, not an antivirus that match
signature the bad program and eliminate it from your system, but in fact threat intelligence
is a method that represents the collaboration of information and providing data on potential
threats with a deep insight of network structure, operations, and activities.
References
● Intelligence and National Security Alliance, "Cyber Intelligence: Setting the Landscape
for an Emerging Discipline," INSA Online Whitepaper, p. 17, September 2011.
● Ernst and Young, "Cyber Threat Intelligence - How to Get Ahead of Cybercrime," 2014.
35
Chapter 6: Incident Handling
By Mohamed Mouldi Chaouch
Introduction
Nowadays cyber threats are growing day by day, there is no way to protect the business
strategy of organizations rather than planning and preparing to prevent security incidents.
The incident handling process has several phases going from preparation to Post-incident
activity.
It is important to note the skills sets for the incident handling and incident response.
Incident Handlers need to be a good communicator and it is necessary to have project
management talents. Incident Response must have good knowledge of technical parts like
networking, malware analysis, forensics, etc.
1. Preparation
The preparation phase involves establishing and training an Incident Response Team and
acquiring the necessary tools and resources.
This phase as its name implies deals with preparing a team to be ready to handle an incident
at a moment’s notice.
There are several key elements to have implemented in this phase in order to help mitigate
any potential problems that may hinder one’s ability to handle an incident.
36
+ Incident handler communications and facilities:
● Contact information.
● On-call information
● Incident reporting mechanisms
● Issue tracking system for tracking incident information, status, etc.
● Smart-phones.
● Encryption software.
● War room for central communications among team members.
● Secure storage facilities.
● Port lists.
● Documentation.
● Network diagrams and lists of critical assets.
● Current baselines.
● Cryptographic hashes.
37
● Access to images of clean OS and application installations for restoration and
recovery purposes.
b - Preventing Incidents:
● The goal of this phase is to prevent overwhelming the incident response team with
incidents; because this can lead to slow and incomplete responses.
● The incident response team can be advocates of sound security practices, but they
are not responsible for securing resources.
● The team can play a key role in risk assessment and best practices for securing
networks, systems, and applications.
This phase is to determine whether the incident is really occurring and analyze its nature.
a - Attack Vectors:
Incidents can occur in countless ways, so to facilitate the work we must define a basis of
attack vectors to provide a classification for incidents, then go through it to handling more
specific attacks, we mention some of it ;
● External/Removable Media.
● Attrition (brute force attack against an a
uthentication mechanism, impair or deny
access to a service or application).
● Web (XSS, SQLI).
● Email.
● Impersonation.
● Improper Usage.
● Loss or Theft of Equipment.
b - Signs of an Incident:
38
The major challenge for the response team is the accuracy in detecting incidents, there are
three factors which make this so difficult:
● Multiple ways to detect incidents (Automated detection (IDPSs, log analyzers, etc.)
and manual detection (problems reported by users)).
● The volume of potential signs of incidents is very high.
● Deep & specialized technical knowledge.
Signs are identified using many different sources, the common are computer security
software alerts, logs, publicly available information, and people.
+ Precursor: Precursor is a sign that an incident may occur in the future. Examples of
precursors are:
+ Indicator: Indicators is a sign that an incident may have occurred or may be occurring now
like:
c - Incident Analysis:
Incident detection and analysis will be conducted efficiently when all of our precursors and
indicators are accurate, unfortunately, this is not the case because of different reasons:
39
● Incorrect indicators provided by users.
● Correct indicators, but has other errors than a security incident like human errors.
So these are recommendations for making incident analysis easier and more effective:
All these recommendations can aid the team to determine the incident's scope, the origin of
the incident, and how the incident is occurring.
d - Incident Prioritization:
40
● Functional Impact of the Incident: Incident handlers should consider how the incident
will impact the existing functionality of the affected systems and the future
functional impact if the incident is not immediately contained.
● Information Impact of the Incident: Incident handlers should consider how this
incident will impact the organization's overall mission.
● Recoverability from the Incident: Incident handlers should consider the effort
necessary to actually recover from an incident.
