DSXchange

nkreddy3 - that is correct.

If you don't have NLS enabled this EBCDIC to ASCII conversion should work. Can you narrow it down to one character in EBCDIC and show the numeric value in EBCDIC and then the value after it is converted to ASCII?

Add a row generator stage with just 1 output. Funnel it with the main data stream so that it arrives after the database stream (I can't remember the exact name of that option right now). In your transform you can then query along the lines of "If this row is @INROWNUM=1 and it contains a special val...

The column is coming in as empty, not NULL so the IsNull() function will never evaluate to true.

The preferred way would be to this via a DataStage job but you could also use the UNIX "orchadmin dump {file.ds}" on both files and "diff" them.

But i am not finding eaxct solution any where, if you have any plz let me know, thnx again. For those of you who don't speak SMS and prefer English: "But I am not finding the exact solution anwhere, if you have any please let me know, Thanks again." Siva512 - What about There is a C-callable API (c...

How did "Vignesh" magically appear instead of the empty value? The text is coming back as empty, not NULL so none of the null handling routines are appropriate.

You can always call a stored procedure, which can execute more than one comand.

Yes, USC2 is almost identical to UTF-16 and if you have characters that are not mapped as you expect then you can always create your own mapping tables.

kavuri wrote:...When I tried to view data, It is throwing some message which is not an error. ..

What is the message (it might be pertinant)?

Is there any message from the MKS install? Does your site have non-standard or complex security?

Bad wording on my part, it isn't a wrapper stage, but putting together a wrapper. You can add a "parallel routine" in the routines section. At v 7 you would do that in the manager, at v8 you can do it right in the designer.

With all due respect Dr. Yeduri, your first suggestion may or may not work but your second is guaranteed to corrupt the job and make it unusable. If you think I am wrong, perhaps you could try deleting the RT_LOGnnn hashed file from one of your jobs and seeing if it will run thereafter. Note, you re...

Have you tried installing MKS separately?

Are you sure you want to uuencode in a transform stage - this would be executed for each and every row in the file. Much better to put this in the output sequential file "filter"