DSXchange

Assuming that JOBS_NAME exists (otherwise the OPEN statement would have taken its ELSE clause): Use of the default file variable (which you have done with the DS_JOBS file) is not recommended. The problem, however, is that your code never assigns a value to the JobName variable, which you are trying...

It looks like a change has been wrought in the engine. Try the following UniVerse command when the job log is not otherwise in use: RESIZE RT_LOGnn * * * If it's a really large log file, and the file system your project is on doesn't have much free space, add a USING clause: RESIZE RT_LOGnn * * * US...

The BCI functions within DataStage can communicate between servers running different releases (which is what I believe the OP is seeking).

Ray Wurlod
Education and Consulting Services
ABN 57 092 448 518

One of the reasons UniVerse was originally chosen as the DataStage engine was that it does not need to have the same structure for every record in a table. RT_STATUSnn takes advantage of this, having different record structures for jobs' records, stages' records and "resource" records. The name of t...

Just a thought - you may have seen it on www.datastagexchange.com; why not go there and do a search for "convention"?

Ray Wurlod
Education and Consulting Services
ABN 57 092 448 518

My post may have been misleading. There is a field in RT_CONFIGnn that records the fact that the job is multi-instance-capable, not the actual invocation ID. That would not make sense; RT_CONFIGnn contains information about the run time configuration. I would expect to find information about the inv...

The only way that you can use SQL in a Routine is to use DSExecute (if it's SQL against the tables in the DataStage repository) or to use the BCI functions (if it's SQL against a database accessed through an ODBC protocol). Derivations in DataStage are done with BASIC expressions (in server jobs), o...

Theoretically, where "expression" is used, it means "any syntactically valid DataStage expression that uses values that are in context". Because DataStage is data-type agnostic, the result of any expression can be regarded as being true or not true in a Boolean context; values 0 and "" are regarded ...

1. The Aggregator stage (in server jobs) has Last as a set function. 2. Use the system variable @OUTROWNUM. 3. (a) count them while you're processing the file (perform a Last on the column being generated using @OUTROWNUM) (b) initialize stage variable by invoking a routine that counts the rows (for...

In PX you can use a Column Generator stage. I've posted a technique to handle the partitioning of data; check the archives.
In server jobs you can use the @OUTROWNUM system variable.

Ray Wurlod
Education and Consulting Services
ABN 57 092 448 518

A good way to think of it is that your DataStage Engine is just another client as far as the Oracle database server is concerned. So it needs to have whatever client software, environment variables, and so on that any other Oracle client would require on that machine. If you want to work via ODBC yo...

If you read the message that pops up when you press the Stop button, you will note that it sends a stop request to the server. Whatever is happening has to know how to check whether this has happened (it's actually an update to a record in the RT_STATUSnn table for the job). That would be the perfec...

If you can't get that to work, switch to the (old fashioned) grid style version of the GUI. To do this, select the ORAOCI8 stage, right mouse click and choose Grid Style from the pop-up menu. You can get closer to the user-defined SQL and more easily use parameters. As to the link names, note the li...

Enrol in the DataStage Essentials class. This will equip you with the knowledge you seek. Essentially, you can work backwards from how you'd do it in SQL. UPDATE table SET column = value WHERE key = keyvalue In DataStage all you need to do is create a job that extracts data from source, generate the...

The operating system does not allow two processes to write to the same text file at the same time. (This is nothing to do with DataStage - it's an artifact of your using the same file name on more than one output stage.) To correct, write to two separate files and - if necessary - merge them subsequ...