[
https://issues.apache.org/jira/browse/HBASE-21745?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16854501#comment-16854501
]
Wellington Chevreuil commented on HBASE-21745:
----------------------------------------------
{quote}Rebuild meta from region metadata in the filesystem, aka offline meta
rebuild.
{quote}
We had seen a few cases requiring this among our customer base, lately. Some
trends observed:
- Happens after upgrading from 1.x to 2.1.0;
- Cluster apparently wasn't on a healthy state prior to upgrading (some
regions were already on an inconsistent state, or wrong permissions defined on
hbase table folders);
- OfflineMetaRepair/hbck1 forcedly used;
Obviously, it appears this situation mostly happens due poor admin actions, but
it may still be valid provide a convenient fix option for this (it has also
been mentioned by some folks on a recent discussion mail thread about hbck2
directions).
To aid our support team in such cases, I had recently written a [simpler
client|https://github.com/wchevreuil/support-tools/blob/master/src/main/java/com/cloudera/support/hbase/RegionMetaBuilder.java]
than OfflineMetaRepair, for re-inserting regions with metadata in hdfs back
into meta. If folks think it's relevant, I can convert it into another hbck2
command.
> Make HBCK2 be able to fix issues other than region assignment
> -------------------------------------------------------------
>
> Key: HBASE-21745
> URL: https://issues.apache.org/jira/browse/HBASE-21745
> Project: HBase
> Issue Type: Umbrella
> Components: hbase-operator-tools, hbck2
> Reporter: Duo Zhang
> Assignee: stack
> Priority: Critical
>
> This is what [~apurtell] posted on mailing-list, HBCK2 should support
> {quote}
> - Rebuild meta from region metadata in the filesystem, aka offline meta
> rebuild.
> - Fix assignment errors (undeployed regions, double assignments (yes,
> should not be possible), etc)
> - Fix region holes, overlaps, and other errors in the region chain
> - Fix failed split and merge transactions that have failed to roll back
> due to some bug (related to previous)
> - Enumerate store files to determine file level corruption and sideline
> corrupt files
> - Fix hfile link problems (dangling / broken)
> {quote}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
