DSXchange

I don't see why not, as all you're going to find under the covers are POOTs - Plain Old Oracle Tables. Should just be a matter of retrieving the metadata and knowing how the tables map / relate to each other. Etc. Just like any other Oracle-based application.

That "you" was a generic you, not the personal pronoun - what changes were made to your DataStage server? By anyone. Of any kind. As Ray would say, the answer obviously is not 'none'.

Damn! Was thinking about this and hoping to beat Ray on this historical fact, but... too slow. Yes, I remember being delighted around the 4.x release to find I no longer had to Validate every dang hashed-file-writing job in every environment to create the damn things. That or transfer them from the ...

If this option is not specified and the file doesn't exist, then you will get an error. Sorry, but this isn't true. One will always be created when needed, it's just that by not checking that option they are created with all default values. As noted, enable that option to override the defaults duri...

Ah... ok. My advice earlier on was to forget the dot but if for some reason it has got to be in there somewhere, that's two somewheres.

Why?

Nothing wrong with simply a full path to the script:


Or were you going for something like this?

So, Cherry - still curious, does this indeed happen for others as well, or just you?

OK, I'll bite. No clue, but if this only happens to you and no others, then a complete removal and reinstall of the client is in order. IMHO.

The INTERVAL datatype is not directly supported, but I'm not sure there's really a need to. Typically they are handled as strings, so a varchar should be perfectly satisfactory. Or a numeric datatype if all you want is an individual component of the interval. Can't speak to DB2 but seeing as how the...

What is your script doing? Is there a significance to the return code of '1'? As noted, the convention is a zero to indicate success and a non-zero return to indicate a problem. Typically. Only you can tell us what 'impact' the change of exit status would mean, seeing as how we have no clue what you...

And the dot. If you still have problems, post the actual error(s) - simply saying it "got aborted" doesn't really help us help you.

Pre-qualify updates v. inserts with a hashed lookup. Split your stream based on the results so you have two links to the OCI stage, one for inserts and the other for updates. Then you can do whatever you need, column wise. Also include hashed file checks for parent relationships to ensure you won't ...

Thanks. I thought I'd heard least year at IOD that that functionality was coming in 8.1... guess that's still in the pipeline. Or I'm misremembering. Or both.

It's not being ignored. I for one would never rely on the NLS_DATE_FORMAT like you are trying to do here. Using the TO_DATE() function will always work regardless: select * from clarity.or_log where clarity.or_log.surgery_date > to_date('25-OCT-08','DD-MON-YY') Or better yet (IMHO) a fully q...

Thought I might start things off and other attendees could add questions that they didn't ask during the webinar. On the topic of 'New/Enhanced stages', you mentioned the Pivot stage and the fact that it is now truly parallel. I was curious if they had added support for both 'directions' yet - verti...