DSXchange

Please don't multi-post the same question. Answered here:

http://www.dsxchange.com/viewtopic.php?t=112345

It's not a version number issue - the 'Command' stage is only available on Windows servers, not UNIX based ones as it is a Windows specific stage. You don't need it.

No, DataStage 7.x is not supported on Vista. No clue if 8.x is or not...

Nope, they will need to be rewritten.

Not that I recall, Senator.

I just can see it as a distinct possibility, depending on how 'robust' their error interpretation and logging is.

If don't recall ever needing to add a process break-point between two transformers to solve a 'job hang' before... did you try just adding an IPC stage instead? As a tangential discussion, I'd be curious what your full job design is to see if we can get a better idea why doing that would solve your ...

You could try splitting off to a flat file before the hashed file and then another afterwards, then you could compare the two and see who is 'missing'. That would probably go a long way towards figuring out the problem.

If I may, why do you 'need' to 'separate' two transformers?

You haven't and you don't need it. It supplies a UNIX-like environment for Windows servers only and you've got the Real Thing.

I'd suggest you ran out of disk space. Resetting and rerunning the job and having it run fine the second time proves nothing other than the fact that there was enough space at that moment. You'd really need to set up some tight monitoring and be checking the space while the jobs run because I doubt ...

What is your 'work around'?

As noted... you will need an 'extra' stage if you want to both read and write to a hashed file in the same job. And then you would check the status of the lookup to know if the value being looked up existed or not - i.e. if the lookup was successful. There is a 'Link Variable' available for that - L...

No, not really. If you do end up having an issue with your job, even if it seems to be related to this topic, please start a new post. Make sure we also get all the gory details of your requirements and job design as well.

medakiran wrote:Our requirement is to delete hash files and at the same time not loose their tuning information.

Then perhaps you should just clear them rather than delete them?

Fits the mood. Besides, just some Pink Floyd lyrics...

As noted, Rnd() is the random number generator but you really do not want to take that approach if what you are doing is generating surrogate key values. You need to look into leveraging an Oracle sequence object or use one of the two 'Key Management' routines in the sdk category. Double-click to op...