Combining the two first impacts determines the business impact of the incident, for
example, a DDOS attack against a public web server may temporarily attrition the service,
whereas unauthorized root-level access to a public web server may result in the exfiltration
of personally identifiable information (PII).
For the recoverability from the incident, the team may intervene when there is a high
functional impact and low effort to recover from this.
The organization can rate the incidents to better prioritize, this is are tables to rate the
factors:
41
Organizations should also establish an escalation process for those instances when the
team does not respond to an incident within the designated time. People may have personal
emergencies. The escalation process should state how long a person should wait for a
response and what to do if no response occurs. Generally, the first step is to duplicate the
initial contact. After waiting for a brief time the caller should escalate the incident to a
higher level, such as the incident response team manager. If that person does not respond
within a certain time, then the incident should be escalated to a higher level of management.
This process should be repeated until someone responds.
3. Containment
This phase is to contain the breach to stop spreading and cause further damage to your
business.
42
Organizations should create separate containment strategies for each major incident type,
with criteria documented clearly to facilitate decision making. Criteria for determining the
appropriate strategy include:
Evidence should be collected according to procedures that meet all applicable laws and
regulations that have been developed from previous discussions with legal staff and
appropriate law enforcement agencies so that any evidence can be admissible in court.
43
4. Eradication
Eradication is the way to eliminate components of the incident, the team must go through
identifying and mitigating all vulnerabilities that were exploited, identifying all affected
hosts, systems should again be hardened and patched, and updates should be applied.
5. Recovery
Take the system through any validation process you have before putting it into production.
6. Post-Incident Activity
This is where you will analyze and document everything about the breach.
44
This form should contain the contact details of the organization's:
● CISO/CIO.
● SPOC of: the incident handling or CSIRT team.
● Legal department contact.
● Public relations contact.
● ISP SPOC.
● Local cyber-crime unit.
b - Incident Detection:
c - Incident Casualties:
d - Incident Containment:
● Isolation activities per affected system (date and time the system was isolated, way
of system's isolation)
45
● Back-up activities per affected system (handler that performed the restoration,
back-up details, etc.)
e - Incident Eradication:
This form should contain information (one form per affected system is advised) such as:
This form can be extremely helpful in improving security measures and the incident handling
process, it provides a reference that can be used to assist in handling similar incidents, and
these reports are good material for training new team members.
At the end, these are the major steps to be performed in the handling of an incident it may
vary based on the type of incident.
1.4 As soon as the handler believes an incident has occurred, begin documenting the
investigation and gathering evidence
2. Prioritize handling the incident based on the relevant factors (functional impact,
information impact, recoverability effort, etc.)
46
3. Report the incident to the appropriate internal personnel and external organizations
6.3 If more affected hosts are discovered (e.g., new malware infections), repeat the
Detection and Analysis steps (1.1, 1.2) to identify all other affected hosts, then contain (5)
and eradicate (6) the incident for them
7.3 If necessary, implement additional monitoring to look for future related activity
+ Post-Incident Activity
Summary
47
This chapter describes the best practices when we go through an incident, how to Prepare
for it, know what to do when it happens, and learn all that you can afterwards
48
Chapter 7: Threat Hunting
By Ali Ahangari
Imagine you are familiar with the most common techniques for threat hunting. As a starting
point, you need suitable data to start the hunting process. In this post, I will explain the
most common datasets every hunter would need them. These datasets fall into two main
categories: Network Data, Endpoint Data.
Network Data
49
● Network session data
● IDS/IPS logs
IDS/IPSs can collect connection-based flow data and application protocol metadata (HTTP,
DNS, SMTP).
● Proxy logs
HTTP data that contains information on outgoing web requests, including Internet resources
that internal clients are accessing.
● DNS logs
Contains data related to DNS domain resolution activity, including domain-to-IP address
mappings and identification of internal clients making resolution requests.
● Firewall logs
Connection data that contains information on network traffic at the border of a network,
focused on blocked connections.
Endpoint Data
50
Contains information on processes run on specific hosts. Critical metadata associated with
process execution includes command-line commands/arguments and process filenames and
ID.
Contains data related to registry objects, including key and value metadata.
