[
https://issues.apache.org/jira/browse/FLINK-23283?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17376322#comment-17376322
]
JING ZHANG edited comment on FLINK-23283 at 7/7/21, 6:21 AM:
-------------------------------------------------------------
[~xintongsong] The unstable case is caused by whether the last two records
dropped as late data is nonderterministic. It depends on the watermark on
current task which is relates to parallelism and what data is processed on each
concurrent task.
And
`GroupWindowITCase#testWindowAggregateOnUpsertSourcePushdownWatermark` has
similar problems, I would set parallelism to 1 for those two cases to avoid
unstable failure.
was (Author: qingru zhang):
[~xintongsong] The unstable case caused by whether the last two records dropped
as late data depends on the watermark on current task which is
nonderterministic which depends on parallelism and what data is processed on
each concurrent task.
I would set parallelism to 1 for those two cases to avoid unstable failure.
> GroupWindowITCase.testWindowAggregateOnUpsertSource fails on azure
> ------------------------------------------------------------------
>
> Key: FLINK-23283
> URL: https://issues.apache.org/jira/browse/FLINK-23283
> Project: Flink
> Issue Type: Bug
> Components: Table SQL / Planner
> Affects Versions: 1.14.0
> Reporter: Roman Khachatryan
> Assignee: JING ZHANG
> Priority: Blocker
> Fix For: 1.14.0
>
>
> https://dev.azure.com/apache-flink/apache-flink/_build/results?buildId=19978&view=logs&j=e25d5e7e-2a9c-5589-4940-0b638d75a414&t=a6e0f756-5bb9-5ea8-a468-5f60db442a29&l=7086
> {code}
> 2021-07-06T10:13:10.6091443Z Jul 06 10:13:10 [ERROR] Tests run: 48, Failures:
> 1, Errors: 0, Skipped: 0, Time elapsed: 14.945 s <<< FAILURE! - in
> org.apache.flink.table.planner.runtime.stream.sql.GroupWindowITCase
> 2021-07-06T10:13:10.6093667Z Jul 06 10:13:10 [ERROR]
> testWindowAggregateOnUpsertSource[StateBackend=HEAP, UseTimestampLtz =
> false](org.apache.flin*k.table.planner.runtime.stream.sql.GroupWindowITCase)
> Time elapsed: 0.329 s <<< FAILURE!
> 2021-07-06T10:13:10.6096965Z Jul 06 10:13:10 java.lang.AssertionError:
> expected:<List(Euro,1,118,1970-01-01T00:00:15,1970-01-01T00:00:20, US Dolla
> r,1,102,1970-01-01T00:00,1970-01-01T00:00:05,
> Yen,1,1,1970-01-01T00:00,1970-01-01T00:00:05)> but
> was:<List(Euro,1,118,1970-01-01T00:00:15,1970-01- 01T00:00:20, US
> Dollar,1,104,1970-01-01T00:00,1970-01-01T00:00:05,
> Yen,1,1,1970-01-01T00:00,1970-01-01T00:00:05)>
> 2021-07-06T10:13:10.6098985Z Jul 06 10:13:10 at
> org.junit.Assert.fail(Assert.java:89)
> 2021-07-06T10:13:10.6099695Z Jul 06 10:13:10 at
> org.junit.Assert.failNotEquals(Assert.java:835)
> 2021-07-06T10:13:10.6100489Z Jul 06 10:13:10 at
> org.junit.Assert.assertEquals(Assert.java:120)
> 2021-07-06T10:13:10.6101292Z Jul 06 10:13:10 at
> org.junit.Assert.assertEquals(Assert.java:146)
> 2021-07-06T10:13:10.6102395Z Jul 06 10:13:10 at
> org.apache.flink.table.planner.runtime.stream.sql.GroupWindowITCase.testWindowAggregateOnUpsert*Source(GroupWindowITCase.scala:421)
> {code}
> Fails locally 3 times out of 1000
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
