Page 1 of 1

Hardware configuration for DataStage 7.x Parallel extender

Posted: Thu Apr 08, 2004 8:52 am
by Agnivesh
Hi all
what is the Hardware configuration required for DataStage 7.x Parallel extender.
Thanks
Agni...

Posted: Thu Apr 08, 2004 4:34 pm
by ray.wurlod
1. UNIX. (PX is scheduled for Windows, perhaps in 7.1?)

2. More. The secret of PX is its ability to share the load over multiple "resources" (= CPU + memory + work disk + scratch disk). These can be in an SMP (symmetric multi-processor) environment or in an MPP (massively parallel processing) environment. The latter can be on a group (cluster) of machines.

Hardware configuration for DataStage 7.x Parallel extender

Posted: Sun Apr 11, 2004 10:06 pm
by Agnivesh
Hi Ray Wurlod,Thanks for ur reply....
Actually we need following information....
1) What is processor speed?
2) Memory requirements....
3) Hardware requirement
4) Other thing requirement for Data Stage 7.X on windows as well as UNIX....

Thanks
Narayana G



1. UNIX. (PX is scheduled for Windows, perhaps in 7.1?)

2. More. The secret of PX is its ability to share the load over multiple "resources" (= CPU + memory + work disk + scratch disk). These can be in an SMP (symmetric multi-processor) environment or in an MPP (massively parallel processing) environment. The latter can be on a group (cluster) of machines.
Ray Wurlod
ABN 57 092 448 518
Education and Consulting Services

Posted: Mon Apr 12, 2004 3:19 pm
by ray.wurlod
CPU speed doesn't really matter, though faster processors will get through the same number of cycles in a shorter period. Your question is currently in the "how long is a piece of string?" category. More, slightly slower CPUs can be better than fewer, slightly faster CPUs in a parallel processing environment.

Memory. At least 512MB, preferably at least 1GB. More than this is always better; many DataStage operations are (deliberately designed to be) memory-hungry.

Disk space: allow 200MB per DataStage project, and LOTS of scratch space. Again, with no indication of what your jobs are intending to do, it's hard to say how much. In addition, if you're going to make extensive use of hashed files and staging files, allow space for these too. What volume of data are you planning to process per run? You need at least this much staging space.

On Windows platforms DataStage server must be installed on an NTFS partition.

Actual hardware configuration is not relevant. For server jobs you are restricted to a single server machine (though it can have multiple CPUs). Parallel jobs can execute in this environment too, or in an MPP cluster.

Parallel Extender's availablility on Windows has not yet been announced.