DSXchange

Imports from other versions of DataStage work just fine - as long as the version being imported is < the version being imported into, i.e. is older. The import process will automatically upgrade your jobs to the current version. What version are you importing from? Other things off the top of my hea...

No, not that I am aware of... as least not directly. You may be able to access the hash via a UV stage and then do 'normal' SQL against the data. You'd need to either have the hash in your Project or create a VOC record for it (if pathed) for the UV stage to work.

I want to duplicate the 44 columns 6 times with each row containing one of the remaining 6 columns. Then, as Vincent noted, you can use the Pivot stage. Bring all of your columns in on the Input side, and then for the Output side list the first 44 columns. In the 45th column, list your 'remaining 6...

Quite possibly! The OCI stage really wants all DATE fields to be declared as Timestamps in DataStage and then you must supply the date to it in 'YYYY-MM-DD HH24:MI:SS' format. So your conversions could very well be happening just fine and then the 'problem' is on the target OCI side. I'd suggest dec...

As I see it I have two options - we are running on Unix AIX so I can either use a flat-file stage and output the files to our Unix box then call a script to perform the actions in Unix or alternatively use an FTP stage to write the data out to a Windows box and then write a VB script to perform the...

Any idea what to do with all these points? Do you think I finally have enough for that pony?

Ah! A blast from the past. Actually, did get an answer to this... just not here, but over on ADN. From the post:

LenG wrote:Roughly: every 2 seconds for the first 2 minutes, then every 4 seconds for the next 4 minutes, then every 8 seconds thereafter.

There you go!

Some random points... You may not have the docs for the Row Splitter/Merger stages, as some of the early disks left them out. Vincent and I got them uploaded to Ascential's Developer Net some time ago and they are here if you have access to their file library. Not really knowing the format or either...

Welcome to our happening forum, Sam! You've got a couple of primary choices to do what you need to do. Some of these answers may need to be tempered by the exact nature of your output rows - are they all of the same format / metadata or will they have different metadata? Your target - database, hash...

That and as another example, the ever popular substring operator - the square brackets.

For example:

MRPG2001[1,3] = MRP
MRPG2001[4,5] = G2001

Best to read up as Vince suggested, so that you know about all of the Weapons of Mass Deconstruction at your disposal.

It's not a query problem, it's a mismatch between the columns declared in the query and the columns defined in your stage. You probably have more columns in the stage than in your custom sql...

If your dsrcpd daemon is refusing to start, you need to take an extra step to find out why. Search the forum for 'dsrpcd' (or 'uvrpcd', either one should yield the same advice) and 'debug' as you need to start it up in debug mode for it to tell you what's wrong. There are quite a few posts here on t...

Same thing I was going to recommend.

Either a script to do the renaming after job, or if you need to deliver the file via FTP do the 'renaming' as part of the file transfer.

Not sure with ODBC, being more familiar with OCI. In the OCI stage, one connection equals one transaction. I personally would never use a Transaction Size of 1 (given a choice), unless you also build in a mechanism to 'pick up where you left off' in the event of a problem. Better, in my opinion, to ...

Technically, yes. You would need to contine to use the 8i client and the OCI8 stage, but you should be able to work with a 9i database that way.