[
https://issues.apache.org/jira/browse/HBASE-20087?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16379528#comment-16379528
]
Andrew Purtell commented on HBASE-20087:
----------------------------------------
bq. I'm trying to think about what are the recent cases in which I've seen
FAILED_OPEN's happening and if automatically trying to assign them would have
caused irreparable consequences (e.g. riding over some data loss). Nothing is
coming to mind
We have kicked FAILED_OPEN assignments a few times, and I've done it more
frequently when testing rsgroups, so much so it became part of the backport. I
think it is safe. If the condition isn't resolved the retries will just
continue to fail.
bq. would be nice to be a little more word-y here, to be more clear that the
region was FAILED_OPEN and we're trying to help.
I changed the log line to
{code}
LOG.info("Retrying failed assignment for " + s.toDescriptiveString());
{code}
The string built by RegionState#toDescriptiveString prints the current state,
which will be FAILED_OPEN, so that should be clear.
> Periodically attempt redeploy of regions in FAILED_OPEN state
> -------------------------------------------------------------
>
> Key: HBASE-20087
> URL: https://issues.apache.org/jira/browse/HBASE-20087
> Project: HBase
> Issue Type: Improvement
> Components: master, Region Assignment
> Reporter: Andrew Purtell
> Assignee: Andrew Purtell
> Priority: Major
> Fix For: 2.0.0, 1.5.0
>
> Attachments:
> 0001-W-4723090-Port-the-RIT-FAILED_OPEN-state-hack-from-R.patch,
> HBASE-20087-branch-1.patch, HBASE-20087-branch-1.patch
>
>
> Because RSGroups can cause permanent RIT with regions in FAILED_OPEN state,
> we added logic to the master portion of the RSGroups extention to enumerate
> RITs and retry assignment of regions in FAILED_OPEN state.
> However, this strategy can be applied generally to reduce need of operator
> involvement in cluster operations. Now an operator has to manually resolve
> FAILED_OPEN assignments but there is little risk in automatically retrying
> them after a while. If the reason the assignment failed has not cleared, the
> assignment will just fail again. Should the reason the assignment failed be
> resolved, then operators don't have to do more in order for the cluster to
> fully heal.
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
