DSXchange

I think it's reasonably safe to infer, because Craig's post was quoted, that Craig's suggestion was the resolution.

What are the precision and scale settings for the Decimal target column? Does it have a specifically set display width?

Before-job and after-job subroutines are always server routines, even in parallel jobs. You can use the BASIC Transformer stage to call server routines for each row processed, as Arnd noted. You can also use server Shared Containers in parallel jobs, subject to some restrictions that are documented ...

This is a stand-alone "UniVerse" program. It is neither a FUNCTION nor a SUBROUTINE. It may have an optional PROGRAM declaration at its beginning. It requires a STOP statement to end it but the friendly compiler will put that in if the programmer forgets.

Within the code find out how the hashed file is opened. If the Open statement is used, there must be a VOC entry in the same project that points to the hashed file; this can be created using a SETFILE command. A hashed file can be opened via its pathname with an OpenPath statement. If that fails, ch...

On the global security page in the WAS administration console. You manage those users in the web console for Information Server; it will select users and groups from whichever user registry is in use. You don't really have to look in the WAS administration console.

Alternatively you can have an operating system command, such as grep or awk, select the rows for you. You would implement this by means of a Filter command in the Sequential File stage.

You don't add users there; you map users defined in whatever user registry WAS is using to authenticate to operating system user(s) on the engine tier.

Specify how the data are partitioned.

Yes. Use values files.

No idea, because you did not vouchsafe what values you used for these.

What was the value given through DB2INSTANCE environment variable?

Which user ID did you use to run ISALite and was the DB2INSTANCE environment variable set for that user?

So where is the source code for CCYCONV ? This one of yours, not one of IBM's.

So, resolved now? Not too bad a turnaround time to produce a patch.

Can you outline how you are trying to use these functions? They are intended to be used when processing groups of records.

SaveInputRecord() pushes the current record onto a queue and returns the number of records now on the queue.

GetSavedInputRecord() pops one record off the queue.