DSXchange

I don't see how a Sequence would change that behaviour, the only difference it brings are the job parameter values. Are you certain you are still processing the same file in that case?

And? Have you successfully reindexed or is there still a problem?

In short, you edit your uvconfig file and then 'regen'. Why do you think you need to change it?

OK... I can see where you'd have to worry about that if your hashed file was being shared while you were working with it, especially with public caching in the mix. For me, I ensure that the hashed file being used for that kind of operation is unique to the job so there's no surprises. One of my ear...

OK, new information here. We've discussed this technique any number of times in the past and this is the first I recall seeing you mention anything about public caching which I thought was all about hashed file cache 'sharing' between jobs. And there's no mention of a 'public caching' prerequisite i...

Well... I was assuming they meant a duplicate within the job run rather than run to run, as that one should be pretty obvious.

I built a stripped down version of what Chuck put together, something that simply searches through a dsx file for a string and reports both the individual occurances of that string and a summary of the object names that string was found in. It doesn't need to know anything about 'server' versus 'par...

You would get normal ORA errors for duplicate or other non-unique problems and duplicate values in a 'key' field would just update multiple records rather than one. This would be something else. How many columns defined? Any problematic data types like a CLOB? What array size are you using? First th...

DBMS_STATS package? Have you tried just "call DBMS_STATS.gather_table_stats(...)" there? Still better to use the SP stage for that IMHO.

Yes. Must be from the same transformer and the lookup cannot be cached. Well, technically it can be cached as long as you use the 'locked for update' flavor of caching but I prefer simply to not cache it.

No, not possible - PL/SQL is a different animal from straight SQL. You should be able to use the Store Procedure stage as long as you uncheck the 'generate procedure call' option, depending on exactly what your code is trying to accomplish. You may be able to use the ODBC stage for that, don't recal...

Ouch. :shock: If you have a "total backup" done daily on that server, why not just have it restorted in its entirety? Unless you've been doing DataStage 'exports' of the projects that would be the only way to get back everything you've lost. Simply reinstalling DataStage won't buy you much...

Only way to know would be for you to try it... and it certainly won't make it any "worse".

Wait, sorry... I thought you were reading from a named pipe in the Sequential stage. If that's an option in the DB2 stage, then I know nothing about that.

Did you check the documentation for that stage?

:!: If you read the other posts, then you should know that a reindex makes all of the jobs appear to disappear if you didn't have exclusive access to the project. That literally means no jobs, no other client connects, no defunct connects, nothing. If you have to restart DataStage to ensure that, th...