DSXchange

Not really sure what it means, other than if you search the forums for IPC or 'mutex' / row buffering issues you'll see most advice is to avoid them as they are problematic. Only turn that functionality on when you actually can benefit from it.

Turn off row buffering in the Performance tab of the job.

No spaces around the equal signs, please.

Table... file... doesn't really make a difference in what you need to do. So whatever advice you got in the other thread should still hold.

Guess not...

Then time to mark the post as Resolved, it would seem.

Now it just sounds like you haven't set the stage up properly.

Perhaps it's time to call Support and let them work with you on the last bits? It gets difficult to analyze things like this without access to the actual job(s) in question.

Ah... the Veterans of the Command Line Wars seem to be dying off. Extract the zip file to someplace easy for you to find - and I don't mean your desktop. Someplace like \temp on your C: drive for instance. Then click Start / Run... and then 'cmd' without the quotes. Then, at the DOS prompt: c: cd \t...

[sigh] You missed the code tags. Please use 'Preview' before you post something to check that it will be ok. And we still can't tell which columns are from 'File A' and which are from 'File B'... why do you insist on putting them side by side? Squinting into my crystal ball, it looks like Ray's resp...

What do you plan on doing with this 'number of rows coming out of the aggregator stage'? What comes after this transformer? And define 'is not getting finished'... how far does it get? Where does it seem to get stuck? I get the impression you still don't understand the implications of what you are d...

Or put all that in the source query - if there is a source query, that is.

Your 'before job' routine InitialHashCreation doesn't seem to exist. Check to see why. And why you would be calling something before job to create a hashed file is another question - what exactly is it doing for you?

Yes, somewhat. You can use CLEAR.FILE on a job's log as long as you know the job's internal number:


Caveat: This removes everything from the log, including the auto-purge settings. You'll need to re-establish those once you've done this.

That's odd. Is this a new problem, perhaps in a new project or something that used to work that stopped suddenly? If you go to the Manager and check the Routines / sdk / Utility branch (from memory) does the routine show there?

Welcome aboard! :D

More details, please. Is this a Server job or a Sequence job? Where / how are you using this sdk routine in the job? If you can't find it, perhaps a custom routine is calling it.