In the realm of Kubernetes, initiating an application is relatively straightforward. However, the transition to a production-ready state presents a distinct set of challenges. How can we ensure the system's performance under actual operational conditions and promptly detect any anomalies?

This presentation aims to shed light on these concerns. We will guide you through our approach to monitoring our Kubernetes cluster fleet and the applications deployed within, with a closer look on Apache Kafka managed via Strimzi as our SUT in our test environment. The Kubernetes ecosystem offers a lot of powerful tools for this purpose, including Thanos, Prometheus, Grafana, Loki, Vector, and many others. Leveraging these tools in conjunction with ArgoCD, we will demonstrate effective ways for monitoring your clusters and applications, establishing and receiving timely alerts, and maintaining access to vital metrics and logs over extended periods.