DSXchange

There isn't one. You could build a routine that selects a list of job numbers from DS_JOBS and then issues a CLEAR.FILE for each one in a loop. SELECT NAME, CATEGORY, JOBNO FROM DS_JOBS WHERE NAME NOT LIKE '//%' AND NAME NOT LIKE '\\%' ORDER BY CATEGORY, NAME; There are extra columns you can remove ...

If it's not there then it has been purged. You need to be careful when you set up AutoPurge for multi-instance jobs - how do you have it set?

No, it's a runtime option and that environment variable should be one of the 'standard' ones that you add to every job. Then you can override it to enable it at any time should the need arise.

jwiles wrote:If you haven't already, I suggest reading the IBM Redbook on DataStage Parallel Framework Standard Practices

http://www.redbooks.ibm.com/abstracts/S ... SEF&mync=E

So... you have an entry in the Director for a run of the base job with a valid Invocation ID shown and that status shows as 'Compiled'? Not sure that I recall seeing that before, if that's what you are saying. :? Compiling the base job again will clear those invocations from the Director, that or yo...

You 'tried to convert' how, exactly?

Do you have a SQL Server DBA there you could ask? They should be able to explain the inner workings to you.

Also note that if it's anything like Oracle, the APPEND will serialize the output. Not really an issue with zero records but something to check.

Meant just what I said. While you can't run a MI job with the same Invocation ID as one that is already running, you can certainly schedule that or build that into a Sequence job. And then it goes kaboom.

The problem is for one of the instance the job is not running. It shows that instance in "Compiled" state. The original compiled job is in its place. I have no idea what that last sentence means. The "original" job will always show Compiled if it is only ever run with an Invocat...

if I ever need to select with an IN list of character values, then I will still face this problem, so any solution or suggestions would be appreciated. As noted, you should be able to find some mechanism that will work to escape the quotes. Sometimes a backslash works, sometimes just another single...

From what little I've seen of Teradata, you don't want a direct 'port' of your Oracle tables to Teradata tables. While you certainly can, they really should be redesigned to take advantage of the unique aspects of the Teradata architecture. Your Teradata Architect or DBA should be able to provide gu...

So, what is your actual issue here? Forget the whole 'from linux' part, you need to be working in the Designer. You are correct that it opens an editor and that you are expected to know DataStage BASIC in order to code anything there. However, if you are opening an existing routine it should not be ...

A Dynamic or Type 30 hashed file is created by default. However, you can certainly override that to create any of the other types available. As to Pathed versus Account-based, that depends on which option you choose: Use account name or Use directory path. And you can see the name in the stage, as i...

Start a new post when you have completely unrelated questions. I've split this one out on its own for you.