Page 2 of 2

Posted: Mon Aug 10, 2015 4:27 pm
by ray.wurlod
Try removing -wait from the dsjob command. Technically -wait and -jobstatus are incompatible options. The -jobstatus option will still cause the command to wait until the job completes, so that it can report the job status.

Posted: Tue Aug 11, 2015 8:25 am
by neeraj
Hello,

I m trying to explain the issue once again and Hope it helps.

I have a multi Instance Job. At same time 2 instances of the same job having different invocation ID are running in Parallel. Most of times, Jobs runs with Success i.e. from 20-July till today Job failed only once.

My Analysis is:-
Sometimes both the instances get executed at exact same time and first instance gets the control and executes while the second instance doesn't get the handle and job aborts. While re-running, the failed jobs again, it run successfully

My Question:-
Do I need to make any change in the environment level to increase the wait time or some other setting which will make both the instances of MultiInstance Job execute in Parallel.

Regards
Neeraj

Posted: Tue Aug 11, 2015 9:24 am
by ArndW
The problem here might lie not in the concurrent execution of job sequences, but the way that DataStage parameter set files are opened. The parameter set files might be opened in exclusive mode for reading, and this could theoretically cause sporadic issues. I've started literally hundreds of concurrent instances in projects before without this problem, but they didn't use parameter sets.

Posted: Tue Aug 11, 2015 11:53 am
by neeraj
Hello Arn,

I am sorry but due to some limitation could not read your complete sentence. But We are not using invoking any parameter set in this command.

Regards
Neeraj

Posted: Wed Aug 12, 2015 8:54 am
by ArndW
In that case I don't see where the problem could be coming from.