[
https://issues.apache.org/jira/browse/HBASE-3130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13105808#comment-13105808
]
Chris Trezzo commented on HBASE-3130:
-------------------------------------
@J-D
Thanks for the response! Attaching a new patch with your comments in mind.
Here are the changes:
1. Fixed the logging in ReplicationPeer so that it uses the commons logging
LogFactory.
2. Fixed the line with 80+ characters.
3. Moved the connection management logic from ReplicationSource down to
ReplicationZookeeper.getSlavesAddresses.
4. I kept the ReplicationPeer.reloadZkWatcher method for a couple reasons. (1)
In order to move all the ZookeeperWatcher logic out of ReplicationPeer, we
would need to have a setZkw method. This potentially allows for null
ZookeeperWatchers within a ReplicationPeer (the reloadZkWatcher method seems
like a slightly cleaner choice). (2) It is basically doing the exact same thing
that the ReplicationPeer constructor is already doing.
Let me know what you think.
Thanks,
Chris
> [replication] ReplicationSource can't recover from session expired on remote
> clusters
> -------------------------------------------------------------------------------------
>
> Key: HBASE-3130
> URL: https://issues.apache.org/jira/browse/HBASE-3130
> Project: HBase
> Issue Type: Bug
> Components: replication
> Affects Versions: 0.92.0
> Reporter: Jean-Daniel Cryans
> Assignee: Chris Trezzo
> Fix For: 0.92.0
>
> Attachments: 3130.txt
>
>
> Currently ReplicationSource cannot recover when its zookeeper connection to
> its remote cluster expires. HLogs are still being tracked, but a cluster
> restart is required to continue replication (or a rolling restart).
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
