// the find
jaegertracing/jaeger-operator
Jaeger Operator for Kubernetes simplifies deploying and running Jaeger on Kubernetes.
A Kubernetes operator that manages Jaeger distributed tracing deployments via a CRD. You describe what you want in a Jaeger CR and the operator handles creating and reconciling all the underlying Kubernetes resources. Aimed at platform teams who need to run Jaeger at scale without hand-managing dozens of YAML files.
The operator handles the full Jaeger topology — agent daemonsets, collectors, ingesters, query pods, Kafka streaming pipelines — all wired together from a single CR. The sidecar injection via a webhook is genuinely useful: annotate a namespace and your apps get the agent injected without touching their deployments. Good test coverage across the pkg/ layer with real controller tests, not just happy-path unit tests. The generate subcommand that spits out plain manifests without running the operator is a practical escape hatch for GitOps workflows that can't tolerate CRDs.
The operator is effectively deprecated for Jaeger V2 — the project itself tells you to use the OpenTelemetry Operator instead, which means you're either stuck on Jaeger V1 or you're learning two operator stacks. The all-in-one deployment mode stores traces in memory, which is fine for dev but the README undersells how quickly you'll hit its limits in anything resembling production traffic. Elasticsearch integration leans heavily on the ECK operator being present, adding another dependency most teams don't already have. The compatibility matrix for Kubernetes versions is maintained but the Strimzi operator compatibility section is vague — 'might work on other untested versions' is not reassuring when Kafka is your production trace buffer.