miaht94 opened a new pull request, #7710: URL: https://github.com/apache/kyuubi/pull/7710
### Why are the changes needed? When Kyuubi Server runs inside Kubernetes without `kyuubi.kubernetes.client.initialize.list`, no Kubernetes client or informer is initialized at startup. As a result, terminated Spark driver pods can remain after their retention period even though cleanup is enabled. Initialize the configured namespace automatically from the in-cluster environment while preserving the explicit initialization list as the higher-priority configuration. Outside Kubernetes, the existing empty-list behavior is unchanged. The stale Guava-expiration behavior is already handled on `master` by #7220, so this patch does not add the Caffeine dependency or unrelated Helm chart changes. Closes #7293. ### How was this patch tested? - `dev/gen/gen_all_config_docs.sh` - `dev/reformat` - `build/mvn test -pl kyuubi-server -am -Pspark-provided,hive-provided,flink-provided -Dtest=none -DwildcardSuites=org.apache.kyuubi.engine.KubernetesApplicationOperationSuite` The focused suite covers no automatic initialization outside Kubernetes, automatic in-cluster initialization for the configured namespace, and explicit-list precedence. ### Was this patch assisted by generative AI tooling? Assisted-by: OpenAI Codex -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
