Application Segmentation

Business applications are the principal target of attackers seeking access to an organization’s most sensitive information, and as application deployment approaches become more dynamic and extend to the external cloud platforms, the number of possible attack vectors is multiplying. This is driving a shift from traditional perimeter security to increased focus on detection and prevention of lateral movement within both on-premises and cloud infrastructure..

Most security pros and industry experts agree that greater segmentation is the best step that an organization can take to stop lateral movement, but it can be challenging to parse the various available segmentation techniques. For example, IT pros and security vendors alike often use the terms application segmentation and micro-segmentation interchangeably. There is, in fact, some overlap between these two techniques, but selecting the right approach for a specific set of security and compliance needs requires a clear understanding of the different ways in which segmentation can be performed.

What is Application Segmentation?

Application segmentation is the practice of implementing Layer 4 controls that can both isolate an application’s distinct service tiers from one another and create a security boundary around the complete application to reduce its exposure to attacks originating from other applications.

This serves two purposes:

  • Enforcing clear separation between the tiers of an individual application, allowing only the minimum level of access to each tier required to deliver the application functionality
  • Isolating a complete application from unrelated applications and other resources that could be possible sources of lateral movement attempts if compromised

Intra-Application Segmentation

It is a longstanding IT practice to separate business applications into tiers to improve both scalability and security. For example, a typical business application may include a set of load balancers that field inbound connections, one or more application servers that deliver core application functionality, and one or more database instances that store underlying application data.

Each tier has its own distinct security profile. For example, access to the load balancer is broad, but its capabilities are narrowly limited to directing traffic. In contrast, a database may contain large amounts of sensitive data, so access should be tightly limited.

This is where intra-application segmentation comes into play, as security teams may, for example, limit access to the database to specific IP addresses (e.g., the application server) over specific ports.

Application Isolation

The second important role that application segmentation can play is isolating an entire application cluster, such as the example above, from other applications and IT resources. There are a number of reasons that IT teams may wish to achieve this level of isolation.

One common reason is to reduce the potential for unauthorized lateral movement within the environment. Even with strong intra-application isolation between tiers in place, an attacker who compromises a resource in another application cluster may be able to exploit vulnerabilities or mis-configurations to move laterally to another cluster. Implementing a security boundary around each sensitive application cluster reduces this risk.

There may also be business or compliance reasons for isolating applications. For example, compliance with industry-specific regulations, such as HIPAA, PCI-DSS, and SWIFT security standards are simplified by establishing clear isolation of in-scope IT resources. This is also true for jurisdictional regulations like the EU General Data Protection Regulation (GDPR).

Application Segmentation vs. Micro-Segmentation

The emergence of micro-segmentation as a best practice has created some confusion for IT pros evaluating possible internal security techniques. Micro-segmentation is, in fact, a method of implementing application segmentation. However, micro-segmentation capabilities significantly improve an organization’s ability to perform application segmentation through greater visibility and granularity.

Traditional application segmentation approaches have relied primarily on Layer 4 controls. This does have value, but firewalls and other systems used to implement such controls do not give security teams a clear picture of the impact of these controls. As a result, they are time-consuming to manage and susceptible to configuration errors, particularly as environments evolve to include cloud services and new deployment models like containers.

Moreover, Layer 4 controls alone are very coarse. Sophisticated attackers are skilled at spoofing IP addresses and piggybacking on allowed ports to circumvent Layer 4 controls.

Micro-segmentation improves upon traditional application segmentation techniques in two ways. The first is giving security teams a visual representation of the environment and the policies protecting it. Effective visualization makes it possible for security teams to better understand the policies they need and identify whether gaps in policy coverage exist. This level of visibility rarely exists when organizations are attempting to perform application segmentation using a mix of existing network-centric technologies.

A second major advantage that micro-segmentation offers is greater application awareness. Leading micro-segmentation technologies can display and control activity at Layer 7 in addition to Layer 4. An application-centric micro-segmentation approach can do more than simply create a coarse boundary between application tiers or around an application cluster. It allows specific processes – and their associated data flows – to be viewed in an understandable way and serve as the basis for segmentation policies. Rather than relying solely on IP addresses and ports, micro-segmentation rules can white-list very specific processes and flows while blocking everything else by default. This enables far superior application isolation than traditional application segmentation techniques.

