Cilium: Up and Running

Chapter 16. Operations

Over the course of this book, you have explored how Cilium provides high-performance networking, security, and observability for Kubernetes. You have learned about its key components, how packets are processed through the datapath, and how features such as service load balancing, network policies, and encryption are implemented. Along the way, you have also seen multiple methods to install Cilium and many examples of commands to verify that those features are operating as expected.

In this final chapter, we shift our focus to day-to-day operations. Running Cilium in production involves maintaining stability throughout upgrades, monitoring performance and capacity, and resolving issues when they inevitably arise. This chapter brings together operational practices, tooling, and considerations that will enable you to run Cilium and Hubble reliably in production environments.

Installation and Lifecycle Management

In the previous chapters we have shown you multiple ways to install and configure Cilium. To achieve operational success, it’s important to understand that the simplest or quickest method is not always the best choice for production environments. If you don’t already have a process, or are looking to improve one, this section describes a few common patterns and offers recommendations based on our experience.

Tip

If you’re targeting a Red Hat OpenShift cluster, Isovalent Networking for Kubernetes, an enterprise distribution of Cilium, provides a certified ...