DSXchange

I would think it could be related to not waiting for the job to complete after issuing the DSRunJob. If it is still running from the invocation made by the first row through the job when the second row comes around, the job will not be in a 'runnable state' - because it is still running.

Or just post a direct link to the thread in question.

Can this be extended beyond 65 secs ? No, valid values are 2 - 65 seconds only. Not sure why, that's just the way it is. Can't say in what way the Director has changed, but I'm sure it has under the covers since the 5.x days. As noted, the best way to speed things up is to minimize the number of it...

Actually, for Oracle - provided it is declared as a Timestamp in Datastage and a DATE field in Oracle - the format should be: " YYYY-MM-DD HH24:MI:SS ". If you are truly using one of the new TIMESTAMP field types in Oracle, then you'll need to use Custom SQL and string fields as the OCI stage doesn'...

Probably cleaner to just include the appropriate header file than to use a DEFFUN like that: $IFNDEF JOBCONTROL.H $INCLUDE DSINCLUDE JOBCONTROL.H $ENDIF Check the online help for the DSGetJobInfo routine and you'll see that it doesn't need the job name but the job handle. You'll need to use DSAttach...

You might want to try again, but make sure the Server comes all the way down. I've got this little quote stashed away from a Ken Bland post: To do a uvregen, get everyone out first, kill all zombies and clients, do an ipcs and then ipcrm to remove leftover shared memory segments, double-check netsta...

Ah - you got me there. I was harkening back to the days when you validated your jobs to create any hash files in them. But since it's not actually a stage in the job...

You're right, 'test' the routine instead via the Manager.

Project is protected in the environment where it gave problem. Is this is the problem? Shouldn't be. However, I haven't done all that much with Protected Projects, preferring to use Version Control and have everything in the non-development areas be Read Only. Out of curiousity, it works now becaus...

Welcome Aboard! :D I'm assuming that you are using Custom SQL in your OCI lookup, not sure how else this particular error is generated. It is complaining about the fact that the Table Name field has not been filled in, which can happen with custom sql. Even though it isn't used for those queries, th...

A word of advice... If this is in your Development or Test Project, be aware that this will also happen the first time the job runs in Production - which can get caught up in any error checking you have and spoil a perfectly good first run. Easiest thing to do is Validate the job before you run it f...

It's what happens when you use the sdk 'key management' routines for the first time in a Project. It has to create the hash file it uses as it doesn't exist yet and for whatever reason, it logs it as a warning.

You won't see it again unless something happens to the DSKSequences hash file.

Can DataStage access two Oracle homes simultaneously? Nope. Or, can we use the OCI8 and OCI9 stages within the same project? Perhaps using the 8i client? Same project, yes. Same job, no. However, I've used an 8i client to access both 8i and 9i databases via the OCI8 plugin without a problem in the ...

That's true... without proper join criteria, you can still select but you'll get a cartesian product - and the table join is missing from the current query. The question as to why it added that particular where clause should be because it thinks it is doing a reference lookup, so it needs bind varia...

As for #1, are you sure those two tables exist in the instance you are connecting to? Not sure why else you'd get that error.

2. It's probably (accidentally) defined as a Reference Link and those two fields are marked as Keys. You got a dotted line from the OCI stage to the Transformer?