[
https://issues.apache.org/jira/browse/HBASE-6617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14709686#comment-14709686
]
Sean Busbey commented on HBASE-6617:
------------------------------------
{quote}
Sorry that somehow I neglected this comment when adding the comment to attach
the patch. The problem here is that in case there're multiple wal groups, log
roll of one group might clean up the not-replicated-yet log of the other group,
especially considering the log set is sorted by the log name.
Take BoundedRegionGroupingProvider for example, assuming using the default
group number 2, then we will have two wal groups, say "<providerId>-0" and
"<providerId>-1", if the log in "<providerId>-1" group got more writes and
rolled first, then all wals in "<providerId>-0" will got removed unexpectedly
{quote}
If the replication code is relying on sort order of WAL file names to represent
order of wal files, that's a problem because there is no guaranteed order even
within a single provider. We [make use of System.currentTimeMillis() in the
name|https://github.com/apache/hbase/blob/master/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/wal/FSHLog.java#L587],
which can go all over the place.
> ReplicationSourceManager should be able to track multiple WAL paths
> -------------------------------------------------------------------
>
> Key: HBASE-6617
> URL: https://issues.apache.org/jira/browse/HBASE-6617
> Project: HBase
> Issue Type: Improvement
> Components: Replication
> Reporter: Ted Yu
> Assignee: Sean Busbey
> Fix For: 2.0.0, 1.3.0
>
> Attachments: HBASE-6617.patch
>
>
> Currently ReplicationSourceManager uses logRolled() to receive notification
> about new HLog and remembers it in latestPath.
> When region server has multiple WAL support, we need to keep track of
> multiple Path's in ReplicationSourceManager
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
