[
https://issues.apache.org/jira/browse/HADOOP-1663?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
lohit updated HADOOP-1663:
--------------------------
Attachment: h-1663-streamexception-as-returncode.patch
Thanks to Owen for inputs. Now instead the code returns error code which is
caught in main of HadoopStreaming and exit accordingly. Attached in the
modified patch.
Thanks
> streaming returning 0 when submitJob fails with Exception
> ---------------------------------------------------------
>
> Key: HADOOP-1663
> URL: https://issues.apache.org/jira/browse/HADOOP-1663
> Project: Hadoop
> Issue Type: Bug
> Components: contrib/streaming
> Affects Versions: 0.13.1
> Reporter: Koji Noguchi
> Assignee: Koji Noguchi
> Fix For: 0.15.0
>
> Attachments: h-1663-streamexception-as-returncode.patch,
> h-1663-streamexception-mod.patch,
> h-1663-streamexception-simpletestcase.patch, h-1663-streamexception.patch
>
>
> In streaming, if I pass an input filename that doesn't exist,
> bash-3.00$ hadoop jar hadoop-streaming.jar -mapper "/bin/cat" -reducer
> "/bin/cat" -input "inputfoo" -output "output"
> ...
> 07/07/28 07:37:18 ERROR streaming.StreamJob: Error Launching job : Input path
> doesnt exist : /user/knoguchi/inputfoo
> bash-3.00$ echo $?
> *0*
> =================
> Copy&Pasting Milind's comment.
> its a hadoop streaming problem. See the code
> below:
> }catch( IOException ioe){
> LOG.error("Error Launching job : " + ioe.getMessage());
> }
> finally {
> if (error && (running_ != null)) {
> LOG.info("killJob...");
> running_.killJob();
> }
> jc_.close();
> Note that it catches exception, and does not rethrow it !!!
> =================
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