● File data
Information on stored file and artifacts kept on a local host. This can include when files were
created or modified, as well as size, type, and storage location information.
Possible Indicators in Endpoint Logs
Endpoint Logs may consist of many data sources including, but not limited to, Anti-Virus
Logs, Windows Logs, HIDS Logs.
51
But what should hunters look for in Endpoint Logs? In a sense, what threats can we find in
Endpoint logs?
You, as a threat hunter, may look for the following indicators in the logs:
52
● Suspicious Group Policy Change
● Suspicious Windows Firewall Changes
● Remote File Copy
● SSH Hijacking
Before doing so, It's necessary to mention that Anti-Viruses sometimes can not detect
malwares but they record logs p rocess calls, changing files, and so on. This is where threat
hunting plays a key role, detecting possible threats from what AVs didn't.
Now, I want to explain items that you can find in AV logs. These items include but not limited
to:
● Identifying known password dumpers, droppers and backdoors (Both Deleted and not
deleted)
● Detecting execution of binary from users APP Data directory
● Identifying process launching without parent process or services (e.g. Svchost.exe
launching without Services.exe being its parent process)
● Identifying scheduled jobs to perform known malicious behavior
● Investigating alerts for executable on web or application server
53
● looking for process launch from odd directory locations
By doing so, you, as a hunter, can find valuable things that AVs didn't.
Respondents provided brief descriptions of their threat hunting processes. Here are some
examples of good processes:
54
These are examples from respondents that are just performing intrusion detection - not
threat hunting:
Common Misunderstandings About Threat Hunting
According to Sqrrl, there are 3 common mistakes when it comes to threat hunting.
Hunting is not a reactive activity. If the main human input in a hunt is remediating the result
of something that a tool automatically found, you are being reactive and not proactive. You
55
are resolving an identified potential incident, which is a critically important practice in a
SOC, but not hunting.
Hunting requires the input of a human analyst and is about proactive, hypothesis-based
investigations. The purpose of hunting is specifically to find what is missed by your
automated reactive alerting systems. An alert from an automated tool can certainly give you
a starting point for an investigation or inform a hypothesis, but an analyst should work
through an investigation to understand and expand on the context of what was found to
really get the full value of hunting. To put this another way, hunters are the network security
equivalent of beat cops; they search for anomalies by patrolling through data, rather than
investigating a call in from dispatch.
2. Hunting can only be carried out with vast quantities of data and a stack of advanced tools
Though it may seem like a new term, security analysts across a variety of sectors have been
hunting for years. Basic hunting techniques can still be very useful and effective in helping
you find the bad guys (e.g. you can perform basic outlier analysis, or “stack counting”, in
Microsoft Excel). An analyst who wants to begin threat hunting should not hesitate to dive
into some of the basic techniques with just simple data sets and tools. Take advantage of
low hanging fruit!
3. Hunting is only for elite analysts; only the security 1% with years of experience can do it
There are many different hunting techniques that have differing levels of complexity.
However, not all these techniques take years to master. Many of the same analysis
techniques used for incident response and alert investigation and triage can also be
leveraged for hunting. The key to getting started is simply knowing what questions to ask,
and digging into the datasets related to them. You learn to hunt by doing it, so if you’re an
analyst who has never hunted before, don’t be afraid to dive in.
Ref: https://sqrrl.com/media/Your-Practical-Guide-to-Threat-Hunting.pdf
56
A Sample Scenario For Threat Hunting
Consider a company that hasn’t recently identified any malicious behavior on the internal
networks. Hunter decides to proactively hunt on the networks. In this sample, he/she uses
the organization’s most recent cyber risk assessment report to determine priorities and
select the assets to hunt on first.
• Preparation
• Investigation
• Adversary Removal
• Synopsis of the Hunt
• Reporting
Preparation:
Before proactive hunting starts, the hunter needs to decide which assets should be the
initial focus of the hunt. This can be gained from cyber risk assessment report of the
company. The hunter reviews the highest-priority group of IT functions in the report and
translates those functions, which are described from a business perspective (e.g., “credit
57
card processing systems”), into the corresponding IT asset information that the hunt team
needs: IP addresses, hostnames, process names, usernames, etc.
