Hi Yasukazu, spark-Kubernetes-operator does support exposing metrics to Prometheus, please refer docs for available metrics and forwarding behavior at https://github.com/apache/spark-kubernetes-operator/blob/main/docs/configuration.md#forward-metrics-to-prometheus
Regards, Zhou > On Jun 29, 2025, at 09:11, Yasukazu Nagatomi <yassan0...@gmail.com> wrote: > > > Hi everyone, > > First off, great work on the spark-kubernetes-operator release cadence—v0.3.0 > landed last month, and I see SPARK-52378 has closed the v0.4.0 release JIRA > issue. It’s also exciting to see the CRD upgrade to v1 on the roadmap for > v0.5.0. > > I have a question: I couldn’t find any built-in support for monitoring the > operator itself. Are there plans in the v0.5.0 roadmap to expose > Prometheus-style metrics (similar to MonitoringSpec/PrometheusSpec in > kubeflow/spark-operator)? Even if the API differs, having a metrics endpoint > to observe the operator’s health and activity would be really valuable. > > Apologies if this is already being considered. > > > Relevant links: > - PR: https://github.com/apache/spark-kubernetes-operator/pull/261 > - JIRA SPARK-52378 (Release Spark Kubernetes Operator 0.4.0): > https://issues.apache.org/jira/browse/SPARK-52378 > - JIRA SPARK-52604 (Release Spark Kubernetes Operator 0.5.0 (v1)): > https://issues.apache.org/jira/browse/SPARK-52604 > > Thanks for all your hard work! > > Best regards, > Yasukazu
