[
https://issues.apache.org/jira/browse/PIG-1124?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12828005#action_12828005
]
Ashutosh Chauhan commented on PIG-1124:
---------------------------------------
Thanks for the review, Dmitriy.
AFAIK, there aint any other property that Pig overrides such that user can't
set it himself using -D switch. Correct me if I am wrong. job.name was
special-cased so that if user doesnt provide job name, pig sets the name itself
to make it easy to track jobs submitted by pig on the cluster. Bug in this
ticket is Pig was always setting the name(instead of setting only when user
wasnt supplying any), thus completely ignoring user supplied name through -D
switch.
> Unable to set Custom Job Name using the -Dmapred.job.name parameter
> -------------------------------------------------------------------
>
> Key: PIG-1124
> URL: https://issues.apache.org/jira/browse/PIG-1124
> Project: Pig
> Issue Type: Bug
> Components: impl
> Affects Versions: 0.6.0
> Reporter: Viraj Bhat
> Assignee: Ashutosh Chauhan
> Priority: Minor
> Fix For: 0.7.0
>
> Attachments: pig-1124.patch
>
>
> As a Hadoop user I want to control the Job name for my analysis via the
> command line using the following construct::
> java -cp pig.jar:$HADOOP_HOME/conf -Dmapred.job.name=hadoop_junkie
> org.apache.pig.Main broken.pig
> -Dmapred.job.name should normally set my Hadoop Job name, but somehow during
> the formation of the job.xml in Pig this information is lost and the job name
> turns out to be:
> "PigLatin:broken.pig"
> The current workaround seems to be wiring it in the script itself, using the
> following ( or using parameter substitution).
> set job.name 'my job'
> Viraj
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