Balancing Application Segmentation with Business Agility

Application segmentation is more important than ever as dynamic hybrid cloud environments and fast-paced DevOps deployment models become the norm. The business agility that these advances enable are highly valuable to the organizations that adopt them. However, heterogeneous environments that are constantly evolving are also more challenging to secure. Security teams can easily find themselves facing a lose/lose proposition of either slowing down innovation or overlooking new possible security risks.

The granular visibility and control that application-centric micro-segmentation offers makes it possible to proactively secure new or updated applications at the time of deployment without added complexity or delay. It also ensures that security teams can quickly detect any abnormal application activity that slips through the cracks and respond rapidly to new security risks before they can be exploited.

For more information on micro-segmentation, visit our Micro-Segmentation Hub.

The Average Cost of a Data Breach, and how Micro-Segmentation can Make a Difference

In the US, the financial cost of a data breach is rising year on year. IBM’s Cost of a Data Breach Report, is independently conducted annually by the Ponemon Institute. This year, the report included data from more than 15 regions, across 17 industries. They interviewed IT, compliance, and data protection experts from 477 companies. As a result, the true average cost of a data breach is more accurate than ever.

Crunching the Numbers: The Average Cost of a Data Breach

According to the study, the average cost of a data breach in 2018 is $3.86 million, which has increased by 6.4% since last year’s report.

While the risk of a data breach is around 1 in 4, not all breaches are created equally. Of course, the more records that are exposed, the more expensive and devastating a breach will be. A single stolen or exposed data record costs a company an average of $148, while 1 million, considered a Mega Breach, will cost $40 million. 50 million may be reserved for the largest enterprises, but this will raise the financial cost to $350 million.

Beyond a Ransom: The Hidden Cost of Data Breach

Although many businesses worry about the rise in ransomware, the cost of a data breach is about much more than any malicious demand from a hacker could be. The true cost can be broken down into dozens of areas, from security upgrades in response to the attack to a drop in your stock price when word of the breach gets out. Research by Comparitech found that companies tend to see a stock price slide of 42% following a breach. Other costly elements of a data breach include Incident investigation, legal and regulatory activity, and even updating customers. These all contribute to the escalating cost when you fail to adequately protect your company against a data breach.

The Ponemon study found that the largest cost comes from customer churn. The US sees the highest cost in the world in terms of lost business due to a data breach, more than two times the average figure, at $4.2 million per incident. Most analysts put this discrepancy down to the nature of commerce in the United States. In the US, there is far more competition and choice, and customer loyalty is both harder to hold onto and almost impossible to retrieve once trust is lost.

Customers also have more awareness of data breaches in the US, as laws dictate they must be informed of any issues as they are uncovered. This kind of reputational damage is devastating, especially in the case of a Mega Breach. In fact, 1/3 of the cost of Mega Breaches can be attributed to lost business.

Of course, there is also the fear that even if you manage to recover from a data breach, the worst is not over. The IBM study found that there is a 27.9% chance of another breach in the following two years after an attack, making your company extremely vulnerable unless you can make considerable changes, and fast.

Preparing Your Business for the Average Cost of a Data Breach

The numbers don’t lie. The speed and impact of data breaches is something to which every company, no matter the size, should be paying attention. There are definitely ways to protect your business and to position yourself responsibly for the worst case scenarios.

According to Verizon, 81% of all breaches exploit identity, often through weak passwords or human error. Malware can piggyback onto a legitimate user to get behind a physical firewall, which is why most IT professionals agree that even next-gen firewalls are insufficient. To limit the potential repercussions of this, all businesses need to be employing a zero-trust model.

With micro-segmentation, perimeters can be created specifically for the protection of sensitive or critical data. This ensures that all networks are considered not trusted. Using a granular approach to limit communications, and tagging workloads themselves with labels and restrictions. Containment of attacks is built into your security from the outset, by limiting the attacker’s freedom of movement and restricting ability for any lateral movement at all. As the financial impact of a data breach rises with the amount of data records stolen, this is a significant weapon to have at your disposal.

Rapid Response Can Limit the Cost of Data Breaches

Efficiency in identifying an incident as well as the speed of the response itself has a huge impact. Rapid response can save money, as well as proving to your customers that you still deserve their trust. According to the IBM report, the average time it took companies to identify the data breach was 197 days. Even once a breach was detected, the average time to contain it was a further 69. When it came to a Mega Breach – it could take an entire year to detect and contain.