The hunter asks for the company’s latest IT network maps and asset inventory information.
He will need to confirm as needed which assets of interest are currently connected to the
network and to get more information on these assets, such as knowing which major
applications each asset is authorized to run, which assets have used the organization’s
networks in the past and what software they should be running.
A hunter who knows what’s normal for an organization’s assets and networks will be much
better prepared to spot deviations from the norm. the hunter is new to the company, so it’s
worth spending time talking with system administrators, incident responders, and other IT
staff members to learn more about normal activity. This can be as simple as knowing what
the typical work days and hours are for people in different roles within the company (e.g.,
standard users, managers, developers, system administrators.) This can also be complex,
including gathering detailed information on which assets people in each role may access,
and which applications are whitelisted (allowed) or blacklisted (prohibited) on the
company’s assets.
For operational reasons, the company might choose an on-demand deployment strategy or
Enterprise-wide deployment. We suppose the company has chosen the first one. The hunter
is responsible for ensuring that the hunt sensor software is deployed to all the assets
included in the initial round of hunting and also configuring the default settings for the hunt
sensors.
● Investigation:
58
So far, the hunter has completed all the preparatory activities. It’s time to investigate the
selected assets for the presence of adversaries. This investigation is documented in four
parts:
The hunter needs to examine the full path of the executable for each currently running
process and service, to include the executable’s filename. Questions to be answered include:
1. Should this executable be running on this machine? Does its full path seem reasonable?
For example, is the executable running from the appropriate system or application directory?
3. Was the executable run from the command line? If so, what were the command line
arguments?
4. Does the code in the running process or service exist in a file stored on the asset’s disk?
If no, why doesn’t such a file exist?
59
5. If the code is in a stored file, does the file’s hash match the vendor’s hash for the file?
Does this file exist on other assets, is it in the same location, and does it have the same
hash as the copies on the other assets?
6. Which network connections are bound to the process or service? For each connection, the
IP addresses and ports used by both endpoints must be noted.
7. Does it appear that all the loaded modules (executables and DLLs) indicated by the
Process Environment Block (PEB) and import tables for a particular process or service are
appropriate?
8. Are all registry keys associated with the process or service in the correct locations?
9. Do all open file handles (the files that the process or service is reading from and/or
writing to) make sense?
The hunter also needs to examine the available information about processes and services
that were recently executed. Such information is available from the "Prefetch" folder and the
registry keys for Application Compatibility Cache (ShimCache).
This information must be evaluated to see if the processes and services run on the asset
make sense. For example, did the full path of the executable seem reasonable?
● Registry auto-run keys. The purpose of these keys is to list executables to be
automatically run in the future, such as when the asset is rebooted or when a local
user logs in.
60
● Scheduled tasks. Users can schedule a particular executable to run in the future and
designate the date and time when it will be started.
At this point, the hunter has defined the scope of the investigation but hasn’t yet collected
the necessary information within that scope. The company’s hunt sensor software can
automatically do this work on the hunter’s behalf, speeding up the hunt considerably.
Remember three key concepts to keep in mind:
The hunter quickly discovers that the primary web server has a running process that seems
suspicious. The hunter also takes a closer look at the compromised web server. Additional
information to be obtained includes the following:
● The usernames for all users currently logged into the server
● The open file handles for all suspicious processes and services executing on the
server
● A copy of each file stored on disk that’s related to the suspicious processes and
services
● A memory dump for each suspicious process or service
● A packet capture of all network traffic for the server
While the hunt sensors are collecting and analyzing more information, the hunter receives an
alert from the hunt sensor on the company’s most critical database server. This alert
indicates the presence of an unknown persistent service on that server, so he immediately
reprioritized hunting to focus on the database server.
61
Throughout the hunt, hunters must constantly reprioritize their focus based on updated risk
assessments, their understanding of the value of particular assets, and the security event
information provided by hunt sensors and other enterprise security controls.
A quick review of the current and recent hunt sensor monitoring within the database server
indicates that the server has an established, encrypted connection to an external IP address.
The hunter can see that a large amount of data is being transmitted from the database
server to the external IP address, and that the service associated with the network
connection is the same service that generated the unknown persistent service alert.
