Datastage PX features for Cluster or Grid

santoshkumar · Post by **santoshkumar** » Mon Feb 13, 2006 2:36 am

I wanted to know how can we leverage Datastage PX features with Grid Computing?
Does DS support grid computing

Ascential/IBM market Datastage enterprise edition as grid enable integration tool.
If yes then how can we use pool of server for running the datastage jobs?
I understand that datastage PX treats grid as MPP cluster. And if we want to use cluster what parameter values needs to set for setting up DataStage server. eg configuration file, rsh etc etc.
SK

ArndW · Post by **ArndW** » Mon Feb 13, 2006 3:46 am

I would suggest starting in the Parallel Job Developer's Guide, pages 58-9 onwards. This details the basic configuration setup required by PX to run on a distributed system configuration - as you've stated this is the way that a future true grid solution will work with PX jobs. Although this is not yet a true grid construct, it will distribute the load across heavy servers - the next generation(s) might have capabilities to distribute to "light" processing nodes. If you think about it, a true grid for ETL jobs isn't always a benefit, since the time spent sending requests out and returning result sets or distributed writing to databases is higher than the amount of processiong time gained by doing things in parallel.

ray.wurlod · Post by **ray.wurlod** » Mon Feb 13, 2006 4:13 am

There was a paper presented at IBM Information Integration Live last November in Las Vegas on using DataStage EE in a grid environment. Can you get access to the papers from this conference?

santoshkumar · Post by **santoshkumar** » Tue Feb 14, 2006 6:26 am

Thanks ArndW/ray.wurlod for your inputs.