With micro-segmentation, the visibility is immediate. All communications are logged, including East-West traffic. This includes private architecture, cloud-based systems, and even hybrid solutions. The best solutions will offer alerts and notifications in case of any unusual behavior, allowing you to stop threats in their tracks, before any damage has been done.

The quicker this happens, the less financial damage will be done. In fact, on average, companies who suffered a breach that managed to contain it within 30 days saved more than $1 million over companies who couldn’t. The larger the breach – the more significant these savings are likely to be.

Ensure You’re Fully Armed Against a Data Breach

The complex nature of most businesses IT systems explains the growing threat of cyber-crime, and the increasing financial cost of lax security holding us all to ransom. Traditional security systems are not enough to ensure adequate protection from a data breach, or rapid detection and response in case the worst happens.

Micro-segmentation offers granular flexible security that adapts to your exact environment, detecting and limiting the force of an attack, and providing the visibility and response tools you need to keep your customers loyal.

Protecting your Business Against Attack Vectors and the Evolving Threat Landscape

Understanding Attack Vectors

An attack vector is the way that an adversary can gain unauthorized access to your network or devices. Over the years, there have been dozens of different attack vectors, many of which have adapted and evolved over time to cause harm or hold companies hostage. Today, networks and organizations are interconnected using both private and public clouds leaving the door ajar for attack vectors that are more sophisticated than ever. What should smart businesses look out for, and how can they protect themselves?

The Evolution of Cyber Attack Vectors

Traditionally, having hardened perimeter security was enough to protect data centers. Layers of security to detect and prevent a breach coming in or out of data centers meant that you could ward off attack vectors to your infrastructure and hardware, which was almost exclusively on-premise.

The Cloud and mobile solutions have changed all of this. The reality for data centers today is keeping data private and secure while running an environment that spans public, private and hybrid clouds. Companies now use a mix of compute resources: Containers, Serverless Functions and VMs. However hackers are not just targeting your compute resources, they are sneaking in via routers and switches, or storage controllers, and sensors. From this vantage point, attackers can then scale their attack, compromising an entire network with lateral movements and connected devices. The MITRE ATT&CK Framework is a great resource to dive deeper on the different initial access attempts¹.

As the way we access the internet changes, cyber attack vectors adapt their own designs right alongside. Assuming that we are plugging all the holes on the IT side is not enough. The human factor has always been a key vulnerability in the security scheme. It has become more prevalent with the advance in end-user technology in recent years. Smartphones are a good example of this. Mobile attack vectors are not something that any organization had to be aware of a decade ago, and now they are an ever-present reality providing an easy gateway into many organizations.

While most people know not to click on dangerous links that arrive via SMS from unknown numbers, and no longer fall prey to email phishing campaigns like unexpected warnings of your bank password being changed, new attack vectors come from unexpected places. The recent Man in the Disk attacks on Android devices are something no one could have anticipated. This malware relies on vulnerabilities in third-party application storage protocols that are not regulated by sandbox restrictions through Android². This careless use of external storage can lead to potential malicious code injection, or the silent installation of unrequested apps to the user’s device. From there, the journey of an attacker to leverage this access to a deeper data center one is very short.

As technology evolves, there are more ways than ever for bad actors to launch attacks. Smart devices and Cloud-solutions only serve to increase the number of platforms which can be used for malicious intent.

Which Attack Vectors Are the Biggest Threats Today?

Email and phishing schemes have been the attack vectors of choice for a large amount of malicious attacks over the past few years. However, as simple attacks are becoming more recognizable, more complex threats are increasingly in vogue. Worryingly, the trend in malware is a movement away from reliance on human error, to clever attack vectors that can strike without any conscious act by the user whatsoever³. Man-in-the-Disk was just one example of this.

Take Drive-by-Downloads. A user only has to visit a compromised website, and malicious code can be injected through their web browser. Once done, this can swiftly move laterally across a network. Mouse Hovering hacking is also growing, a technique that launches javascript when a user hovers over a link to see where it goes. This has been seen in familiar applications such as PowerPoint, showing that even what users consider to be ‘safe’ environments can be dangerous. Increasingly sophisticated attack vectors that can spread without a user’s knowledge or their initial action are only going to become more common over time. If these tactics are leveraged against a user with administrator access to your data centers, the results could be catastrophic.

