DSXchange

There is only ever one copy of a multi-instance job. There may be log and status entries from various invocations, but these are just entries in tables. Log entries are removed during the normal purging process. Status entries are kept until the job is re-compiled. But there are never multiple copie...

Create an External Filter stage that runs oerr with the error code passed in.

Yes, but if you delete a category you must also delete the objects contained in it.

If you do this in the client, a warning message pops up indicating that this will happen. If you do it from the command line you're on your own.

Not in a routine.

For preference use SQL.

If you must use a DataStage job, use a selection filter (or a constraint in a Transformer stage) and aggregate the sum either in a stage variable or in an Aggregator stage downstream of the Transformer stage.

There is no such thing as a DataStage user. DataStage relies upon operating system authentication.

You can edit DataStage roles on the server by editing the .*.adm hidden files in a project. Each contains a list of UNIX group names.

In my experience the machine shutdown shuts down DataStage quite cleanly without having to change anything.

You don't. You write SQL to do it.

Use a WHERE clause to restrict your date range. Use a SUM function in the SELECT clause.

Hah!! MS Security Patches interfere with EVERYTHING in my experience, even with themselves. There are other possibilities. Has this problem recurred? What you see in Directory is not the actual status, it's the most recently recorded status. The job updates its status periodically. If something prev...

Given that this is the 8118th job you've created, and there are six hashed files created per job, you may be hitting an operating system limit "number of subdirectories in a directory". This is 32765 on Solaris. Time to clean up your project, maybe split it into two projects.

There are two techniques for obtaining a parameter value into an after-job subroutine. (1) Put the parameter reference (for example #DirectoryPath#) into the Input Values field - it becomes available in the routine as InputArg. (2) Invoke DSGetParamInfo() function with DSJ.ME as the job handle argum...

Most of the columns in the Repository tables are (deliberately) un-named. This allows the vendor to maintain the mystique and to make hacking the repository more difficult. It is also the case that many of the repository tables contain more than one record type (and therefore record format), which m...

It's a vertical pivot. Search the forum for this term. At least three techniques have been published here.

The Hashed File stage has a Selection tab does it not? An appropriate selection phrase phrase to put in here is
If that builds a WITH phrase that causes a syntax error, trick DataStage into generating a legal phrase, for example

If this error ever recurs, try clearing the job's status file (Tools menu in Director, or DS.TOOLS menu on server). I'm guessing that the status record for the job somehow got "stuck" (there are a few possibilities, which I will not canvas now, since the problem is resolved).

That suggests one of three things.
(1) You are not on the server where DataStage is installed.
(2) You did not install DataStage server successfully.
(3) You did an -itag installation, so that the location is in the DSHOME environment variable.