DSXchange

Good idea but that again is a unix solution Just what kind of solution are you looking for? :? Anything run on the server side will be a 'UNIX solution' by definition, are you looking for something on the client / Windows side for this? You have to leave the GUI to automate anything like this and t...

You only? Everyone? One particular job? All jobs?

You could always try searching the forums for "Converting a nullable source to a non-nullable result", I mean it's not like no-one has ever asked this same question before.

[channelling Ray]

It's a Sequence job (no 'r') - a Sequencer is an object used in a Sequence job to join multiple flows into one.

[/channelling Ray]

There's also a decent discussion / tutorial in one of the Server job pdfs, from what I recall.

Look at the known Sequence job(s) for this job and make sure the job is included in the Dependancies tab of the Sequence.

Pretty much just like you'd think - source from the hashed file, write to your database. Nothing tricky.

Did you check at the O/S level for any processes associated with the job? Could it be the job was killed/died so hadn't reset the Status? If so, you'd need to use the Clear Status File option in the Director to 'fix' that.

Now you just need to let us know how your utility comes out.

'Breaking from the loop' is easily done - it's just a matter of branching out of the loop to a point after the End Loop stage based on your 'custom condition'.

Backwards Link.

Or perhaps DB2 tables? I would assume you'll need to import these using the appropriate DB2 utility, but perhaps that can be into a set of work tables, then use DataStage to clean/transform from there.

Yes, Version Control is still alive and well in 7.5.3.

7.5.3 is just a 'maintenance' release and so is mostly just a collection of patches. Nothing changes with the repository or Version Control... you'd need to actually upgrade to 8.x to feel those pains. :wink: I would guess that Vincent meant the patch level of the Server engine in 8.0.1 is equivalen...

That was why I was asking, to narrow this down to either the definition of the ODBC entry or to something specific to the job / query. I would triple-check the former before investigating the latter.