DSXchange

A job parameter reference is the only kind of "variable" allowed in the Table Name field. Therefore you can not achieve what you ask.

Does the job work if you use ODBC rather than OLEDB? ODBC does not need to create any COM objects. Meanwhile, have you reported this - with a reproducible case - to your support provider? It may be that the OLEDB stage does not support so many output links; whether this is a bug or not I can not say...

Something like Rule 1 happens automatically during the blocking phase of your prior MATCH job. The Master (the record with the highest information weight) is marked as XA, the others in its block as DA. There is no simple rule to work across multiple fields, as you choose to do, though there is a si...

It is. But the output link handling the failed lookups is not a reject link. The reject link handles only failures reported by the database.

It can all be done. Of course, you can only boot one operating system at a time, so you will not be able to run your DataStage server on Linux and expect to be able to run your DataStage clients on Windows at the same time. Probably not what you had in mind. Also, your laptop probably doesn't really...

The only version of DataStage for Windows that will run parallel jobs is 7.5x2, at least until version 8.0 is released.

7.5x2 is NOT the same as 7.5.2. Version 7.5.2 will not permit execution of parallel jobs on a Windows platform.

All of which assumes that the OP can actually get microseconds from the system clock. On Windows platforms I suspect not, hence my first post on this thread.

Do you wheel in a third developer to find the majority solution? What if the third developer comes up with a third solution, for example using Perl?

What you actually need to do is to enable the "Add Checkpoints so sequence is restartable on Failure" property in the job sequence's compilation options (on the General tab of job properties).

This piece shows that operators for four stages are being combined into one process: op9[2p] {(parallel APT_CombinedOperatorController: (APT_LUTProcessOp in ManFeed_lkp) (APT_TransformOperatorImplV0S114_Rsn321_Test_Transformer_114 in Transformer_114) ( _SurKeyGen) ( Ins_flt) ) on nodes ( sps111[op9,...

Try a reject link without any constraint expression. Use two Transformer stages if you want to segregate the two types of reject. Check the output link execution order to make sure that the reject link is executed last.

The Sequential File stage does have append and overwrite as write methods. But the question remains: how are you parsing the results transmitted by the Folder stage into separate files?

Do you have any external (ActiveX) functions defined and being used in this job?

Does the job work if you use ODBC rather than OLEDB?

7.5x2 is the ONLY version where you can run parallel jobs on Windows, until version 8.0 is released.

You should get "duplicate" when the value of Key is the same as in the previous row. And, with the logic you've stated (ignoring the typo for Else) you should get "NO" when the value of Key is different from that in the previous row. If this isn't happening, check your work, particularly your stage ...