DSXchange

Ah, sorry, I see the confusion. Let me find a little time to build this locally and get it working, that should help make sure that (a) I'm not lying to you and (b) to get you more better instructions.

There's also the dssearch command but that relies on the metadata in the Sequence job, specifically the information on the Dependencies tab that isn't always 100%.

Pretty sure Kim Duke has posted that here somewhere...

You can't do that. You need to loop through all jobs and then check their status individually.

So... dates like 03-MAR for March 3rd? Kind of hard to fully validate them without the year but I guess you could hard-code on a leap year and then validate. Unless 29-FEB is never valid in your world.

The 'DRS' stage is just that - dynamic, meaning it can be switched to use different drivers and thus access different databases natively or via ODBC. The other stages 'just' do what they're named for.

Search here for SETFILE which will establish the VOC record you need.

Technically, the Execute Command stage. That or Windows servers have a Command stage as well, from what I recall. Not really sure of the difference, though.

ray.wurlod wrote:When I first saw Rocket Software in the U2 announcement, I immediatly thought of Bullwinkle J Moose (aka Mr Know-It-All) for some reason.

Why, because of Rocket J Squirrel, of course!

As long as it is an account-based hashed file or you've created a VOC record for a path-based one, you can use SELECT or LIST on them from there. I still prefer View Data from the stage, however.

Why would you think I would suggest something you couldn't do?

Add the year column to the lookup as a Key and work it into your SQL as noted.

Can't check myself, but it will have the word KEY (MAXKEYSIZE?) in it and be set to 768 unless it has been changed. Really shouldn't be all that hard to find.

Have you searched? This has been discussed a fair amount here. Also, did you check the ulimit from inside a running job? Checking from the command line doesn't get you the proper information.

Sure. They would need to be completely separate installs with different "itags" and you'd also need to install both clients into separate locations, using the Multi-Client Manager to switch back-and-forth between them. There's a pdf document that explains all that in your Docs directory, i...

Welcome. While normally you won't have to do anything special, there are some cases where bad things in a lower level activity don't get picked up as bad in the calling job and thus go undetected. Seems that this BCP issue is one of those cases. For this you should probably consider putting together...