hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Doug Cutting (JIRA)" <j...@apache.org>
Subject [jira] Updated: (HADOOP-111) JobClient.runJob() should return exit status for a job.
Date Sat, 08 Apr 2006 00:05:36 GMT
     [ http://issues.apache.org/jira/browse/HADOOP-111?page=all ]

Doug Cutting updated HADOOP-111:

    Component: mapred

> JobClient.runJob() should return exit status for a job.
> -------------------------------------------------------
>          Key: HADOOP-111
>          URL: http://issues.apache.org/jira/browse/HADOOP-111
>      Project: Hadoop
>         Type: Bug

>   Components: mapred
>     Reporter: Andrzej Bialecki 

> JobClient.runJob() doesn't return any values. Any information about the exit status of
a job is discarded (or appears only in logs).
> It should be possible to return the exit status, so that JobClient users can determine
whether a job was successfully completed or not. This is also important when using cmd-line
tools in shell scripts - currently they don't return any exit codes, because it's not possible
to determine the outcome of a job submitted through JobClient. As a consequence, it's difficult
to automate repetitive jobs using shells scripts.
> It would be also nice to have an exit message in case of errors, for human consumption.
> I propose to implement one of the following:
> * change the return type of this method from void to int
> * or, better yet, to put the exit code and optional exit messages inside the JobConf
instance under pre-defined keys.

This message is automatically generated by JIRA.
If you think it was sent incorrectly contact one of the administrators:
For more information on JIRA, see:

View raw message