datastage

[sree27]

How the process is going in parallelisam?

[quality_quantity]

If you are upgrading from server to parallel jobs then consider an iterative implementation rather then a big bang.

Small volume jobs may actually be faster if left as server jobs as parallel jobs can have a longer start up time.

Server and parallel jobs will run quite happily from the same sequencer or scheduler.

BASIC transformers and server job containers can be put into parallel jobs. This gives you a parallel job that re-uses your old transformation routines, it will be slower then a 100% parallel job but this might not matter, it will still be a lot faster then the old server job and be faster to build.

PeopleSoft-EPM upgrades will continue to deliver server jobs so the more you can leave as server versions the easier the upgrade path.

Those with an SOA requirement can turn either server or parallel jobs into real time services.

Just change the very high volume jobs and do some performance testing to compare different designs. Release a small number to production to see how they run.