[
https://issues.apache.org/jira/browse/HADOOP-2490?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HADOOP-2490:
--------------------------
Attachment: 2490-v4.patch
Fix it so hdfs exceptions don't fail tests, move the verify of mapreduce job
before shutdown of cluster to see if makes a difference up on hudson and add
some debug output so can get more info on why TestScanner2 hangs on cluster
startup.
> [hbase] Failure in nightly #346
> -------------------------------
>
> Key: HADOOP-2490
> URL: https://issues.apache.org/jira/browse/HADOOP-2490
> Project: Hadoop
> Issue Type: Bug
> Components: contrib/hbase
> Reporter: stack
> Assignee: stack
> Attachments: 2490-v4.patch, totext-2.patch, totext-3.patch,
> totext.patch
>
>
> hbase tests failed in the nightlies and in a few previous patch builds.
> Suspect is use of TextSequence in place of creating new Text instances every
> time (though, it passed its patch build and a nightly succeeded with
> TextSequence in place). I reviewed the patch. Looks like all TextSequence
> uses should be safe. Looking at console for last night failures, at least
> one seems to be complaint about hdfs accesses. Will wait till hudson starts
> working again building patches regularly before trying experiments -- e.g.
> backing out TextSequence patch to see if it makes a difference.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
