DSXchange

It's actually an excellent idea, because one process "warms the cache" for the others. :D

The Reporting Assistant has been deprecated in favour of MetaStage.

The Reporting Assistant can not report on job sequences or on parallel jobs.

The only alternative is MAT, but this must assign the same value to every element, for example Otherwise you have discovered the correct answer. Well done.

But does it get a null sent to it because you're using default Format, and there's a "" in the source file?

Main sources of error codes are the hashed file SYS.MESSAGE (for the Engine) and the PDF manuals (for example for InterCall, the underlying architecture used by connected clients). I am fairly confident that there is no consolidated list of error codes generated by DataStage itself, as the OP reques...

If you're on Windows XP sp2, this error can also occur if the network card is inactive. Make sure that you're connected to a LAN of some kind. It's weird, the other machine in the network doesn't even have to be on, but the cable must be plugged in. It may even be enough to be plugged in to an activ...

... and a DSDetachJob to indicate that you're finished with it and another process doesn't get a "locked" message.

Would you care to share the actual expressions you used, to save someone else this research task when they search the Forum?

The "date picture" string is taken literally. Same as it is in Oracle. The moral is that you make sure that your date strings exactly match the date picture.

The Lookup stage supports one or more reference inputs. More than one of these can read from the same source (virtual Data Set). The only exception is a sparse lookup, in which case the Lookup stage can have only one reference input link. However, you can have more than one Lookup stage in a job.

Strictly speaking, it doesn't. At version 6, when IBM bought UniVerse (and a heap of other database products) from Informix, Ascential had to change the name of the database to "DataStage Engine". However, the storage mechanism in the Repository is unchanged through to version 7.5.

dsjob -jobstatus is documented as returning the job's status code. So you could use dsjob -run -jobstatus project jobname ; echo $? to get the exit status into a shell script. And, of course, the -jobinfo option returns a heap of stuff out of which you can parse the job status. I appreciate that the...

Is this really a DataStage TX question?

That may not be enough. If there are data in the hashed file, you will have an empty index. Command="CREATE.INDEX " : vHashFile : " " : vIndexCol Call DSExecute('UV', Command, Output, SystemReturnCode) Command="BUILD.INDEX " : vHashFile : " " : vIndexC...

It's the job ScrubClm that's not in a runnable state, not the job sequence. This information is in the error message. The job ScrubClm may be aborted, or not compiled, or running, or stopped, or crashed. Any of these status values mean that the job is not in a runnable state. That's why job sequence...