[
https://issues.apache.org/jira/browse/HDFS-15674?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17233582#comment-17233582
]
Masatake Iwasaki commented on HDFS-15674:
-----------------------------------------
The test intermittently fails if count of reported block in IBRs does not
reached to expected value(4000).
{noformat}
[ERROR] Failures:
[ERROR] TestBPOfferService.testMissBlocksWhenReregister:341 Timed out wait
for IBR counts FBRCount = 2441, IBRCount = 3999; expected = 4000. Exception:
Timed out waiting for condition. Thread diagnostics:
Timestamp: 2020-11-16 12:45:57,425
{noformat}
This waiting condition does not make sense because pending IBRs are possible to
be deleted by {{clearIBRs()}} in {{BPServiceActor#reRegister}}. It is ok
because succeeding full block report covers them based on the discussion of
HDFS-15113.
The tests should check that all blocks are reported rather than counting number
of reported blocks in FBR/IBR independently.
I submitted [PR #2467|https://github.com/apache/hadoop/pull/2467] for this.
> TestBPOfferService#testMissBlocksWhenReregister fails on trunk
> --------------------------------------------------------------
>
> Key: HDFS-15674
> URL: https://issues.apache.org/jira/browse/HDFS-15674
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Ahmed Hussein
> Assignee: Masatake Iwasaki
> Priority: Major
> Labels: pull-request-available
> Time Spent: 10m
> Remaining Estimate: 0h
>
> qbt report (Nov 8, 2020, 11:28 AM) shows failures timing out in
> testMissBlocksWhenReregister
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
