[
https://issues.apache.org/jira/browse/HADOOP-2029?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HADOOP-2029:
--------------------------
Attachment: logrolling-v3.patch
Add logging. New supposition is that an updating client thread got a region
before it was moved to retiring queue inside of HRegionServer#batchUpdate.
Thread then moves into HRegion#startUpdate and blocks on a read lock while
concurrently region is being shut down. Moved check on whether region is
closed inside the read lock rather than check before.
> [hbase] TestLogRolling fails too often in patch and nightlies
> -------------------------------------------------------------
>
> Key: HADOOP-2029
> URL: https://issues.apache.org/jira/browse/HADOOP-2029
> Project: Hadoop
> Issue Type: Bug
> Components: contrib/hbase
> Reporter: stack
> Assignee: stack
> Priority: Minor
> Fix For: 0.15.0
>
> Attachments: hlogrolling.patch, logrolling-v2.patch,
> logrolling-v3.patch
>
>
> Looks like we fixed TestRegionServerAbort yesterday. Now TestLogRolling has
> been acting up of late.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