Adversary Removal
All evidence points to an adversary exfiltrate sensitive data from the company via the
compromised database server. The hunter must act as quickly as possible to stop the
compromise by disrupting the exfiltration communications. Possible methods for handling
this situation include the following:
● Suspend the thread or threads associated with the exfiltration. Pat must consider the
likelihood that the activity is malicious and the criticality of the service being
compromised by the thread.
● Kill the service associated with the exfiltration. The hunter would rather suspend the
thread than kill the service because the latter can cause a significant disruption to
operations. On the other hand, the entire service could be malicious, in which case
suspending the thread won’t be an effective form of remediation.
● Disable network access. Isolating the database server from other servers prevents
further exfiltration and blocks the adversary from accessing the database server. Of
course, this isolation also prevents all operational use of the database server, which
will cause major production outages.
In this case, the hunt sensor indicates that there’s a single malicious thread responsible for
exfiltrating data, and the service in question isn’t malicious. This makes the decision easy.
62
Pat orders the database server’s hunt sensor to suspend the malicious thread. This action
breaks the connection between the database server and the adversary while avoiding any
disruption of the critical services that the database server provides to the company’s
customers, employees, vendors, business partners, and contractors.
Let’s fast forward to the end of the hunt investigation and recovery actions related to the
compromised database server. Here’s a synopsis of the most noteworthy actions:
1. Based on information provided by the hunter and the database server’s hunt sensor, the
hunter searched all assets across the enterprise for the following:
a. Any network connections involving the same external IP address that the database server
was connected to and from which data was exfiltrated.
b. Any files with the same filenames and/or file hashes as the malicious files found on the
database server
c. Any registry keys matching the malicious registry keys found on the database server
3. The company’s incident response team forensically collected all files on the two servers
associated with the malicious services, and they initiated chain of custody procedures. The
files were turned over to the company’s malware reverse engineers for further analysis.
4. The hunter removed all of the malicious files (including executables and other DLLs) and
associated malicious registry keys from the compromised assets.
63
5. System administrators discovered that the compromised web server was missing a patch.
Without this patch, the server was susceptible to users executing arbitrary code. The
system administrators also determined that the web server and the database server had the
same username and password for their local administrator accounts. This is how the
adversary gained privileged access to the database server.
6. System administrators forced the change of all passwords for all local accounts on the
web server and database server to prevent future reuse of any credentials that were
compromised.
7. When the hunter finished hunting on the database server, the hunt sensor was removed
from the server in accordance with the company’s on-demand deployment strategy.
Hunt Reporting
The hunter documented the findings and results for the hunt on an ongoing basis, issuing a
separate report for each major compromise investigated through hunting. His report for the
database server compromise includes the following information:
● An executive summary that reiterates the most important points from the report
● The scope of the compromise, such as which business processes, IT assets, users,
data sets, etc. were affected, in what ways they were affected, and how long they
were affected
● The identity of the adversary, if known
● A timeline of the activities involving the compromise and the hunt investigation into
that compromise
● A narrative corresponding to the timeline, indicating what the findings were at each
step in the hunt process, which tools or techniques were used for each step, who
received information from the hunters and when, what challenges or other issues
were encountered and how they were resolved (or why they aren’t yet resolved), etc.
● The root cause or causes of the compromise
64
● Recommendations for improving the company’s security and for making future hunts
more effective, and the consequences to the company if these recommendations are
not followed
Malware Beaconing: How To Hunt
One of the most common malware's behavior is ”beacon”, which implies that infected hosts
communicate to Command and Control servers at regular intervals that have relatively small
time variations. To be more specific, there are two key points about malware beaconing:
65
66
6. Finally, group every relevant attribute/field by URL, including the count of unique
source hosts and start checking the output.
