Warning: Job control process (pid XXXX) has failed

manish1005 · Post by **manish1005** » Fri Jan 05, 2007 4:55 am

I am facing a strange problem. When I run a job it executes successfully with correct results except for the fact that one warning [ Warning: Job control process (pid XXXX) has failed ] is logged in between.

As I already said results are correct and no other effect of this warning is visible. This warning comes *most* of the times in all the server jobs and I am unable to trace the source/cause of the warning.

The problem comes when I use sequences. Because this warning doesn't let the first job stage to trigger another job(even though the first stage gave correct as expected results) and execution of the sequence stops after first jobstage only.

Below is the log scrap as seen through Director:

[Green Y] Environment variable settings: ALLUSERSPROFILE=C:\Documents and Settings\All Users BELL=^G......
........

[Red !] Warning: Job control process (pid 4152) has failed

[Green Y] Control: Job CopyOfproduct_class aborted

[Green Y] Info: CopyOfproduct_class: Set NLS locale to US-ENGLISH,US-ENGLISH,US-ENGLISH,US-ENGLISH,US-ENGLISH

[Green Y] Info: CopyOfproduct_class..Transformer_1: DSD.StageRun Active stage starting, tracemode = 0.

[Green Y] Info: CopyOfproduct_class..Transformer_1: Set NLS locale to US-ENGLISH,US-ENGLISH,US-ENGLISH,US-ENGLISH,US-ENGLISH

[Green Y] Info: CopyOfproduct_class..Transformer_1: DSD.StageRun Active stage finishing.
110 rows read from DSLink7
110 rows written to DSLink14
0.047 CPU seconds used, 0.203 seconds elapsed.

[Green Y] Info: Finished Job CopyOfproduct_class.

Please suggest whats wrong?

kumar_s · Post by **kumar_s** » Fri Jan 05, 2007 5:39 am

What is the JobSequece desing?
Do you have any before job subroutine? Is it pertaining to only to Jobsequence or to all other?

manish1005 · Post by **manish1005** » Fri Jan 05, 2007 6:05 am

>What is the JobSequece desing?
My Job sequence is nothing but two jobstages linked one after the another, so that the second job executes after the first gets done. (This is the simplestic case which I am trying and later on will proceed to more complex one).

>Do you have any before job subroutine? Is it pertaining to only to Jobsequence or to all other?
No I dont have any before or after subroutines.

The problem is not with the sequence but with the jobs as even when I run individual jobs I get that warning(though results are correct and as per expectations).

DSguru2B · Post by **DSguru2B** » Fri Jan 05, 2007 7:41 am

Reset the job and see if there are any messages that pop up from the 'previous run' that are helpful.

ray.wurlod · Post by **ray.wurlod** » Fri Jan 05, 2007 2:05 pm

Do the individual jobs that generate the "job control process has failed" message include job control code? Are they reported as being run under control (that is, is DSJobController reported in the detailed Status view or in the "job starting" and "job finishing" log events?

If not, I have to disagree with your diagnosis that the problem is not in the job sequence.

manish1005 · Post by **manish1005** » Sat Jan 06, 2007 4:31 am

DSguru2B
>Reset the job and see if there are any messages that pop up from the 'previous run' that are helpful.
Resetting the Job also does not give any additional useful info. Even the same warning comes during the reset.

Ray
>Do the individual jobs that generate the "job control process has failed" message include job control code? Are they reported as being run under control (that is, is DSJobController reported in the d ...

Through Director I am not able to find any Job control code. The user remains "machineName\Administrator" for all the log entries.

kumar_s · Post by **kumar_s** » Sat Jan 06, 2007 9:32 pm

In the example that you have given, CopyOfproduct_class is it a job or the Jobsequence that calls the Job?

manish1005 · Post by **manish1005** » Sun Jan 07, 2007 10:30 pm

CopyOfproduct_class is a job. The Director logs that I have given in the first post are of only one job(CopyOfproduct_class) running independently and not in sequence. (Though the logs for the job are exactly similar when I run it in sequences).

Any idea anybody?

ray.wurlod · Post by **ray.wurlod** » Sun Jan 07, 2007 10:46 pm

This is intriguing. I've been able to create a job that has exactly this issue (using the DCS Strategies site). The job is not run from a job sequence; it looks like this:

Code: Select all

SequentialFile  ----->  CopyStage  ----->  DataSet
   |                         |                |
   | Reject                  |                | Reject
   V                    Aggregator            V
  Peek                       |               Peek
                             |
                             V
                            Peek

The only warning message had this text:
Job control process (pid 18831) has failed

The immediately following (informational!) message has this text:
Job Engagements_Aggregator. aborted

However, the job ran successfully to completion. The expected number of rows was processed on every link.

Mysteriouser and mysteriouser!

manish1005 · Post by **manish1005** » Sun Jan 07, 2007 11:50 pm

Hi Ray,
My job is also quite similar, and the errors are same.

Any wild thoughts about how to tackle the issue. I can not make sequences with this pseudo error which gives the right results but with warnings which actually abort the conditional sequence.

kumar_s · Post by **kumar_s** » Mon Jan 08, 2007 12:03 am

Ray's job is PX. How about yours?
If your also on PX, you have a work around. Promote that warning to Info using Message handler. If server, better approach Support.

manish1005 · Post by **manish1005** » Mon Jan 08, 2007 12:15 am

Hi kumar,
Mine is a server job.

Contacting and getting response from support takes quite some time, I believe someone must have experienced the same problem earlier.......

kumar_s · Post by **kumar_s** » Mon Jan 08, 2007 12:47 am

Check if the same appears in other projects?

ray.wurlod · Post by **ray.wurlod** » Mon Jan 08, 2007 4:44 am

Further, my test was on a Linux machine, not a Windows machine, so that's another variation on your scenario. Version of DataStage was 7.5.1A. Haven't had another chance to look at it. Have a horrible feeling it's a spurious message, but how to prove that?

The topic ID for message handler is DSTAGE_RUN_E_0441. Ask your support provider what might generate this particular message.

manish1005 · Post by **manish1005** » Mon Jan 08, 2007 5:25 am

Kumar,
yeah, it appears same in other projects as well

Ray,
Mine is win2000 server machine, and the version of DS is same 7.5.1A. (coundn't read rest of your message as I am not a premium member yet).