Details
-
Bug
-
Status: Open
-
P2
-
Resolution: Unresolved
-
None
-
None
Description
Didn't have time to look very closely into the root cause, but in taking a look at flaky cross-language tests I noticed a pattern of different suites on different runners flaking at the same time. The specific ones that I've noticed so far are:
Samza: https://ci-beam.apache.org/job/beam_PostCommit_XVR_Samza/
Spark: https://ci-beam.apache.org/job/beam_PostCommit_XVR_Spark/
Dataflow: https://ci-beam.apache.org/job/beam_PostCommit_XVR_PythonUsingJava_Dataflow/
Example flake (Mar 29, 12 PM):
https://ci-beam.apache.org/job/beam_PostCommit_XVR_Samza/993/
https://ci-beam.apache.org/job/beam_PostCommit_XVR_Spark/3530/
https://ci-beam.apache.org/job/beam_PostCommit_XVR_PythonUsingJava_Dataflow/242/
Example flake 2 (Mar 30, 6 PM):
https://ci-beam.apache.org/job/beam_PostCommit_XVR_Samza/998/
https://ci-beam.apache.org/job/beam_PostCommit_XVR_Spark/3535/
https://ci-beam.apache.org/job/beam_PostCommit_XVR_PythonUsingJava_Dataflow/247/