Splunk Query:
| eval current_time=_time
| sort 0 + current_time
| eval diff_time=current_time-previous_time
67
| eventstats count, stdev(diff_time) AS std by src, url
68
Chapter 8 MSSP outsourcing
By David Nathans
Most organizations think about outsourcing cybersecurity due to its high costs and lack of
resources. Unless you are in the business of cybersecurity building your own SOC many not
have a high return on investment. Larger organizations see cybersecurity as a cost avoidance
(Saving money they would have to spend if they were breached) whereas smaller businesses
see it as a cost of doing business (money they need to spend to stay in business) if they
invest anything in cyber to begin with. Very different perspectives but the results are the
same. Not everyone can have their own Security Operations Center (SOC).
Before you can really think about MSSP outsourcing you need to understand what an MSSP is
and how different organizations use the acronym to describe themselves. Having a good
understanding of how organizations are different, yet call themselves the same thing will
help in getting closer to what you ultimately want in an outsourcing relationship.
● What is an MSSP?
● What is a Pure MSSP?
● What is an MSP?
● What to look for in an outsourcing relationship?
● Key questions not to ask?
● Setting expectations?
● Being a good partner?
What is an MSSP?
69
As defined by Gartner[1], a Managed Security Service Provider (MSSP) provides outsourced
monitoring and management of security devices and systems. This definition goes on to say,
“Common services include managed firewall, intrusion detection, virtual private network,
vulnerability scanning and antiviral services.”
This definition does not go far enough and could be considered outdated as there are many
more types of services and cross-over offerings available today. Here is an attempt to
provide a sample of examples opposed to an exhaustive list:
The key to any organization calling themselves an MSSP would be that their focus is
providing services that the name squarely states. They are a provider who M anages Security,
as a service. Long gone are the days where MSSP’s just provided managed firewall and
Intrusion Detection (IDS). Today, a key differentiator of an MSSP is that they offer a wide
range of security services backed by a 24/7 Security Operations Center.
● AT&T - https://www.business.att.com/portfolios/cybersecurity.html
● Capgemini - https://www.capgemini.com/service/cybersecurity-services/
● IBM - https://www.ibm.com/security/services/managed-security-services
● Secureworks - https://www.secureworks.com/
● Symantec -
https://www.symantec.com/services/cyber-security-services/managed-security-serv
ices
70
● Verizon -
https://enterprise.verizon.com/resources/articles/managed-security-services/
Typically, MSSP’s have been large businesses that provide a long list of services but have a
Managed Security Service (MSS) offering as either part of their extended internal operations
or as a separate business unit. This can be the lifeblood of large enterprise organizations
who need outsourcing due to the size and volume of needs. Large organizations need to
outsource to other large if not larger organizations to ensure that the provider can handle
their workload and have the resources to be successful for long periods of time.
Over the last few years, a new breed of MSSP has begun to grow and address a much
different market. The pure-play MSSP is an entire company whose primary and only focus is
on offering security services. These new types of MSSP’s also bring with them a new crop of
service offerings that are boutique, focused, and address specialty markets. Entire security
companies can be dedicated to the protection and security of the evolving landscape of
Internet of Things (IoT), Internet of Medical Things (IoMT), or specialize in industries such as
oil and gas to include Industrial Control Systems (ICS) and Supervisory Control and Data
Acquisition (SCADA) systems. These organizations also bring with it, expertise in protecting
small businesses, Government, cloud and more.
● Dragos - https://dragos.com/
● MedSec – https://medsec.com/
● Red Canary - https://redcanary.com
● Rapid7 - https://www.rapid7.com
● SOCSoter – https://www.socsoter.com
● FireEye – https://www.fireeye.com
There has been no better time than now to find a company that specializes in security
services who is as unique as the organizations with whom they are protecting. Not only can
71
you now find Pure MSSP’s who specialize in your industry but also bring with them products
and services unique to business size and offer price points that make sense. MSSP’s are
also bringing new capabilities to market such as Managed Detection and Response (MDR).
This is service that is typically provided by pure MSSPs and has a strong focus on threat
intelligence, threat hunting, security monitoring, incident analysis, log analysis, and incident
response through a SIEM or even a managed SIEM infrastructure. Using advanced security
analytics on the network and endpoints to track user behavior, application behavior and
mis-usage. MDR provides deeper detection by combining security analytics with machine
learning and augmented or artificial intelligence compared to traditional MSSPs, who mostly
rely on static rules and signatures. Endpoint Detection and Response (EDR) is another
emerging yet game changing technology pure MSSPs have been leveraging. Threats are
constantly changing, and anti-virus software is only able to stop some of the problems some
of the time. New EDR offerings by MSSPs include the capability to perform searches of all
endpoints in an organization or across all devices an MSSP manages leveraging threat
intelligence to find known and unknown threats, isolate compromised devices and stop
malicious activity instantly.
