DSXchange

Hi,

When calling a parallel job from a job sequence, sometimes the job fails because its parameters are empty. The job sequence and the parallel job are multi-instanced.

When we try again it works.

Well... with nothing else to go on other than "when we try again it works", I'm going to give you my standard response to that: resources. As in not enough at that time for things to run properly. When these linked multi-instance processes run, how many are typically running at the same time?

Hi,

Sorry for the delay, I missed the notification.
Your answer is the one that i was afraid of, even if my previous post was incomplete.

We have two behaviours:

1) Relaunch ok without any actions
2) Relaunch ok after a clear status of the aborted instance

Thank you.

Okay. Not seeing anything that would change my answer or the direction I think you need to go with the investigation. And from what I recall of your second behavior, that's normal - unless we're talking about a Sequence job with the 'restartable' property enabled? Otherwise you need to clear the status.

Of course you could always open a support case and get some official help, but failing that I'd still be looking into how many you are attempting to run at the same time. Seeing as how that's the whole point of making things multi-instance, we at least should get that off the table first. So... how many are you spooling up at once?