[
https://issues.apache.org/jira/browse/PHOENIX-6104?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17255313#comment-17255313
]
Istvan Toth commented on PHOENIX-6104:
--------------------------------------
Looking at the code, it seems that 2.0 will simply refuse to split regions that
RegionPolicy.shouldSplit() doesn't want to split.
org.apache.hadoop.hbase.regionserver.RSRpcServices.getRegionInfo(RpcController,
GetRegionInfoRequest) seems to be the culprit.
This seems dumb to me, as there are legitimate reasons to pre-split an existing
table, I don't see why the RS conflates +_isSplittable_+ with +_shouldSplit_+.
If you are not on PTO, it would be great if you could get more information on
why this change was instituted.
It's a pain for our tests, but I imagine that it would also be a pain for
anyone doing bulk loads, etc., where they want to pre-create regions on
existing tables.
In 2.4, a _canSplit_ filed has been added to SplitPolicy, but it's not actually
used by the RS to set the _splittable_ field.
> SplitSystemCatalogIT tests very unstable with Hbase 2.3
> -------------------------------------------------------
>
> Key: PHOENIX-6104
> URL: https://issues.apache.org/jira/browse/PHOENIX-6104
> Project: Phoenix
> Issue Type: Bug
> Components: core
> Affects Versions: 5.1.0
> Reporter: Istvan Toth
> Assignee: Istvan Toth
> Priority: Major
> Attachments: 6104-testouput.log
>
>
> The failure is in the test preparation code, where we split the system
> catalog table, and it seems to be a HBase issue, rather than a Phoenix one,
> but we need to track the issue.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
