DSXchange

And is that an error? I am not sure what you want.

If you do what is your output?

The Zulu format for xml is a good one. What is your actual error when going to a service?

configuration in DataStage isn't that difficult, all you need to know is your Queue Manager name and Queue name to read and write messages.

Please, don't use all-capitals, that is considered SHOUTING. The maximum value will probably be a 15 digit number (depends upon your EXACTNUMERIC uvconfig file setting). It should be plenty big enough for surrogate key generation. Actually, now that I think about it, it should increment beyond that ...

Can you write a dummy job, reading your 2 values into 2 columns in a sequential file as decimal(12,2), then doing the subtraction in a transform and writing to a sequential file? Is it reproduceable like that?

Use the routine "KeyMgtGetNextValueConcurrent" in both streams or use a database surrogate key (i.e. a database sequence).

After you have executed ". ./dsenv" in your $DSHOME directory, please enter "env | grep LIBPATH" and "env | GREP DSHOME" and "ls -al $DSHOME/lib/libgciudt6.so" then post the results here.

Is this a server job as indicated? Server doesn't distinguish between numeric types, so I can't see how this error happens and a short job I did (v8,UNIX) didn't reproduce your problem.

What column is the primary key defined as in Teradata?

It might be that you specified double-quotes in the attributes and I think that in that place they might not be necessary. I am assuming you specified column metadata.

DB2 does not have a truncate function, at best one can issue a "db2 import from /dev/null of del replace into MyTable".

You could write a server job and put the code that does the actual work into the job control code and keep the job empty; this is similar to how job sequence work. It is hard to give real suggestions since I don't know what you need to do, it might be possible to do this with the norma job sequence ...

You should think about just writing a DataStage BASIC routine to do this for you. If you have your metadata stored in a file the rough logic would be: {Use OPENSEQ to open the sequential file} LOOP UNTIL EndOfFile READSEQ MetaData FROM SequentialFilePtr ELSE EndOfFile {process metadata to fit into p...

My stage variable response also applies when you use the pivotr stage.

Odd, I just tried an exact search for "When checking operator: Dropping" and received 46 matches, many of which describe the same problem and contain an answer.