Page 1 of 1

Process Id Failed

Posted: Mon Jan 28, 2008 4:03 am
by asitagrawal
Hi Everyone,

I have a Sequence, which invokes many jobs. Jobs and the Sequence are with multiple Invocation ids (ofcourse all unique).

There are 10 Invocations of the sequence, and each sequence invoking 10 instances of the jobs.

One particular job, let me call it as Job1, is resulting in error Job control process (pid xxxx) has failed.
Follow this link for the job design: JobDesign

This error has been occuring randomly during the past month. The Sequence is scheduled to run every week, from Mon to Fri. :(

At the time of failure of the process id, an error even is logged in Windows Log, which is:

Code: Select all

Event Type:	Error
Event Source:	Ascential DataStage
Event Category:	None
Event ID:	1004
Date:		26-01-2008
Time:		20:36:55
User:		N/A
Description:
DataStage error: Unhandled Exception raised at address 0x1017DC8C : Access violation.  Attempted to read from address 0x01391000.  Binary data is processor CONTEXT structure..
Data:
0000: 3f 00 01 00 00 00 00 00   ?.......
.
.
.
.
I have recompiled the job once, after the error occurred for the first time.

Please help me to understand the possible reasons, for Why the same error is repeating in the same job !

Please advice.

Posted: Mon Jan 28, 2008 8:20 am
by chulett
I'm guessing you are having resource issues but can't know for sure. Next time one of the jobs abort, try Resetting the job and let us know the contents of the 'From previous run...' log message it (hopefully) generates.

Posted: Mon Jan 28, 2008 8:41 am
by asitagrawal
I will definitely do this...
But, unfortunately, in the latest run, there is no such entries logged...
Are these always logged or they are conditional ?

Posted: Mon Jan 28, 2008 8:43 am
by asitagrawal
Hey..I have it...
Sorry, I did not check that earlier..

The same error occurred on 23Jan also...
This is the "From previous run:"

Code: Select all

From previous run 
DataStage Job 3 Phantom 468
Program "DSD.RUN": Line 593, Unable to unlink the operating system file "DSD.RUN_79218_14633".	  
Program "DSD.RUN": Line 593, Unable to unlink the operating system file "DSD.StageRun_79219_14633".	  
Job Append_CurrentBatch_PP_RevCDeqR.S01_IN05 must be reset before it can run.

Posted: Mon Jan 28, 2008 7:27 pm
by Teej
For some reason, there was a different process/run that have those files locked, and it was not cleaned up properly.

This sounds like a potential bug. What version of DataStage are you using, exactly?

Posted: Tue Jan 29, 2008 5:46 am
by asitagrawal
Teej wrote: This sounds like a potential bug. What version of DataStage are you using, exactly?
Could be... actually I went for a version upgrade.. so we actually had to upgrade from 7.5.1A to 7.5.2.. thing went fine..but the actual screwup happened when we had overwritten the project directories with the backup of 7.5.1A !!!

So, for that matter... I can relate, anything which is now going wrong, to this act of overwriting the project directories !!

So, now what I see is, client version as 7.5.1A and server version as 7.5.2.