While relatively new in the way of MSSP offerings, MDR and EDR is proving to be valuable for
organizations looking to increase their cyber defenses. If your organization is looking to
improve its incident response and threat detection programs with a pro-active, hands on
approach to cyberdefense, an MSSP offering MDR and EDR could be a great way to achieve
these goals.
What is an MSP?
Over the last few years there has also been other changes in the technology business. Value
Added Resellers (VAR) of technology products and services, technology resellers and IT
break/fix companies, which traditionally operates on a transactional and short-term basis,
have all struggled as computing platforms move to the cloud and end users become more
technically proficient and self-supportive. The emergence of the Managed Service Provider
(MSP) has evolved to become the entrusted lifeblood of businesses who rely on technology
but cannot afford full time in-house staff to support the varied needs, or they can
72
supplement full time in-house staff that are overwhelmed with their responsibilities. The
MSP partners with their customers over annual, or multi-year periods. They can handle a
number of technical and administrative duties such as help desk support, network and
application management and ensuring that the systems are working seamlessly with as
little (ideally NO!) downtime as possible.
The first thing you need to do before you begin outsourcing is to look in a mirror. You have a
lot of choices out there and you can either shop for the best fit for your needs or be swayed
by slick marketing and stories of how great one company over another. The best
relationships are built on a firm ground of understanding what the business needs and what
it is getting from the outsourced relationship. If those two things are not well understood,
then the relationship is not going to work even with the best of intentions. A good
outsourcing relationship is not about getting an advantageous contract or getting more for
your money, but rather gaining a partnership that extends, enhances, and augments a
business’ ability to grow and benefit from that relationship. The MSSP partner you choose
should be willing to protect your business as if it’s their own and you should also be willing
to make them a key member of your team.
How to test the relationship? The difference between selling and recommending is trust. If
you trust your outsourcing partner and they make recommendations, then you do trust them.
73
If all you hear is them trying to sell you, then maybe you don’t trust them nor have a good
relationship.
Another thing you should look for in an MSSP relationship is innovation and enhancements
and if those come free or at a charge. Did the MSSP build proprietary technology and do all
their own inhouse development or are they just reselling other companies’ products and
services? In the case of a pure MSSP, cybersecurity is all they do. They live and breathe for
cyber and should always be improving their capabilities. Of course, MSSP’s need to make
money but they should have clear pricing models and set clear expectations on what you get
versus what is an extra charge.
Additionally, you need to see if the MSSP is going to be aligned with your needs not just from
an offering perspective but also from a compliance perspective. If there are country
requirements that don’t allow you to outsource outside your own country, if data backups are
going overseas. These and many more questions should be asked of an MSSP as you begin to
build a relationship with them.
The list of questions you can ask an MSSP are endless and there are many sites out there
that have an exhaustive list of things that you can ask. Instead it is more fun and beneficial
to explore questions that maybe you should not ask when looking for a new MSSP partner.
Contracting with a big-name player may be great. If you are a small fish then will you get the
attention you deserve? Let’s be real, you get what you pay for with the big guys. There is no
doubt that they will do everything they can to service your contract and provide you the best,
and you can be very happy with them. But let’s be honest, if your only spend $1,000 and
another company’s spends $10 million, who is going to get more attention? Do you really
need to ask this question?
Growth questions always seem to come up and never seem to make real sense at the time.
Asking questions like “what’s your staff growth rate?” is a waste of time for everyone unless
you are specifically looking for a job. Growth of a company is a distractor to servicing a
74
customer’s account. An MSSP should do all they can to ensure that any growth does not
upset their customer base. Besides, having a ton of employees and an impressive office
means they have big bills and need you to help pay for them.
