hadoop-pig-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Thejas M Nair (JIRA)" <j...@apache.org>
Subject [jira] Commented: (PIG-1570) native mapreduce operator MR job does not follow same failure handling logic as other pig MR jobs
Date Mon, 30 Aug 2010 17:23:54 GMT

    [ https://issues.apache.org/jira/browse/PIG-1570?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12904258#action_12904258
] 

Thejas M Nair commented on PIG-1570:
------------------------------------

Regarding 
bq. Another thing to investigate (somewhat related) - there seems to be a problem when PigServer
is used to execute query having native mr operator - i was unable to run the tests in local
mode . But i am able to run query in local mode from commandline.

The problem was that in test setup, the MiniCluster hadoop-site.xml (~/pigtest/conf/hadoop-site.xml)
is in classpath. The WordCount.jar would end up trying to run the MR job using minicluster
and fail, if rest of the test is using local mode.


> native mapreduce operator MR job does not follow same failure handling logic as other
pig MR jobs
> -------------------------------------------------------------------------------------------------
>
>                 Key: PIG-1570
>                 URL: https://issues.apache.org/jira/browse/PIG-1570
>             Project: Pig
>          Issue Type: Bug
>            Reporter: Thejas M Nair
>            Assignee: Thejas M Nair
>             Fix For: 0.8.0
>
>
> The code path for handling failure in MR job corresponding to native MR is different
and does not have the same behavior.
> For example, even if the MR job for mapreduce operator fails, the number of jobs that
failed is being reported as 0 in PigStats log.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message