DSXchange

The dsenv file is an artifact of a UNIX server and yet you've marked your post as being a Windows server. Assuming you are running on UNIX... For Oracle you need to have your $ORACLE_HOME defined in the dsenv file and your SHLIB_PATH (or LD_LIBRARY_PATH depending on your platform) needs to include a...

You cannot call routines from a parallel job because PX does not understand the BASIC language that routines use. I'm afraid this isn't true. You can call routines from the BASIC Transformer as the OP is trying to do. My guess (don't know for sure) is that PX jobs may not have a 'User Status' area ...

ArndW wrote:After such an exhaustive post by Ray

Affectionately known here as a 'Full Wurlod'.

If your DBA can't shed any light on the problem, let us know what the stored procedure is doing. More than likely, there is another way to accomplish what it is trying to do, something more appropriate for a DataStage job.

And btw - where exactly is 'lhklhl'?

Nope... hash lookups are basically an all-or-nothing affair. If the hash was created with three key fields then all three need to be referenced or the results will be... unpredictable. So, unless Dept can be 'hard-coded' or derived from the source data you won't be able to perform the lookup. Sounds...

You can't. Not unless you are willing to totally replace that process with one of your own making.

Something to keep in mind - the failed job will be the last job listed in the email when you tick that option.

I'm trying to design a simple server job which would do the same as this SQL query: SELECT max(MAXDATE) as MAXDATE FROM ( SELECT Max(DATE1) as MAX1 FROM TABLE1 UNION SELECT Max(DATE2) as MAX1 FROM TABLE2 ) where TABLE1 and TABLE1 are located in two different datasources. Define what you mean by dif...

StefL wrote:OK, I didn't realise I had to maintain the Dependencies grid myself, seems like something that definitely should be automatic - or nonexistent...

You do, except for (as you found) some aspects of Sequencer jobs. However, that being said, it isn't used by Version Control.

As Ray noted, the current job is already attached so don't try to attach to it again. Skip that step and use DSJ.ME as the handle instead of what the 'attach' would have gotten you.

At worst case, post your code.

Actually, it seems to read through the jobs themselves to find what it would consider to be 'dependant' components. However, I don't recall that functionality ever working for Sequencer jobs... just Server jobs.

True... of course... that was a classic Brain Fart on my part. Sorry.

I've seen them too. Warnings when you install and warnings in every job you run. Haven't seen where it wouldn't actually install but that wouldn't surprise me.

Are you sure you don't have that backwards, Ray? Or is this just applicable to the EE edition?

PIC 9(8) is COBOL notation for an eight digit numeric field. So, pick one of the numeric data types (Numeric, Decimal, Integer) and set the Length to 8.

How to run Analyze Table statement from a DataStage job? You won't be able to just do this easily, which is what the other posters are pointing out. However, if what you want to do is analyze a table after you're done loading it, that is simple - just put the full 'analyze' command in the After Sql...

kollurianu wrote:By the way i am working on DS 7.5 sever Edition. Hope the options given by would be ok for this version.

That just means you will have to take the DS.TOOLS option as the 'Cleanup Project' button was removed from the Administrator in your version.