Administrator access could be the weak link when it comes to keeping your data centers safe overall. By accessing admin privileges, adversaries have access to the most valuable information you store, and can therefore cause the most harm. It’s important to think about the way your business works in a crisis when you’re planning preventive security measures. Used in an emergency, local authentication options are often not logged in the same way as your admins usual activity, and the credentials may even be shared across workloads and hosts for the sake of ease of use.

As well as smarter attack vectors, the growth in threats such as file-less attacks show that attackers are getting better at learning how to cover their tracks. 77% of cyber-crime in the US last year used a form of file-less attack⁴. Research shows that this type of malware is ten times as likely to succeed as traditional file based attacks, and helps attackers stay well beneath the radar.

AI is also an area that is likely to be compromised in the near future, with many companies creating chatbots and machine learning tools as the customer-facing representative of their websites and apps. As virtual assistants are built by humans, they are subject to the same gaps that human knowledge has. Studies are beginning to show that AI has problems with hallucinations and recognition⁵. Let loose on customer data and processes, it’s easy to see how advanced malware may slip through the cracks.

More than ever, in preparation for the next stage of intelligent malware, companies need to secure their data centers effectively against the latest attack vectors.

How Can Businesses Protect Themselves from Cyber Attack Vectors?

Keeping your IT environment safe from the latest attack vectors means being able to detect threats faster, and with better intelligence.

This starts with visibility. Being able to identify application flows across your entire infrastructure means that you have granular visibility across your whole IT stack. Dynamic deception tactics automatically trap attackers, even when the end-user isn’t aware of what is going on under the surface. Reputation analysis instantly uncovers anything suspicious or out of the ordinary, from unexpected IP addresses and domain names to file hashes within application flows. Even new attack vectors are isolated in real-time, with mitigation recommendations so that incident response is streamlined.

Ring-fencing, the separation of one specific application from the rest of the IT landscape is one way that companies are limiting the reach of the latest attack vectors from their most sensitive data or valuable assets. This and other kinds of micro-segmentation allow your business to truly limit the attack surface of any potential breach.

There are a number of benefits to this. Regardless of operating system limitations, communication policy can be enforced at the layer 4 transport level as well as the Layer 7 process level. By segmenting your flows by the principle of least privilege, even if a breach occurs, you ensure that it is quickly isolated, and attackers are unable to make lateral moves or scale their intrusion any further. When micro-segmentation is enforced alongside breach detection and threat resolution, even new attack vectors can quickly become a known quantity, and are unable to pose real danger.

Staying Safe Against Future Cyber-Attack Vectors

The way that data is stored and transferred is dynamic in and of itself. Our methods and processes are always changing as the capabilities of the cloud and the hybrid nature of our IT environments continue to grow. In direct response, attack vectors will never stay the same for long, and hackers will always have new tricks up their sleeve to compromise the latest solutions and catch us unaware. As well as current attack vectors that take advantage of IoT devices and no-fault infiltration, predictions for the future include AI-driven malware and an increase in file-less malware attacks, allowing hackers to hide their activities from detection.

The only solution is true visibility of all your applications and workflows. Using this mapping alongside segmentation policy that controls communication flows can restrict attackers in their tracks at the smallest sign of an anomaly. Even against new or unknown attack-vectors, these tools enable true threat resolution that can protect your entire infrastructure in real-time.


1. https://attack.mitre.org/wiki/Main_Page
2. https://research.checkpoint.com/androids-man-in-the-disk/
3. https://churchm.ag/was-it-human-error/
4. https://www.securityweek.com/fileless-attacks-ten-times-more-likely-succeed-report
5. https://www.wired.com/story/ai-has-a-hallucination-problem-thats-proving-tough-to-fix

GuardiCore’s Journey from Vision to Best-in-Class Micro-Segmentation

Micro-segmentation as we know it today has gone through several stages in the last few years, moving from a rising trend for securing software-defined data centers to a full-blown cyber security technology and a top priority on the agenda of nearly every CISO.

Built on the vision of securing the hybrid cloud and software defined data centers, we started our journey in 2013, thinking how to solve what in our opinion was a huge challenge for a market that did not exist at that time. In this post we’ll share how we created the micro-segmentation solution that is considered the best on the market – from vision to execution.

