[
https://issues.apache.org/jira/browse/HBASE-18016?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16002005#comment-16002005
]
stack commented on HBASE-18016:
-------------------------------
Agree that if you look at TruncateTableProcedure, it doesn't seem to make sense
adding an abort but fact is that Procedures will have corner cases.
In this case, what we saw was an infinite loop of the following:
{code}
2017-04-18 21:06:05,392 WARN
org.apache.hadoop.hbase.master.procedure.TruncateTableProcedure: Retriable
error trying to truncate table=XXXXstate=TRUNCATE_TABLE_CREATE_FS_LAYOUT
java.io.IOException: java.util.concurrent.ExecutionException:
java.io.IOException: The specified region already exists on disk:
hdfs://nameservice1/hbase/XX/.tmp/data/default/YYYYYY/b76c62fb69c0aa8e667adfd604a22c69
....
2017-04-18 21:06:05,176 WARN
org.apache.hadoop.hbase.regionserver.HRegionFileSystem: Trying to create a
region that already exists on disk:
hdfs://nameservice1/hbase/xx/.tmp/data/default/YYYYYYY/b76c62fb69c0aa8e667adfd604a22c69
{code}
Truncate had borked itself. The above condition is probably fixable... i.e.
Truncate should recognize this state where it has mangled events and deal
appropriately ... but ti will take a while till we have ironed out all corner
cases.
Meantime, operators should be able to kill/abort Procedures gone wild whatever
the reason.
> Implement abort for TruncateTableProcedure
> ------------------------------------------
>
> Key: HBASE-18016
> URL: https://issues.apache.org/jira/browse/HBASE-18016
> Project: HBase
> Issue Type: Sub-task
> Components: proc-v2
> Reporter: Umesh Agashe
> Assignee: Umesh Agashe
> Fix For: 2.0.0
>
>
> TruncateTableProcedure can not be aborted as abort is not implemented.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
