[
https://issues.apache.org/jira/browse/HADOOP-5437?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Amareshwari Sriramadasu updated HADOOP-5437:
--------------------------------------------
Attachment: patch-5437.txt
Attaching the patch fixing the testcase.
The problem was: The configuration mapred.job.reuse.jvm.num.tasks wa set to -1,
but never passed to cluster. Moreover, it should be job configuration, and not
cluster configuration, since the configuration is a job-level configuration at
tasktarcker.
With the patch I verified jvm reuse is on and each jvm runs multiple tasks.
> Unit test for jvm -reuse
> ------------------------
>
> Key: HADOOP-5437
> URL: https://issues.apache.org/jira/browse/HADOOP-5437
> Project: Hadoop Core
> Issue Type: Bug
> Components: mapred
> Affects Versions: 0.19.0
> Reporter: Amareshwari Sriramadasu
> Priority: Minor
> Fix For: 0.21.0
>
> Attachments: patch-5437.txt
>
>
> TestMiniMRDFSSort.testMapReduceSortWithJvmReuse actually does not test jvm
> running multiple tasks. In the test, each jvm runs only one task.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
