[
https://issues.apache.org/jira/browse/HBASE-15631?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16207800#comment-16207800
]
Andrew Purtell commented on HBASE-15631:
----------------------------------------
bq. ServerName.java seems to have brought HostAndPort back.
ServerName copied verbatim from branch-2 to be compatible. This is not my
doing. Should we patch ServerName everywhere? We could on a separate JIRA.
bq.Changes in ProtobufUtil.java seem unnecessary
I thought I removed these. Let me check. There is a new RSGroupsProtobufUtil
that should contain all new changes.
bq. The HMaster changes look "interesting". Looks like there could be NPE
condition with regionState in the reshuffled if-statement (the original seems
more correct). And I also do not follow the unconditional call to
randomAssignment there. Are those changes needed at all?
These changes come from the original branch-1 backport patch. I'm not sure. We
can try changing this code to look more correct and see if something expected
by RSGroups breaks. Thanks for having a look. This is great feedback.
> Backport Regionserver Groups (HBASE-6721) to branch-1
> ------------------------------------------------------
>
> Key: HBASE-15631
> URL: https://issues.apache.org/jira/browse/HBASE-15631
> Project: HBase
> Issue Type: New Feature
> Affects Versions: 1.4.0
> Reporter: Francis Liu
> Assignee: Francis Liu
> Fix For: 1.4.0, 1.5.0
>
> Attachments: HBASE-15631-branch-1.patch, HBASE-15631.branch-1.patch,
> HBASE-15631.patch
>
>
> Based on dev list discussion backporting region server group should not be an
> issue as it does not: 1. destabilize the code. 2. cause backward
> incompatibility.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
