[
https://issues.apache.org/jira/browse/HBASE-14659?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14968421#comment-14968421
]
Heng Chen commented on HBASE-14659:
-----------------------------------
I notice that in this testcase, some test function use
{{MiniMapReduceCluster}} to run MR job, some not. For example,
{{testWritingPEData}} not use {{MiniMapReduceCluster}} but
{{testMRIncrementalLoad}} use it.
And tests which use {{MiniMapReduceCluster}} run very slowly.
Could all tests run MR without {{MiniMapReduceCluster}}? It seems to be much
faster.
I don't know what's the difference between the two approaches.
> Fix flakey TestHFileOutputFormat
> --------------------------------
>
> Key: HBASE-14659
> URL: https://issues.apache.org/jira/browse/HBASE-14659
> Project: HBase
> Issue Type: Bug
> Reporter: Heng Chen
>
> As i said in HBASE-14654
> This testcase was hanged twice recently.
> the two QA information:
> https://builds.apache.org/job/PreCommit-HBASE-Build/16118/console
> https://builds.apache.org/job/PreCommit-HBASE-Build/16117/console
> I notice that the two QA running simultaneously on same machine H0.
> I doubt If there are some common resources the two QA conflicts.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
