DSXchange

A job handle is opened through a call to the DSAttachJob(<jobname>,<errormode>) function.

A transform is a 1-line function that is actually replaced in the job, a routine is a separate function call that is compiled and stored elsewhere. If you change a transform then you need to recompile any jobs that use it. If you change a routine you only need recompile it once and the new version w...

Take your incoming data stream and sort it by ascending identifier and descending occurrence. Process the sorted file through a transform with a stage variable call LastIdentifier holding the last identifier. In the constraint put in a "In.IDENTIFIER<>LastIdentifier" and output that row. You don't r...

Ray,

do you wish to limit it to only those who have taken that particular class?

It seems that something timed out, either the TCP/IP connection or the database commit. 1. What is your commit frequency? 2. Is the problem reproduceable either with number of rows processed within the same job or with a similar runtime for the same job? This might help to see if it is the DB or not.

Since using Hashed files and UniVerse are identical, there is no need for you to define ODBC connectivity, just point your jobs at the the hashed files you want and you can access them.

Neo,

It sucks, doesn't it?

I once had a Pascal Professor who allowed us to use GOTO statements; but said we would need to justify their use in comments. I almost never use GOTO statements and structured coding, even in BASIC routines, usually allows you to dispense with GOTOs.

The best document is Parallel Job Advanced User Guide, chapter "Specifying your own Parallel Stage"

RStone, DataStage can connect from a UNIX machine to DB/2 tables on the mainframe both from Server and PX. If one can set up a DB/2 connect then one can get the data using DataStage. If DB/2 connectivity is not set up then DataStage cannot retrieve data - this is more of a configuration question and...

Neo, I've done that many times - unfortunately when doing a test in the routine it executes immediately and a cntrl-c won't stop it. We just need to be better programmers. Sometimes I've put counters into my programs and put in a 'IF Counter > 10000 THEN STOP' - n.b. that calling DSLogFatal() within...

The definition of "doubt" means "to lack confidence, or disbelieve" in something.

What is your problem with the column import stage in PX? You use it to make several columns out of one column.

The output link execution order can be changed as I suggested.

Go into the Designer and double-click on the transform stage in question to open it. In your window you will see, at the top and beginning at the left a line of little icons. Choose the rightmost.

One of the icons on the top of the transform window in the designer lets you specify the column ordering.

The metadata you are looking for is stored in several hashed files and is generally not directly accessible. Even if you have only the base product you can still use the reporting assistant and the export facility (to an XML or DSX format file) to make this information available. Are you looking for...

Your error message shows that you are accessing UniData, not UniVerse. It is equivalent to write to hashed files (and quicker, as well) if you don't have specific SQL conditions that you need to satisfy.