DSXchange

According to your original post in this thread, DataStage IS running on a UNIX box.

A pair of stage variables, not a function. One stage variable effects the comparison while the second "remembers" the value from the previous row. The technique has been outlined on DSXchange many times.

It's a difficult question. Theoretically all triggers in a Nested Condition will fire very close to simultaneously, so the combination of jobs will start at very close to the same time. You could use a Routine to determine which jobs are running and wait for all of those (and only those) to finish. ...

Try connecting during US business hours. The numbers are often higher than that.

Yet again, it depends. If process 1 is busily updating, then it will be taking at least record level locks that will prevent process 2 from obtaining a table level lock. So how are the deletes being accomplished? Through SQL (which will start with record level locks but will probably end up promotin...

That's a new problem - your original SQL wouldn't handle that either.

You can still do it in a Transformer, by outputting only rows where the key (user_id) is different from that in the previous row.

Please note that Kim's routine does not find jobs invoked from the DSRunJob() routine within a parent job.

Provided the environments are identical and you have migrated all required components, then you can use the compiled code on another machine. Technically the fact that it's a trial version of the compiler should not affect code that you have successfully compiled but there will come a time when your...

At least 50 columns. As many as - or more than - the number of distinct values in the STATE column. Filter out nulls or empty values after the Pivot has occurred.

It is possible but not recommeded due to the danger of establishing a self-deadlock situation in the database.

Nothing's different in DataStage - it's just another client as far as the database is concerned. If you issue two DELETE statements on the same table, the first to start executing will take an exclusive table lock and the second will wait until that lock is released (or a timeout error occurs).

One means the common repository for Information Server, the other means the "UniVerse" local repository for DataStage. What the actual letters mean has never been revealed.

Please read this post.

Talk to your DB2 DBA about why bind20 may have failed. This is a DB2 client command.

Nagaraj wrote:Did you get answer to this query?

Do you see one?