Cynet: Optimizing Kubernetes Cost to Strengthen Competitive Advantage

Cynet is a pioneer and leader in advanced threat detection and response on a mission to make cybersecurity easy. By combining high-fidelity detection, decoy interactions, network analytics, and expert analyst assistance, Cynet provides accurate findings with associated risks without a lot of complexity and noise, so security teams can prioritize and respond to what matters.

Cynet believes that cybersecurity doesn't need to be complicated, and their focus on simplifying the complexity is a value they bring into their business operations.  In fact, Leonardo Da Vinci’s saying, “Simplicity is the ultimate sophistication,” is a part of the company’s DNA. This factored into their decisions to select Kubernetes as their container orchestration platform. 
‍
Kubernetes is a critical technology that helps Cynet simplify their operations.  It standardizes the way they deploy, allocate compute resources, and monitor the health and performance of their system.  It also provided a simplified approach to scaling their applications up and down to support the demands of their customer base. 

Today, Elad Gotfrid, Director of Operations at Cynet, and his team of DevOps Engineers and SREs are managing a large, multi-cluster, multi-region AWS EKS environment, that supports all levels of the application delivery process, from dev to staging to production.

Limited visibility into pod costs hampers Kubernetes cost optimization

With the increasingly competitive cybersecurity landscape, Gotfrid and his team were under pressure to reduce cloud costs and optimize compute resources. By reducing costs, his team can help improve business outcomes and give Cynet an additional competitive advantage.   

“The cost of our cloud infrastructure, especially with Kubernetes, makes a direct impact on the cost of our endpoints that are deployed in our customer’s environments,” Gotfrid explained. “By keeping our cloud costs optimized, we are enabling our business to provide a powerful solution to our customers, at a low cost.”

His team was successful in optimizing the cost of the majority of their cloud services, but when it came to Kubernetes, they simply did not have the cost visibility to efficiently optimize their environment. 

“One big gap we had was we were unable to see the cost of each pod and service,” Gotfrid said “We needed a better visibility into our pods and what was happening underneath to properly optimize the costs.”

This made cost optimization efforts a very manual process and something they struggled to find the time to accomplish. Additionally, the lack of visibility could present potential risks when trying to reduce costs. If optimization efforts are too aggressive they can negatively impact system reliability and stability.  

“Finding the right balance between stability and cost-effectiveness is the most challenging part.  You need clear cost visibility, the ability to prioritize the needs of critical services, and to understand the impact and potential risks of any changes you are making to the environment,” Gotfrid explained. “PerfectScale made finding this balance extremely simple.”  

Centralized Solution to Identify Resiliency Issues

Since implementing PerfectScale, the path to ensuring their Kubernetes environment is continually stable and cost-effective has been quick, efficient, and convenient. 

“The solutions provided us with a quick time to value. Implementation took only a few minutes, and we instantly started getting the level of Kubernetes cost visibility we were looking for,” Gotfrid said. “The solution made it easy to identify wasteful workloads and provided us recommendations on what changes to make to cut costs while maintaining resilience.” 

Additionally, PerfectScale provided Gotfrid’s team with a centralized solution to identify resiliency issues.  Previously, the team was using multiple Prometheus instances to monitor and manage the stability of their clusters. This made it complex and time-consuming to investigate and resolve issues like OOM or CPU Throttling.

“PerfectScale provided us with a single pane of glass to identify stability issues across all of our clusters. It has actually found several issues we didn’t know were happening,” Gotfrid explained. “Now, when an issue is detected, we get alerted instantly with the recommended changes we need to resolve the issues quickly.”

Balancing Optimization with Stability and Driving Business Results 

The DevOps team at Cynet was able to leverage PerfectScale to quickly reduce unnecessary cloud waste without impacting the stability of their environment. The goal is to continue to use the tool to ensure the perfect balance of cost and performance in the future.  

“PerfectScale has allowed us to fulfill our commitment to the business, and run our Kubernetes environment as lean as possible,” said Gotfrid. “As our development team continues to develop innovations that are disrupting the cybersecurity market, PerfectScale will help us ensure our solution is stable, resilient, and competitively priced for our customers.”

‍