[
https://issues.apache.org/jira/browse/HBASE-21530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16706441#comment-16706441
]
Andrew Purtell commented on HBASE-21530:
----------------------------------------
That makes sense but imagine this situation we recently experienced in
production: An external misbehaving process makes a large number of procedure
calls (to create and delete namespaces in this case). The work is not actually
useful and is seriously impacting operations. The queue of procedures remaining
to be executed is hundreds if not thousands of requests deep. They’re in the
proc WAL so boincing the master isn’t going to clear the queue. What do you
suggest we do here without a shell command or admin API for cancelling
requests? (We can delete the proc WAL manually, which is what was done, but
only out of desperation.)
We shouldn’t have any API where once a request is made there is no way to shed
the work. Mistakes will happen. We need to be able to clean up from the
mistakes. Suffering a major production incident with nothing to be done for
days is not acceptable.
> Abort_Procedure should be able to take a list of proc IDs
> ---------------------------------------------------------
>
> Key: HBASE-21530
> URL: https://issues.apache.org/jira/browse/HBASE-21530
> Project: HBase
> Issue Type: Improvement
> Reporter: Geoffrey Jacoby
> Priority: Minor
>
> As a convenience, it would be helpful if the HBase shell's abort_procedure
> call had the option of taking in multiple procedure ids at the same time,
> rather than relying on operators to use a loop in an external script.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
