DSXchange

Why not just select user_code column without the CASE construct and handle the non-KSKS values within the job?

Perhaps you can use some protocol like Samba to mount the disks from the UNIX machine on yours.

If you're using a CFF stage you need to have imported from a COBOL file definition. If you're importing a Sequential File definition, you need to use a Sequential File stage.

None, because you have the records, and U (another of our posters) doesn't.

Close a few windows, so as to reclaim some memory.

You would use the IsValid() function in a Transformer stage that has two outputs, one for times when the record is in some way invalid. This would server as your link to your rejects table.

This needs to be narrowed down in terms of definition. Please answer the questions I posed earlier.

Good points. Just remember that the two "unhandled" user overrides are applied after the pattern action language script has been executed.

Why do you want to know how U (one of our posters) can achieve it?

Put the parameter value somewhere such as in a file or in the first job's user status area.

Change comma to pipe on the Format tab when importing.

Are the messages coming from the first or second aggregator in each pair? Are the grouping keys in the second aggregators the same, or a subset, of those in the first? Could there be a problem in the Teradata stage that is being driven by the second aggregator stage? Have you tried enabling inter-pr...

Are the dependencies correctly recorded on the Dependencies tab of every job? Are any jobs started with a call to the DSRunJob() function?

No other messages, such as "row out of order"?
Are you aggregating sorted data?

Give us examples of the logged messages. That may shed some light. The limited information you have provided thus far does not.