We're currently running flink on kubernetes as a job cluster using this helm template: https://github.com/docker-flink/examples/tree/master/helm/flink ( with some added configuration ).
If I want to shutdown the cluster, re-deploy a new image ( due to application code update ) and re-start, how would I go about restoring from a savepoint?
The jobmanager command is strictly set on the standalone-job.sh command and if I add a savepoint in the parameter in the deployments k8s resource, then if flink restarts ( due to some system error ), it will always restart from that savepoint, which is not what we want.
Is there a way to restore from the latest savepoint and if that savepoint doesn't exist, it will just start fresh with the kubernetes job cluster helm configuration?