2015: First steps towards segmentation

Throughout the second half of 2015, we started delivering our micro-segmentation methodology after realizing that understanding how applications communicate inside the cloud was the key to success and as such – must be addressed first. “You can’t protect what you can’t see” wasn’t coined by GuardiCore but was immediately embraced by us when we started planning our micro-segmentation solution. We started developing our visibility solution Reveal, a visual map of all the applications running in the data center, all the way down to the process level. Reveal allows you to view applications and the flow they create in real time while also providing historic views. For the first time, admins and security teams were able to easily discover the running applications, one by one, and then review relations between the application tiers. Early releases supported general data center topologies as well as Docker containers.

2016: Gartner names micro-segmentation a top information security technology

We launched our segmentation solution at the RSA conference 2016 with a big splash. Reveal gained a lot of coverage and was well received by security teams who were lacking the proper tools to see the application flows in their data centers. It was one of the hottest security products at RSA 2016 and for a good reason!

Important to note that when micro-segmentation was introduced in Gartner’s Top 10 Technologies for Information Security in 2016 time in June 2016, many security professionals were unaware of the concept. In that report Gartner stated that to prevent attackers from moving “unimpeded laterally to other systems” there was “an emerging requirement for microsegmentation of east/west traffic in enterprise networks”. Enthusiasm was then at its peak, micro-segmentation was widely covered in the media and conferences dealing with the technology abound.

2017: Micro-segmentation for early adopters

Micro-segmentation was gaining traction as one of the most effective ways to secure data centers and clouds, but organizations learned the hard way that the path to meaningful micro-segmentation was full of challenges. Incomplete visibility into east-west traffic flows, inflexible policy engines and lack of multi-cloud support were among the most cited reasons. Throughout 2017 market penetration was around 5% of target audience and micro-segmentation was far from being mainstream. Andrew Lerner, Research Vice President at Gartner, noted in a blog post that “Micro-segmentation is the future of modern data center and cloud security; but not getting the micro-segmentation-supporting technology right can be analogous to building the wrong foundation for a building and trying to adapt afterward”.

That year GuardiCore tackled these challenges head on and based on the feedback we received from our growing customer base, we added flexible policy management and moved on from using only 3rd party integration to add native enforcement at the flow and process levels. Customers were able to move from zero-segmentation to native enforcement in 3 easy steps, based on revealing applications, building policies and natively enforcing policies.

2018: Our solution takes complexity out of micro-segmentation

Today, micro-segmentation serves as a foundational element of data center security in any data center. According to a Citi group’s report, cloud security is the number one priority among CISOs in 2018, with micro-segmentation the top priority in plans to purchase in this category. Concentrated effort on the part of organizations from different industries has resulted in better understanding of the technology. This year we were able to deploy micro-segmentation across all types of environments, from bare metal to virtualized machines, through public cloud instances and recently to containerized environments.

So if you are planning a micro-segmentation project let’s talk. We can show you how to do it in a way that is quick, affordable, secure, and provable across any environment.

Lateral Movement Security

Security teams often focus significant effort and resources on protecting the perimeter of their IT infrastructure and tightly controlling north-south traffic, or traffic that flows between clients and servers. However, several major transformations in enterprise computing are causing east-west traffic, or server-to-server communication within the data center, to outgrow north-south traffic in both volume and strategic importance.

For example:

  • Traditional on-premises data centers increasingly use horizontal scaling techniques that employ large sets of peer nodes to service the requests of clients, rather than a simple north-south flow.
  • The emergence of big data analytics as an essential competency is also driving substantial growth of east-west traffic, as processing of large data sets distributed across many nodes is generally required.
  • The growing adoption of public cloud infrastructure makes the traditional concept of a network perimeter obsolete, increasing the importance of securing east-west traffic among nodes.

While many organizations remain heavily invested in perimeter security, they are often extremely limited in their ability to detect and prevent lateral movement within their data center and cloud infrastructure.

What is Lateral Movement?

Lateral movement is the set of steps that attackers who have gained a foothold in a trusted environment take to identify the most vulnerable and/or valuable assets, expand their level of access, move to additional trusted assets, and further advance in the direction of high-value targets. Lateral movement typically starts with an infection or credential-based compromise of an initial data center or cloud node. From there, an attacker may employ various discovery techniques to learn more about the networks, nodes, and applications surrounding the compromised resource.

