DSXchange

What else do you need to do? There must be more than just calling this routine X times.

Either build a looping Sequence job as noted or create a new routine that can do whatever with multiple files.

Well, you could always flex your C++ muscles to create a more 'parallel' solution.

No such thing as NULL values from a flat file, hence your issue. Rather than use the IsNull() function, check for an empty string instead:

If a.SALARY="" then 999 else a.SALARY

Apparently that's a Java implementation of an "ISO 8601" format and I keep seeing search results calling it a SimpleDateFormat. For example, go here and scroll down to Method Detail and the isoFormat entry. The 'Z' seems to be a timezone offset.

I'm unclear what exactly the "bug" that was allegedly found might be.

What issue would that be? Have you tried this and gotten an error? If you're just asking if one works with the other, I have no clue but it sure seems like something your official support provider could answer for you rather easily.

What is your source for this column - flat file? Database table? Something else entirely? Assuming the first but looking for confirmation.

Ray, 'Named Batches' are a specific ability in Version Control to group objects into logical named (rather than simply numbered and timestamped) groups and are completely unrelated to the 'Batch' jobs you are referring to.

OK... let me take a shot at this from memory. I've been using VC since before it was free and part of the product suite but I have to warn you I really never found a need to use the 'Named Batch' option so haven't played with it for quite some time. 1) There's no mechanism to rename it. All you can ...

I like my answer more better.

Does Oracle Enterprise stage need a primary key on the Oracle table itself to perform an upsert properly? or just a unique index?? Or neither? An upsert requires a unique index to function correctly, be that PK or otherwise. Actually, let me amend this somewhat. Since the first action must fail for...

The latest release of Oracle supports TDE or Transparent Data Encryption. No clue if Teradata has something equivalent, however. But then again, this is a completely different discussion if they think you need to encrypt everthing in the 'ETL pipeline data' rather than just at the source/target.

It would help if you gave us a hint as to what stages you have in your job and what "ODSTBD_PRS_PORTF_STORICO_ST1" is.

Your "value expressions" for both are identical hence the identical result. You need to throw a Field() in there to pull the first field or the second field from the output rather than the entire string.

DataStage is notorious for stripping quotes. You may be able to get it to work by "escaping" them and from what I've seen here, differnet methods work for different folks. So try: \' '' (two singles) ''' (three singles) "'" (double single double) For each one and see if any of th...