Cisco Tech Blog CiscoTech Blog Close
Home Contact

What is Zero Trust Security? 🔗︎

Cisco Cloud Native Security Solution

The “Trust no one, always verify!” motto is at the core of the zero trust security model and its popularity is growing as the zero trust framework addresses many of today’s common cybersecurity issues, such as:

  • ‘‘‘Security perimeters are becoming porous’'': as the need for connectivity between systems increases, maintaining a tight security perimeter is increasingly difficult.
  • ‘‘‘Cloud-native and hybrid infrastructure increase risks’'': one of the main advantages of migrating all or parts of a system to the cloud is its agility and flexibility. Traditional cybersecurity models are ill-equipped to secure a microservice infrastructure.
  • ‘‘‘Hackers are getting better’'': a growing variety of AI-assisted types of malware are now available off-the-shelf or as SaaS on the Dark Net, enabling malicious actors with limited tech-savviness to perform advanced attacks.
  • ‘‘‘Compliance breaches’ cost is skyrocketing’'': GDPR, HIPAA, and other regulations now impose a high threshold to meet their minimal requirements for securing data protection, imposing stiff fines for non-compliance, and opening the door to costly damage lawsuits in case of breach.

Why the Shift to Zero Trust Security Framework? 🔗︎

Several factors lead to the advent of the zero trust security model:

The Evolution of the Technology 🔗︎

Until the advent of microservices in the last few years, organizations’ digital infrastructure was based on a monolithic architecture. Network security was inspired by military strategy. Creating a security perimeter around the company’s virtual assets from data to code was the recommended strategy. Gateways were deemed the only access points where external packets had to be scanned, validated, and authorized to access the secure perimeter and interact with the protected assets. This became increasingly problematic as a hacker worming his way inside the perimeter could roam free, steal all the data, and manipulate or crash the entire system. To mitigate those risks, the secure perimeter was compartmentalized, attempting to limit the damage in case of breach by making it harder to move from one compartment to another within the secure perimeter. Yet, this still meant the data was hosted and coding was on-site. It could not handle the emerging inter-systems networking. When reliance on SaaS and other external virtual services and data centers became increasingly unavoidable, the risk of a hacker hitching a ride on an authorized source mushroomed. In parallel, the Bring Your Own Device (BYOD) rising trend and business necessity multiplied the number of potentially hackable devices and made it impossible to fully secure all devices gaining access to data centers and other services needed for organizations to function.

The Soaring Financial Costs 🔗︎

The costs resulting from a breach in a business system are steadily rising. The average cost of a data breach in 2020 is $3.86M globally, jumping to $8.64M for the USA. These figures cover the costs deriving from:

  • ‘‘‘Loss of business revenue’'’, both during the attack and resulting from tarnished reputation.
  • ‘‘‘Regulatory fines and legal costs’'': Data protection authorities are mandated to issue fines for non-compliance, such as the $124 million fine imposed on Marriott in 2019. This was dwarfed by the $575 million settlement agreed to by Equifax for the 2017 famous breach that exposed the personal identifying data of hundreds of millions of people.
  • ‘‘‘Incident Response’'': From containment to mitigation and recovery, handling a breach is costly in staff-hours. Furthermore, the lack of proper expertise to handle a breach leads to delay in finding the right expert and higher cost stemming from emergency fees.

However, they do not cover the costs resulting from loss of intellectual property, even though these might end up being the highest of them all.

Implementing the Zero Trust Security Model in Kubernetes 🔗︎

Kubernetes are the leader in the containerized architecture that is now the prevalent development paradigm. As per the zero trust security core tenants, it’s based on assuming that everything and everyone is a potential hacker or malware, so every interaction must be verified.

Zero Security Best Practices for Kubernetes 🔗︎

  • Applying security control to every single entity, whether software or hardware, and at every location. This includes exhaustive workload identity definition and monitoring.
  • Authenticating network connection at each end by identifying both the client and the server, systematically enforcing authentication policy, and excluding non-whitelisted connection.
  • Encrypting every network connection
  • Applying the least privilege principle
  • Continuously monitoring network connections, transactions, and communications

Zero Security Best Practices Zero Security Best Practices

In a containerized environment such as Kubernetes, this implies involving DevOps in baking in zero trust security policies from building to development and runtime. Structured in containers, nodes, pods and clusters, Kubernetes are designed to maximize microsegmentation security practices. Proper segmentation magnifies zero trust network security through several best practices Kubernetes includes an array of policies that need to be configured individually and consistently, as Kubernetes default values are generally designed to accommodate a wide range of possibilities. This is antithetical to a tight zero trust policy that aims to grant access on a “need to know” basis to provide maximum insulation between segments and thus minimize the potential damage in case of a breach. Then the successful hacker can only access a minuscule fraction of the infrastructure.

Zero Security at Runtime 🔗︎

Nevertheless, as Kubernetes architecture often relies on external sources, and as hackers continuously create new attack vectors, the network needs to be consistently monitored. Efficient monitoring implies a complete mapping of every single one of the organization’s assets, plug-ins, and inputs from external sources. It also implies a continuous assessment of their vulnerabilities.

Zero Security at Runtime Zero Security at Runtime

The illustration here shows a runtime scan report, displaying the assets scanned, their origin, and the detected vulnerabilities classified by color reflecting their degree of risk. Launched in March 2020, Kubei is the first open-source Kubernetes runtime scanner and is fully integrated into its Cloud Native Security solution. Learn more about it here.