DSXchange

In PX 7.0 you can use a BASIC Transformer stage (not the same as a Transformer stage) and, from this, make use of routines written in BASIC. Otherwise BASIC routines are not accessible from parallel jobs. They are, of course, accessible from job sequencers and job control code, though you may need D...

You used the wrong stage type.
In PX 7.x there is a Transformer stage, and there is a BASIC Transformer stage.
Use the BASIC Transformer stage. It's the one you're used to from server jobs.

In short, you use the output link constraint from a Transformer stage to identify those rows that will be sent on for further processing (in your case, updating rows in a table). There is no limit on what you can do (except that, if it's too complex, you may need to write a routine; there is an uppe...

Don't do it, is my advice.

You're re-inventing the wheel.

You can call stored procedures with the appropriate Oracle stage. Do that.

The processes are in parallel, unless they are started in a fashion that makes one depend upon another. This is completely handled by the operating system. Each is give a "time slice" - a short period during which it has use of the CPU, and other processes sleep. A process may relinquish its time sl...

Since you're on 7.x, why not insert a BASIC Transformer stage? Use it to simply pass through any existing data, but to generate a new column containing @LOGNAME for every row? Or it could have a separate output link that only executes for row #1, and sets the value of @LOGNAME into, for example, the...

If you can detect the wrong data in the last line, why not simply have your job ignore that line? For example, change the "missing columns" rule to pad with some data that don't exist elsewhere in the file, then have the Transformer stage output constraint expression filter that pad data out. Much e...

I don't know why this doesn't work? I want to do some pattern matching on the ride side, both sides etc. I want my function argument to compare to what is in my table in my routine. EQMASKEntries=4 Dim EQMASKTbl(EQMASKEntries) Mat EQMASKTbl="" EQMASKTbl(1)="'ATT'..." EQMASKTbl(2)="'GAR'..." EQMASKT...

Ray can you post a working example, I can’t even get my code to compile and when I did the match didn’t work. The documentation is no help to me! My examples follow: Thanks, Larry EXAMPLE1 EQMASKEntries=1 Dim EQMASKTbl(EQMASKEntries) Mat EQMASKTbl="" Ans=@False EQMASKTbl(1)="...'GAR'......

Could still be worth checking the file dictionary to see whether such a column already exists! (Or ask the third-party vendor.) If it's there, your problems are non-existent.

You can inspect that code if you wish. Step 1. Determine the job number. It's probably a part of the error message - or you can SELECT NAME, JOBNO FROM DS_JOBS WHERE NAME = 'jobname'; Step 2. Find the directory called RT_BPnn in your project, where nn is the job number. Step 3. Review the file JOB.1...

I'm with Ken on this one. Let's go one step further. There's nothing to stop me hacking the code produced by a Transformer stage, because I know how to get it compiled without recompiling the whole job. But would I do that, knowing that the next developer, lacking that technical knowhow, would recom...

Windows can be pernickety about your user identification. Did you try a qualified user name, for example machine\user or domain\user ?

There is one other possibility, and that is that a required column is indeed missing. In the Columns grid of the Sequential File stage, scroll to the right and you can change the rules about what happens when a column is missing, for example from aborting the job to substituting a pad character (may...

As for Ray's post, would them tmp file that the records whose id's have been changed be a subfile of the one whose records have been changed? I'm still reading the docs I downloaded the other day. My idea was that there would be something in the records themselves that would timestamp their most re...