DSXchange

Certification is always a good idea. It helps in creating more awareness about Datastage. I have no idea if any ETL tools conducting certification exams. But i know of Reporting tools like Cognos having certification exams.

Regards

A workaround:

We had a similar problem. OCI9i stage was not working. Hence we
reverted to OCI8 box pointing to Oracle 9i database

Regards

Problem of incorrect Aggregation is very critical functional issue.
We found a solution after a trying a lot of combinations.
Mark as key only the "Group" columns. Rest of the columns
should neither be key nor be marked as "Group

You can add/View the enviormental variables from inside the lookup stages (doing a right click on the mapping box->Job Parameter-> Add envirornment variabl) inside the transformer.

Regards

Write the output of batch"A" to a flat file. In Batch B read the contents
of the flat file to which batch "A" has written.

Regards

First You need to check which all server jobs need to be converted to parallel jobs. Those jobs should be converted to parallel which are really slow and those in which there are large number of updates/inserts

Regards

Surely there could be a locking problem either with at Table level or Row level in the target Database. Please Check whether there is this type of locking when the sequencer is waiting.

Regards

Sure Ray!!!

I would post the solution as soon i as solve this.

Regards

Its fine. I got a hunch of how to solve this problem
Thanks a lot for the help

Regards

Ray, I am using the Hash file for lookup (not for writing into it). Hence if there are multiple matching records it would take the first matching record. If i use a UV stage for lookup then i am able to transfer more than one matching record to the target. But the issue is that the UV stage is slow....

Its a server job.
We have to remove duplicates from the source query. Is there a way in which the source stage does not process duplicate records(by using keys in the source stage) instead of using a hashed file with stage variables.

Regards

The General rule of the thumb is - Two processing nodes per cpu

Regards

The parallel jobs are slow in compilation but fast on execution.
I am also facing this issue.

Regards

Modify this Sample Rountine $INCLUDE DSINCLUDE JOBCONTROL.H $INCLUDE DSINCLUDE DSJ_XFUNCS.H ErrCode = 0 ;* set this to non-zero to stop the stage/job hJob1 = '' hJob1 = DSAttachJob(Arg1, DSJ.ERRFATAL) If NOT(hJob1) Then Call DSLogFatal("Job Attach Failed:" : Arg1, "JobControl") A...

Thanks Sanjay.
Creating a seperate job for these works but i wonder why it does not
work in a single job since datastage processes records one after another
(In server jobs)

Regards