// the find
strimzi/strimzi-kafka-operator
Apache Kafka® running on Kubernetes
Strimzi is a Kubernetes operator for running Apache Kafka clusters, handling the full lifecycle: provisioning, configuration, TLS, upgrades, and scaling. It's a CNCF incubating project with broad OpenShift support and is the de facto standard for Kafka-on-Kubernetes. If you're running Kafka in k8s without Strimzi, you're probably reinventing it badly.
The CRD surface area is extensive in a good way — KafkaNodePool, KafkaMirrorMaker2, KafkaBridge, KafkaConnect are all first-class resources, not afterthoughts. KRaft mode (ZooKeeper-free) is now the default path, which removes a painful operational dependency. Container signing with cosign and SBOM publication from 0.38+ is production-grade supply chain hygiene that most operators skip entirely. The CI pipeline is unusually thorough: regression, migration, KRaft, feature-gate, and namespace RBAC scope pipelines all run separately, which means they actually catch regressions rather than masking them.
The Java codebase is enormous — the operator, API module, and system tests are all in one monorepo and the build is Maven-heavy, so local iteration is slow and painful if you want to contribute. The CRD API is expressive but verbose; defining a production-grade Kafka cluster with proper TLS and RBAC takes hundreds of lines of YAML, and there's no higher-level abstraction for common patterns. Upgrade paths between Strimzi versions are non-trivial and version-locked to specific Kafka versions, so falling behind means a multi-step migration rather than a simple chart bump. Documentation is thorough but split across the website, in-repo docs, and GitHub Discussions in ways that make it easy to end up reading outdated instructions.