Gyula Fora created FLINK-30529:
----------------------------------
Summary: Use simple cancel for savepoint upgrades on failing jobs
Key: FLINK-30529
URL: https://issues.apache.org/jira/browse/FLINK-30529
Project: Flink
Issue Type: Improvement
Components: Kubernetes Operator
Reporter: Gyula Fora
In some cases where savepoint upgrades are required we should consider simply
cancelling the job and observing the last checkpoint/savepoint available.
We should think through what are the caveats of this approach but this could
help upgrade some clusters that would be stuck now.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
