[
https://issues.apache.org/jira/browse/HBASE-3044?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jean-Daniel Cryans updated HBASE-3044:
--------------------------------------
Attachment: HBASE-3044.patch
Fixes the issue and adds a safeguard later on in the code.
> [replication] ReplicationSource won't cleanup logs if there's nothing to
> replicate
> ----------------------------------------------------------------------------------
>
> Key: HBASE-3044
> URL: https://issues.apache.org/jira/browse/HBASE-3044
> Project: HBase
> Issue Type: Bug
> Affects Versions: 0.89.20100924
> Reporter: Jean-Daniel Cryans
> Assignee: Jean-Daniel Cryans
> Fix For: 0.90.0
>
> Attachments: HBASE-3044.patch
>
>
> ReplicationSource only calls
> ReplicationSourceManager.logPositionAndCleanOldLogs if it replicated
> something, meaning that a region server that doesn't host a region with
> replicable edits will read through the logs but never report the position in
> zookeeper and won't clean the processed hlogs. This happened on one cluster
> here and shutting down a region server took a few minutes just to delete all
> the znodes.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
