DSXchange

If num_rec is your join key you will NEVER see 000129 in a left outer join. In a left outer join you will only see matching key values (000123) or NULL.
This is the definition of a left outer join.

Have you checked for internal corruption? Try a query against RT_CONFIG1000.

You could have asked someone else at your site (if it's WF) - there's a lot of experience there.

Create the number of columns that will accommodate the maximum possible number of delimited values and set them to null if missing.

Just set your Lookup Failed rule to Continue, propagate column A to the output as well, and check downstream in a Transformer stage whether C is null - if so use A otherwise use C.

Youse guys have forgotten!!! The kind of batch scottopizza is talking about is the simple job sequence that can be (and always has been able to be) created in Director. It runs like a regular job, but has a funny name (its name begins with "Batch::"). These are regular job control routines but are a...

Are you using the same user ID/password in the authentication fields in the DRS? Are you the owner of the said tables? (If not, try qualifying the table with the owner name in the DRS stage.)

I read the problem as the more general one where, if the second column is 1 you need a lookup WHERE KEYVALUE LIKE '?%' but, if the second column is 2 you need a lookup WHERE KEYVALUE LIKE '%?%' - you can not do the both in the same Transformer stage. You also can't solve it using Hashed File stages.

Try a Search for this particular error message (leave out the number). Curiously you are not the first to have encountered it.

Both temporary directory and scratch space should ALWAYS be local disk - just think for a moment how they're used. Resource disk should also be local disk wherever possible.

Umm... no. My bad.

I don't think this is possible from the client side.

They were going to use 1313 as the error code, but that number had already been taken for The Munsters' address.

:D Well done for solving the problem.

Can you now mark the thread as Resolved?

Two Transformer stages, with the second lookup being conditional upon success of the first (a constraint in the first Transformer stage). Use other output links if needed to capture rows that fail either lookup.

Craig has almost certainly identified the problem, without saying so. I believe that the TWS process has not executed the DataStage dsenv script and therefore does not have the appropriate environment variables set. Once you post the actual error messages we will be on firmer ground with our diagnos...