We are currently migrating from 7.5 to 8. All our jobs indivdually run fine, but when we start scheduling everything fails. We've been able to track this down to a multi-instance job which is used in all our flows. If this is triggered to start too much at the same moment(with other invocationid's ) it fails. (status 99 or 3).
We've created a workflow which is just triggering this job 5 times in parallel. This fails even without any other activities on the server. CPU is then used for about 5% .
IBM support says:
I find this very hard to believe and if it is the case we have no choice but to return to version 7 where we can trigger a lot more jobs.Following my investigations on DataStage architecture it appears that you cannot run more job in a second than the number of CPU installed on the server, there would be a dispatch issue. This is valid for DataStage 8
Anybody had simular problems and what would then be the solution?