DSXchange

Welcome Aboard. :D

First suggestion would be - no guessing allowed. Timeout problems throw -14 errors, did you get any of those? What makes you think the first job 'never got triggered'? The Sequence job should log a 'Summary run' message when it ends, have you carefully examined it?

Ok, what is the current Array Size in the job? Apparently, way 'more bigger' than it should have been.

Always best to mention the 'real situation' right off the bat.

Or just ensure the load jobs abort in the event of a problem. No 'audit' required.

Exactly, all that would do is not send rows - however actions like truncate would still happen when the stage is 'opened' and the db logged into. There's no way to "ignore" a link. As noted, you should drop a Sequential File stage on the canvas and direct your link there temporarily. You'll lose som...

Define 'disable'. If that includes not even logging into the database, you'll need to redirect the link to something like a Sequential File stage. Otherwise, setting the Constraint on the link to something that equates to false - like @FALSE for instance - will stop rows from going down it.

Where here? There here? <Young Frankenstein springs to mind>

Did you check out on the Unicode link?

Mostly - one of the stages exists, and one of them doesn't. One type only works on SMP systems.

It's in the documentation, one such place would be under Job Properties.

I can't help with the SQL Server question about LongVarBinary support. For the Oracle side, however, I would definitely avoid LONG or LONG RAW datatypes. They're really something you were stuck with in Oracle7 but since then have been deprecated and one should now use the 'new' LOB fields, introduce...

In particular, it's complaining about the Oracle LongRaw datatype not being supported.

I asked that earlier on. This all seems to be driven by the 'dataset stage' issue and the hope that somehow doing it in C instead of BASIC will rectify the issue. And I suggested the urgency be communicated to their Support peoples, but apparently they were 'helpless' - whatever that ends up meaning...

Welcome aboard. :D You have use of the Field function in PX, yes? Two of those, inline, would do the trick. The first to take the '3rd equal sign delimited field' and then the second to use a space delimiter: Field(Field(Link.ETL_ERR_MSG_TXT,"=",3,1)," ",1,1) = 38...

As noted, for any 'abnormal termination' situation, you need to Reset the aborted job from the Director to try and get more diagnostic information.

Have you tried that? Did the 'Reset' add a 'From previous run...' message to the log? If so, please post the contents of that message.

That was the other reason. Not just the 'Set' versus 'Get' aspect but the fact that parameters are not in play here. And I assume that means either job parameters or having to pass it to the routine as an argument. I thought there was some way to reference it directly? #$Whatever# for example, but p...

DSSetParam?