// the find
opensearch-project/opensearch-k8s-operator
OpenSearch Kubernetes Operator
A Kubernetes operator from the OpenSearch project itself that manages the full lifecycle of OpenSearch clusters and Dashboards via a CRD (OpenSearchCluster). It's for teams running OpenSearch on k8s who want the operator pattern instead of hand-rolled StatefulSets and scripts for upgrades, scaling, and TLS.
It handles the operational stuff that's actually hard to get right by hand: quorum-safe rolling restarts during version upgrades, online volume expansion, and TLS certificate hot-reloading without pod restarts. Per-node-pool configuration (separate manager/data/ingest/coordinating pools with independent scaling) maps well to how OpenSearch clusters are actually run in production. It ships as a Helm chart on Artifact Hub plus an OLM bundle, and CI covers build, lint, and functional tests on every push.
The operator is mid-migration from the opensearch.opster.io API group to opensearch.org, so both CRD groups exist in the codebase right now (two full copies of every CRD type under api/v1 and api/opensearch.org/v1) — expect CRD churn and some doc/config drift until that settles. The upgrade-test workflow is sitting in a workflows-disabled folder, which is a bit ironic for a project whose whole pitch includes 'rolling version upgrades.' The version compatibility table only covers a handful of explicitly-tested OpenSearch releases, so anything outside that range is 'proceed with caution.' 342 forks against 584 stars is a high ratio for this size of project, suggesting a fair number of users are running patched forks rather than upstreaming fixes.