beam The PostCommit Java Dataflow V2 job is flaky

The PostCommit Java Dataflow V2 is failing over 50% of the time Please visit https://github.com/apache/beam/actions/workflows/beam_PostCommit_Java_DataflowV2.yml?query=is%3Afailure+branch%3Amaster to see the logs.

Mar 24 '24 09:03 github-actions[bot]

.take-issue

Mar 25 '24 12:03 tilgalas

Reopening since the workflow is still flaky

May 04 '24 12:05 github-actions[bot]

Reopening since the workflow is still flaky

Jan 23 '25 09:01 github-actions[bot]

This time is due to testReadFailsBadSession (org.apache.beam.sdk.io.gcp.spanner.SpannerReadIT) failed, since Jan 18, 2025

however in between https://github.com/apache/beam/actions/runs/12850039366 and https://github.com/apache/beam/actions/runs/12846858051 there is no change there is no code change

could this be caused by Spanner service change?

Jan 23 '25 16:01 Abacn

Job has been stable for 3 days. Closing as resolved.

Feb 07 '25 06:02 Amar3tto

Reopening since the workflow is still flaky

Mar 04 '25 03:03 github-actions[bot]

StorageApiSinkSchemaUpdateIT has been flaky. These are tricky tests in terms of timing that has schema setup asynchronuously. They'd better be exercised on direct runner where timing varies less. My suggestion here is to disable or just keep one test in StorageApiSinkSchemaUpdateIT on Dataflow runner, to reduce the flakiness

Mar 05 '25 15:03 Abacn