datastage job parallelism on XP platform

Aggie99 · Post by **Aggie99** » Tue Feb 24, 2009 5:48 pm

I have datastage server/client installed all in one XP machine. This is to facilitate my learning of the tool.

Given that is my hardware limitation!

My question is, does it make any difference (in processing time) whether I set up the configuration as 2 nodes, 4 nodes, etc. Can I still enjoy parallelism in DataStage?

thanks for your inputs.

IBM Analytics Champion 2009 - 2020 · Post by **asorrell** » Tue Feb 24, 2009 7:05 pm

I assume that you are on a small XP box, though you didn't say what kind of CPU horsepower it has. You can certainly configure any number of nodes for practice. There is no actual connection between CPU's and nodes (something I hammer into my students during DataStage classes). DataStage lets the O/S figure out the best way to service its requests.

However, with that said - remember that the more nodes that are active the more simultaneous requests for processing are taking place on that system. In your case I'd recommend just configuring two nodes. Any job that works correctly with two nodes should work correctly no matter what number of nodes are in place. Two nodes allows you to test parallelism without overloading your machine.

ray.wurlod · Post by **ray.wurlod** » Tue Feb 24, 2009 9:50 pm

We don't have enough information about your personality to make any statement about what you find enjoyable.

Other than that, as Andy said, it's simple (resource) supply and demand.

Aggie99 · Post by **Aggie99** » Tue Feb 24, 2009 11:31 pm

thanks Andy, that surely helps.

Ray, thanks for your humor remark regarding my "datastage enjoyment"