Software at Scale 55 - Troubleshooting and Operating K8s with Ben Ofiri
Everything you wanted to know about running Kubernetes, and more.
Ben Ofiri is the CEO and Co-Founder of Komodor, a Kubernetes troubleshooting platform.
We had an episode with the other founder of Komodor, Itiel, in 2021, and I thought it would be fun to revisit the topic.
Highlights (ChatGPT Generated)
[0:00] Introduction to the Software At Scale podcast and the guest speaker, Ben Ofiri, CEO and co-founder of Komodor.
- Discussion of why Ben decided to work on a Kubernetes platform and the potential impact of Kubernetes becoming the standard for managing microservices.
- Reasons why companies are interested in adopting Kubernetes, including the ability to scale quickly and cost-effectively, and the enterprise-ready features it offers.
- The different ways companies migrate to Kubernetes, either starting from a small team and gradually increasing usage, or a strategic decision from the top down.
- The flexibility of Kubernetes is its strength, but it also comes with complexity that can lead to increased time spent on alerts and managing incidents.
- The learning curve for developers to be able to efficiently troubleshoot and operate Kubernetes can be steep and is a concern for many organizations.
[8:17] Tools for Managing Kubernetes.
- The challenges that arise when trying to operate and manage Kubernetes.
- DevOps and SRE teams become the bottleneck due to their expertise in managing Kubernetes, leading to frustration for other teams.
- A report by the cloud native observability organization found that one out of five developers felt frustrated enough to want to quit their job due to friction between different teams.
- Ben's idea for Komodor was to take the knowledge and expertise of the DevOps and SRE teams and democratize it to the entire organization.
- The platform simplifies the operation, management, and troubleshooting aspects of Kubernetes for every engineer in the company, from junior developers to the head of engineering.
- One of the most frustrating issues for customers is identifying which teams should care about which issues in Kubernetes, which Komodor helps solve with automated checks and reports that indicate whether the problem is an infrastructure or application issue, among other things.
- Komodor provides suggestions for actions to take but leaves the decision-making and responsibility for taking the action to the users.
- The platform allows users to track how many times they take an action and how useful it is, allowing for optimization over time.
[8:17] Tools for Managing Kubernetes.
[12:03] The Challenge of Balancing Standardization and Flexibility.
- Kubernetes provides a lot of flexibility, but this can lead to fragmented infrastructure and inconsistent usage patterns.
- Komodor aims to strike a balance between standardization and flexibility, allowing for best practices and guidelines to be established while still allowing for customization and unique needs.
[16:14] Using Data to Improve Kubernetes Management.
- The platform tracks user actions and the effectiveness of those actions to make suggestions and fine-tune recommendations over time.
- The goal is to build a machine that knows what actions to take for almost all scenarios in Kubernetes, providing maximum benefit to customers.
[20:40] Why Kubernetes Doesn't Include All Management Functionality.
- Kubernetes is an open-source project with many different directions it can go in terms of adding functionality.
- Reliability, observability, and operational functionality are typically provided by vendors or cloud providers and not organically from the Kubernetes community.
- Different players in the ecosystem contribute different pieces to create a comprehensive experience for the end user.
[25:05] Keeping Up with Kubernetes Development and Adoption.
- How Komodor keeps up with Kubernetes development and adoption.
- The team is data-driven and closely tracks user feedback and needs, as well as new developments and changes in the ecosystem.
- The use and adoption of custom resources is a constantly evolving and rapidly changing area, requiring quick research and translation into product specs.
- The company hires deeply technical people, including those with backgrounds in DevOps and SRE, to ensure a deep understanding of the complex problem they are trying to solve.
[32:12] The Effects of the Economy on Komodor.
- The effects of the economy pivot on Komodor.
- Companiesmust be more cost-efficient, leading to increased interest in Kubernetes and tools like Komodor.
- The pandemic has also highlighted the need for remote work and cloud-based infrastructure, further fueling demand.
- Komodor has seen growth as a result of these factors and believes it is well-positioned for continued success.
[36:17] The Future of Kubernetes and Komodor.
- Kubernetes will continue to evolve and be adopted more widely by organizations of all sizes and industries.
- The team is excited about the potential of rule engines and other tools to improve management and automation within Kubernetes.