[
https://issues.apache.org/jira/browse/HBASE-12891?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dave Latham updated HBASE-12891:
--------------------------------
Attachment: HBASE-12891-v3.patch
Thanks, Stephen, for jumping in, identifying the concurrency problem and the
new patch.
Churro is out right now and we've been up to some other things.
I reviewed the patch, and I'd like to propose an alternative that separates the
lazy load of regionsFromMeta into a synchronized method and leaves use of it
alone. I've attached that as HBASE-12891-v3.patch
> Parallel execution for Hbck checkRegionConsistency
> --------------------------------------------------
>
> Key: HBASE-12891
> URL: https://issues.apache.org/jira/browse/HBASE-12891
> Project: HBase
> Issue Type: Improvement
> Components: hbck
> Affects Versions: 2.0.0, 0.98.10, 1.1.0
> Reporter: churro morales
> Assignee: Stephen Yuan Jiang
> Labels: performance, scalability
> Fix For: 2.0.0, 1.1.0, 0.98.13
>
> Attachments: HBASE-12891-v1.patch, HBASE-12891-v3.patch,
> HBASE-12891.98.patch, HBASE-12891.patch, HBASE-12891.patch,
> HBASE-12891.v2-branch-1.patch, HBASE-12891.v2-master.patch,
> hbase-12891-addendum1.patch
>
>
> We have a lot of regions on our cluster ~500k and noticed that hbck took
> quite some time in checkAndFixConsistency(). [~davelatham] patched our
> cluster to do this check in parallel to speed things up. I'll attach the
> patch.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
