[
https://issues.apache.org/jira/browse/FLINK-21659?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17325723#comment-17325723
]
Chesnay Schepler commented on FLINK-21659:
------------------------------------------
There is some regex parsing being applied to the result of a REST request, but
we don't check that the request was actually successful. We end up trying to
parse a checkpoint id from a stacktrace:
{code:java}
/home/vsts/work/1/s/flink-end-to-end-tests/test-scripts/common_ha.sh: line 151:
[: 58)\n\tat
org.apache.flink.runtime.rest.handler.job.AbstractAccessExecutionGraphHandler.handleRequest(AbstractAccessExecutionGraphHandler.java:
integer expression expected {code}
The reason this request fails is because, if a job is in the INITIALIZING
state, then the ArchivedExecutionGraph is created with a nulled
CheckpointStatsSnapshot, which is mistakenly interpreted by the handler as
checkpointing not being enabled.
This is not logged, because for this specific case the logging was disabled in
https://github.com/apache/flink/pull/14129.
> Running HA per-job cluster (rocks, incremental) end-to-end test fails
> ---------------------------------------------------------------------
>
> Key: FLINK-21659
> URL: https://issues.apache.org/jira/browse/FLINK-21659
> Project: Flink
> Issue Type: Bug
> Components: Runtime / Coordination
> Affects Versions: 1.13.0
> Reporter: Guowei Ma
> Assignee: Chesnay Schepler
> Priority: Critical
> Labels: test-stability
> Fix For: 1.13.0
>
> Attachments: screenshot-1.png
>
>
> https://dev.azure.com/apache-flink/apache-flink/_build/results?buildId=14232&view=logs&j=4dd4dbdd-1802-5eb7-a518-6acd9d24d0fc&t=8d6b4dd3-4ca1-5611-1743-57a7d76b395a
> “Completed checkpoint” is more than two times(42 times in the
> "flink-vsts-standalonejob-2-fv-az83-563.log") but the test still fail.
> !screenshot-1.png!
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
