DSXchange

Use a Sequencer set to "All". This will not fire its output until all the activities upstream of it have completed. If necessary you can have an Execute Command or Routine activity downstream of the Sequencer to execute a sleep or nap command.

Funnel is best characterised as performing a UNION ALL, rather than any kind of join. Assuming two inputs, with M rows on one and N rows on the other, you will get (M+N) rows out of a Funnel stage and as many as (M*N) rows out of a full outer join.

You can use DB2 functions in DataStage in the same way you use it in any other client - within a SELECT statement. This implies that the stage type is a DB2 or ODBC stage of some kind.

There is no "sparse feature" in server jobs.

Difference stage, Compare stage, Change Capture stage?

Replace them in an export DSX then re-import.

Get the SQL out of your Oracle stage (it will be in the log) and inspect it to figure out what may have caused the -1036 error code. Without the SQL we can't provide any more cogent advice than that.

Are the header and footer identifiable by anything in their data structure, such as a "record type" field? If so, constrain on that.

Mod(sv,3) can never be 3.

Use DateFromDaysSince() function.

My bad.

Should have been &h0D

So? Just design it in. This is a great opportunity to create a generic Shared Container for error trapping and reporting. Enrich the error information (which passes through the container using RCP) with information about the job run. If you want to capture the DataStage job log information could I s...

DOS-style Record Delimiter String gives \x0D\x0A (CR/LF).

A simple CR at the end is a very unusual requirement. Are you totally certain that's what you want?

If so try using hex notation (\x0D).

"DataStage" is the parallel canvas. I'm surprised, though, just to see "licensed". Normally this interface reports processor value units (PVUs) or "quantity". Could this environment have been upgraded from a version 7 system? Do you have an "enterprise license"...

Scroll to the right in the Columns grid until you find the "Incomplete Column" rule. Working up from the final column, change the possibly missing columns' rule to "Replace", which will replace each with NULL. Once you've done the pivot you can filter out the rows that have null,...