As attackers are learning about the environment, they often make parallel efforts to steal credentials, identify software vulnerabilities, or exploit misconfigurations that may allow them to move successfully to their next target node.

When an attacker executes an effective combination of lateral movement techniques, it can be extremely difficult for IT teams to detect, as these movements often blend in with the growing volume of legitimate east-west traffic. The more they learn about how legitimate traffic flows work, the easier it is for them to attempt to masquerade their attacks as a sanctioned activity. This, combined with many organizations’ insufficient investment in lateral movement security, can cause security breaches to escalate quickly.

Assessing Lateral Movement Security

One fast, simple, and inexpensive step that organizations concerned about lateral movement security can take is to test how vulnerable their environment is to unsanctioned east-west traffic. GuardiCore Labs offers a free, open-source breach and attack simulation tool called Infection Monkey that can be used for this purpose.

Infection Monkey scans the environment, identifies potential points of vulnerability, and attempts predetermined attack scenarios to attempt lateral movement. The output is a security report that identifies the security issues that were discovered and includes actionable remediation recommendations.

Infection Monkey Warns of Danger of Lateral Movement

Visualizing East-West Traffic

Organizations seeking more proactive lateral movement security can begin by visualizing the east-west traffic in their environment. Once a clear baseline of sanctioned east-west traffic is established and viewable on a real-time and historical basis, it becomes much easier to identify unsanctioned lateral movement attempts.

This is one of the flagship capabilities of GuardiCore Centra. Centra uses network and host-based sensors to collect detailed information about assets and flows in data center, cloud, and hybrid environments, combines this information with available labeling information from orchestration tools, and displays a visual representation of east-west traffic in the environment.

Visibility for Lateral Movement

This added visibility alone delivers immediate benefits to organizations seeking a greater understanding of potential lateral movement risks. It also provides the foundation for more sophisticated lateral movement security techniques.

Improving Lateral Movement Security

Once an organization has a clear view of both sanctioned and unsanctioned east-west traffic in its data center and cloud infrastructure, it can use this information to take active steps to stop lateral movement. An optimal approach includes a mix of both proactive and reactive lateral movement security techniques.

Micro-Segmentation Policies

Once an IT team has visualized its east-west traffic, the addition of micro-segmentation policies can significantly reduce attackers’ ability to move laterally. Micro-segmentation applies workload and process-level security controls to data center and cloud assets that have an explicit business purpose for communicating with each other. When strong micro-segmentation policies are implemented, attempts at lateral movement that do not explicitly match sanctioned flows – down to the specific process level – can generate alerts to the security operations team or even be blocked proactively.

Detecting and Responding to Unauthorized East-West Traffic

While micro-segmentation policies significantly improve lateral movement security, it is important to complement policy measures with additional detection and response capabilities. In addition to providing information-risk alerts when policy violations occur, GuardiCore Centra can detect and respond to unauthorized east-west traffic by leveraging deception technology to monitor and investigate suspicious behavior within east-west traffic.

Deception

GuardiCore Centra applies deception technology to analyze all failed attempts at lateral movement and then redirect suspicious behavior to a high-interaction deception engine. The attacker is fed responses that suggest that their attack techniques are successful, but all their tools, techniques and exploits are being recorded and analyzed in a fully isolated environment.

Deception

This helps IT teams learn more about the lateral movement being attempted in the environment and assess how to best improve security policies over time.

A Growing Strategic Priority

While strong perimeter security remains essential, the transition from traditional on-premises infrastructure to hybrid-cloud and multi-cloud architectures is increasing the strategic importance of lateral movement security.

It’s essential for security teams to:

  • Gain ongoing visibility into their organization’s east-west traffic
  • Develop techniques for differentiating between sanctioned and unsanctioned east-west traffic
  • Implement controls like micro-segmentation to tightly govern infrastructure activity
  • Actively monitor for unauthorized lateral movement to both contain breaches quickly and continuously refine policies based on the latest attack techniques.

Organizations that move beyond perimeter-focused thinking and place greater emphasis on lateral movement security will ensure that their security measures remain in step as IT infrastructure becomes more dynamic and heterogeneous.

For more information about Micro-Segmentation, visit our Micro-Segmentation Hub