Chesnay Schepler created FLINK-24887:
----------------------------------------
Summary: Retrying savepoints may cause early cluster shutdown
Key: FLINK-24887
URL: https://issues.apache.org/jira/browse/FLINK-24887
Project: Flink
Issue Type: Bug
Components: Runtime / REST
Affects Versions: 1.15.0
Reporter: Chesnay Schepler
Assignee: Chesnay Schepler
Fix For: 1.15.0
If an operation is retried we potentially access the result of a previous
attempt to see if it has already failed and eagerly fail the trigger request.
If that attempt is already complete then this may lead to an unexpected
shutdown of the cluster.
Beyond this issue, the eager checking of previous attempts makes error handling
more complicated, because you have to cover all cases for both the trigger and
status-retrieval operations.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
