How to decide which partition to be used in which stage

Rameshgoldenhill · Post by **Rameshgoldenhill** » Sun Mar 14, 2010 9:16 am

How to use parallelism in data stage job and what is the best possible partition selection in which stage to optimize performance and parallelism. Thanks in advance

ray.wurlod · Post by **ray.wurlod** » Sun Mar 14, 2010 11:56 am

Begin by reading Chapter 2 of the Parallel Job Developer's Guide then post any specific questions that you may have.

As it stands your question is too vague. There is no single "best" algorithm, there's no general solution to "optimize parallelism", and you would need to define what you mean by "performance" in an ETL context (rows/second is an almost meaningless metric).

John Smith · Post by **John Smith** » Sun Mar 14, 2010 4:59 pm

Have you attended the IBM datastage training? that might be a good start.