[ 
https://issues.apache.org/jira/browse/UIMA-2842?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13655232#comment-13655232
 ] 

Lou DeGenaro commented on UIMA-2842:
------------------------------------

Although a significant part of Job processing, the Job Driver (JD) is not 
responsible for management of overall Job state.  That responsibility falls on 
the Orchestrator (OR).  Until now, the OR did not venture into user space (ie 
the user's log directory).  With newly added functionality to ducc_ling, the OR 
can now easily record to the user's log directory the states (and rationale, if 
any) that occur during the Job's lifetime into newly established file ducc.log.

Example of successful job:

degenaro@bluej62:~/ducc/logs/267$ tail -f ducc.log
2013.05.11 06:51:17 Sat Received
2013.05.11 06:51:17 Sat WaitingForDriver
2013.05.11 06:51:37 Sat WaitingForResources
2013.05.11 06:51:49 Sat Initializing
2013.05.11 06:57:22 Sat Running
2013.05.11 06:59:37 Sat Completing
2013.05.11 06:59:59 Sat Completed

Example of failed job, due to missing service dependency:

degenaro@bluej62:~/ducc/logs/268$ tail -f ducc.log
2013.05.11 06:53:03 Sat Received
2013.05.11 06:53:03 Sat WaitingForDriver
2013.05.11 06:53:27 Sat WaitingForServices
2013.05.11 07:03:20 Sat Completing
2013.05.11 07:03:20 Sat ServicesUnavailable
2013.05.11 07:03:20 Sat service manager reported 
UIMA-AS:FixedSleepAE_1:tcp://bluej9999:61616:not-available;
2013.05.11 07:03:39 Sat Completed


                
> Some reasons for job failures are not displayed in WS or recorded in 
> jd.err.log
> -------------------------------------------------------------------------------
>
>                 Key: UIMA-2842
>                 URL: https://issues.apache.org/jira/browse/UIMA-2842
>             Project: UIMA
>          Issue Type: Bug
>          Components: DUCC
>            Reporter: Burn Lewis
>            Assignee: Lou DeGenaro
>            Priority: Minor
>             Fix For: 1.0-Ducc
>
>
> A job (P7-2828) failed with hovers of "not runnable" & "no more work for 
> process"... no other clues in WS or logs.  But I found in the OR log it was 
> because of "services unavailable".  "Not runnable" is not a very informative 
> reason!

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to