You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@flink.apache.org by "Dale Lane (Jira)" <ji...@apache.org> on 2026/09/30 22:55:00 UTC

[jira] [Created] (FLINK-40864) Add the FlinkMiniCluster Flink service and reconciler

Dale Lane created FLINK-40864:
---------------------------------

             Summary: Add the FlinkMiniCluster Flink service and reconciler
                 Key: FLINK-40864
                 URL: https://issues.apache.org/jira/browse/FLINK-40864
             Project: Flink
          Issue Type: Sub-task
          Components: Kubernetes Operator
            Reporter: Dale Lane


Adds the components that deploy and manage the lifecycle of a {{{}FlinkMiniCluster{}}}.

{{MiniClusterFlinkService}} will extend {{{}AbstractFlinkService{}}}, supplying the deployment-specific behaviour (creating and deleting the Deployment, Service and ConfigMap; locating the MiniCluster pod) while inheriting REST-based job management, since a MiniCluster exposes the same REST endpoint as a distributed JobManager. Session-cluster deployment and scaling are not applicable and will be rejected.

{{MiniClusterReconciler}} will extend {{{}AbstractJobReconciler{}}}, so upgrade modes, suspend and redeploy semantics match {{{}FlinkDeployment{}}}. The job should be submitted under a fixed JobID that is assigned once and reused across deployments, so that checkpoint lineage survives pod restarts and the launcher can rejoin a job the dispatcher has already recovered from HA metadata.

Nothing calls this yet, the controller wiring will come in a later sub-task



--
This message was sent by Atlassian Jira
(v8.20.10#820010)