You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hbase.apache.org by "Lars Hofhansl (JIRA)" <ji...@apache.org> on 2015/02/09 06:41:34 UTC
[jira] [Resolved] (HBASE-1406) Refactor HRS
[ https://issues.apache.org/jira/browse/HBASE-1406?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Lars Hofhansl resolved HBASE-1406.
----------------------------------
Resolution: Won't Fix
Doesn't look like this is going to happen.
> Refactor HRS
> ------------
>
> Key: HBASE-1406
> URL: https://issues.apache.org/jira/browse/HBASE-1406
> Project: HBase
> Issue Type: Improvement
> Components: regionserver
> Affects Versions: 0.20.0
> Reporter: Nitay Joffe
>
> The current HRS implementation is a mess, especially after ZooKeeper additions to handle Session Expired events. See HBASE-1311.
> It contains logic to restart itself which caused a lot of fields to be non-final that can be. It should be split out to two separate class, one that runs HRS duties and one that watches over the first in a join/restart loop. Using this means a special event like a ZooKeeper Session Expired wouldn't require special restart code, just an abort. The new wrapper class will handle restarting.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)