DSXchange

There are no bugs in 8.5.
Only undocumented features and opportunities.

---
Paul Christensen
Information Agenda Architect - IBM Senior Certified Professional
Distribution, Industrial & Life Sciences

Yes, as noted above.

Did you bother to check the things I suggested?

Or are you looking for Harry Potter to come along and magically repair your environment?

qutesanju wrote:Does datastage 7.5.1.A server edition is compactible with windows vista?

No.

qutesanju wrote:has anybody installed datastage 7.5.1.A server edition into windows vista?

Not as far as I am aware.

No osh process with the -APT_PMsectionLeaderFlag option is a job; it is a section leader in a parallel job.

Do you want this solved as a DataStage task or as a "batch code" task?

If the latter, why did you post in a DataStage forum?

What you are asking is not reasonable. Where would it end? How many levels of "run a job from the before-job routine of a job" would you have? While the technology does support such an approach, a sequence (as Craig noted) is by far a better, and more easily maintained, approach.

Change your expectations. There is a finite upper limit.

Don't do the parsing in the file stage - read the entire line as a single string and do the parsing in the Transformer stage.

(Test read performance with just a string - don't do any parsing - this is as fast as you'll ever get.)

The DSN you used is not defined in the uvodbc.config file being used by the project. Or, perhaps, that DSN is not defined:

• (a) if you are on UNIX in the $ODBCINI file
  
  (b) if you are on Windows as a System or File DSN (DataStage does/can not use User DSNs)

Why do you believe there's a problem? DataStage uses STREAMS I/O module, so should be pretty slick. Create this job to test read speed. SeqFile -----> Transformer -----> SeqFile In the Transformer stage, specify @FALSE as the constraint expression. The target sequential file stage can append to the ...

This is a horizontal pivot, and is accomplished with a Pivot stage. First, parse the country codes into separate fields. Then apply the pivot. Downstream of that, filter out those rows with NULL country codes.

Either use the full pathname of the dsjob command or include its parent directory in your command search list (PATH environment variable).

Have you checked with IBM for any known issues? What platform is your Engine tier running on?