DSXchange

That really makes no sense. It should be working fine for you, not really sure what in the heck we are missing. At this point, all I can think to suggest is to open a case with your official Support provider and see if they can shed any light on this. Sorry.

Ah... MinOccurs=0 makes it 'optional' which means you can just not mention it in your call. Not sure that's really a 'work around', I'm thinking that's the honest-to-goodness resolution.

Why the Transformer? That would mean one call per record, is that really what you want to do? Consider doing it 'after job' for a 'once per run' implementation.

What stage? The Execute Command stage in a Sequence job pops to mind as the easiest, but there are other ways.

Dump the Sequential stage and use an External Source stage to capture just the filename, then point the XML Input stage to that URL/File path. Ernie Ostic has a nice little write-up for this on his blog site.

The NOTFOUND Link Variables are known to not work with Database reference lookups. At least I can vouch for the fact that it doesn't work with the OCI stage. So you need to go Old Skool and check a Key field for null to indicate a failed lookup. It is not only using the first column for the lookup i...

samsuf2002 wrote:The Format i want to generate with datatstage ----

Are you SURE you want that format? Those would need to be separate XML files to be valid.

chucksmith wrote:In DataStage Designer, select Tools, then Options. Choose "Default Trigger Colors", then make your trigger colors unique.

First thing I do when I install a new client.

You don't need the quotes. The 'failed to open Project' only happens when it can't authenticate you from what I've seen, hence my request. Note that when you pass in '-user' and '-password' you must also include the '-server' option as well, even if you are currently on that same server. So... $ dsj...

Search the forum, the server function is one Ray wrote.

You can also go to Kim Duke's Tips page and download the XML Best Practices document we've stashed there. It will help greatly with the 'XML in DataStage' questions, but you'll still need to learn XML itself.

Are you sure your LOAD isn't the bottleneck? Replace it with a Sequential stage to check, then tune your source query to be as optimal as possible. Oracle, not DataStage is where you need to solve this 'problem'.

Yes, he does... and that only for 'pathed' hashed files.

I'd wager it was more the &PH& clearing than turning NOTIFY off that solved your problem. As noted, get NOTIFY turned back on!

No. It should create the columns based on your metadata in the job. Decimal gets translated to NUMBER and the Length will be the precision used. I've never seen the behaviour alledged here - define it with one size and it is created with a different size. ??