Asking why they think their technology is the best. Is this a game? MSSP marketing
departments can drown you in whitepapers, marketing materials, social media and more. If
you want to ask technical questions then go for it. Don’t just ask questions to show off your
technical wizardry or play “stump the chump.” Make the questions meaningful because you
are interested in how they address a real problem. You want to know how they are going to
protect you, not how they are going to keep you entertained on twitter.
There are many more other questions you should probably not ask. Next time you get a blank
stare, bewilderment or long pause, stop and ask yourself it that was a good question and if
the answer really matters to establishing a good relationship with your MSSP.
Setting expectations
Previously we discussed looking in the mirror and understanding your needs before going out
to look for an MSSP. That is never truer than when looking to set expectations. If you are not
sure of what you want, then it will be impossible for any MSSP to make you happy. This does
not mean you need to be the most technical person in the room, but you should know what is
important to your company and what your company needs to protect. You should also clearly
understand your challenges in providing cybersecurity services yourself. You need to know if
it is cost issues, internal knowledge, resource or time problems that prevent you from doing
it yourself. You not only need to know what is important from a cyber protection perspective
but also understand and be able to communicate what they expected outcomes are. If you do
not have the capability to run a full incident response and containment of a breach, then you
need to ensure that your MSSP or even multiple MSSP’s can and will offer it to you if you
need it.
Setting expectations goes both ways. An MSSP should be willing to share with you their
development roadmap. Knowing what you want out of your MSSP is great but if you are able
to strengthen your cybersecurity along with the growth and development of new features and
75
capabilities of your MSSP then it will be easier to plan, budget and ultimately meet company
business objectives.
Lastly, every business is different and as such there are times when something custom will
be needed. Knowing what services, you are going to be contracted for, it would be
worthwhile to understand if there are any capabilities that could be customized. Not only can
they be custom but also if there is a charge for any customization.
You need to understand where the MSSP services begin and end and your responsibilities
begin and end.
The best relationships for MSSP and customers alike are ones where the customers are
engaged and the MSSP is responsive. The customer needs to be responsive and willing to
work collaboratively in cybersecurity efforts alongside the MSSP. This does not mean you
have to have a dedicated resource but at least someone who has ultimate responsibility to
be available to the MSSP when needed. It is always best to assign a dedicated resource to
work with the MSSP that is not the CEO. It is not that the CEO is a bad person, but they have
other stuff to do and may not be able to respond to issues or requests from the MSSP.
Obviously, you will need to decide in your own organization who is the best for this role but
don’t skip assigning it, it could mean the difference between an incident and a breach.
Understanding what are the depth of services and capabilities that the MSSP has is critical
for you to understand what they can do for you especially when it matters most. If all you get
is a notification of a potential incident without the ability to question, engage or invoke full
incident response services, then are you in an outsourced relationship or did you just buy a
product. MSSP’s and customers should have regular meetings and touchpoints to ensure
that everything is working smoothly. This should be a time that is set aside regularly to
review reports, plan new projects, discuss issues and provide feedback. Feedback is a
critical component to being a good partner. If you are happy or not, letting your MSSP know
76
will help them either continue doing what they are doing or make course changes to help
resolve issues.
Summary
While there are certainly similarities and difference between an MSP, MSSP, and pure-play
MSSP. It essentially boils down to the fact that MSP’s do not offer their own security in their
portfolio of services outside of infrastructure, Firewall, Patch, and anti-virus management
unless they partner with an MSSP or a cybersecurity vendor. MSSP’s offer focused
cybersecurity offerings as part of a larger organization, and pure-play MSSP’s only deal in
cybersecurity and even sometimes provide specialized cyber services in specific focus areas.
Regardless of who you choose and why, in the end, you need to focus your decision and
outsourcing efforts in achieving a few simple objectives. Find a partner that cares about you
and is skilled in your size and type company. Minimize costs by leveraging security experts
working for you at a fraction of the cost of having your own in-house team. Increase
efficiency by allowing your outsourcer to apply proven procedures and policies into the
management of your infrastructure that ultimately helps reduce false positives and improve
responsiveness to detect and contain cyber-attacks. And lastly, your MSSP needs to enable
you to focus on your business while they focus on protecting it.
References:
77