Efrat Levitan created FLINK-40570:
-------------------------------------
Summary: Speed up CI by cancelling jobs upon test completion
Key: FLINK-40570
URL: https://issues.apache.org/jira/browse/FLINK-40570
Project: Flink
Issue Type: Sub-task
Components: Connectors / Kafka
Reporter: Efrat Levitan
While analyzing CI
failures[[1]|https://github.com/apache/flink-connector-kafka/actions/runs/33749050730/job/100628141595?pr=285]
[[2]|https://github.com/apache/flink-connector-kafka/actions/runs/33780517843/job/101106906200?pr=295]
I noticed that {{{}testTopicIntegritySuccess{}}}, though successfully passing,
yields >1000 lines of {{TimeoutException: Kafka server timed out}} stack traces
due to closing the kafka server before the minicluster teardown, and also holds
back minicluster slots from being rescheduled for the next test, slowing down
CI.
*Suggested fix:*
Cancel the jobs when the test completes to improve suite speed and reduce
misleading logs
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
