[
https://issues.apache.org/jira/browse/HBASE-4402?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13121559#comment-13121559
]
stack commented on HBASE-4402:
------------------------------
Looks like the testMergeTable failure is because of this patch. Adding the
below gets us past the testMergeTable failure (NPEs causing master shutdown and
then test would hang looking for the absconded master). This test is weird too
anyways hand-creating regions in fs then bringing all online; needs to be
refactored but meantime....
{code}
68,69c140,143
< + List<ServerName> localServers = serversByHostname.get(
< + oldServerName.getHostname());
---
> + List<ServerName> localServers = new ArrayList<ServerName>();
> + if (oldServerName != null) {
> + localServers = serversByHostname.get(oldServerName.getHostname());
> + }
76c150
< + oldHostsNoLongerPresent.add(oldServerName.getHostname());
---
> + if (oldServerName != null)
> oldHostsNoLongerPresent.add(oldServerName.getHostname());
{code}
> Retaining locality after restart broken
> ---------------------------------------
>
> Key: HBASE-4402
> URL: https://issues.apache.org/jira/browse/HBASE-4402
> Project: HBase
> Issue Type: Bug
> Components: master
> Affects Versions: 0.92.0
> Reporter: Todd Lipcon
> Assignee: Todd Lipcon
> Priority: Blocker
> Fix For: 0.92.0
>
> Attachments: hbase-4402.txt, hbase-4402.txt
>
>
> In DefaultLoadBalancer, we implement the "retain assignment" function like so:
> {code}
> if (sn != null && servers.contains(sn)) {
> assignments.get(sn).add(region.getKey());
> {code}
> but this will never work since after a cluster restart, all servers have a
> new ServerName with a new startcode.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
