DSXchange

recently i have seen a job, in which a flat file will be given a input to a dynamic file write stage.......flat file willl be holding xml data......

can anybody tell me how dynamic file write stage works? and why it is used..>?
i didnt find any content about it anywhere

That's because there's no such thing.

Well how can you bve so confident?

if you go to the utility section of pallet....you will find that stage....

any idea? how it works?

My Palette, which is stock standard out-of-the-box, has no Utility section. Might I suggest that your Palette has been customized?

Secondly there is no stage type of this name in the Repository, either in DS_STAGETYPES (the local repository) or - in version 8 - in the unified metadata repository for Information Server. Might I further suggest therefore that someone at your site has created a custom stage type (it might be a Build stage, a Custom stage or a Wrapped stage)?

Finally, as to how I can be so certain, put that down to 12 years intense experience with the product, being properly certified (and a preferred trainer to IBM), and checking my facts. Might I finally suggest that you compare my post count on DSXchange with your own?

Hahahaha...you smashing like anything.....i know you are genius.....but when i saw such a thing in my designer i got little bit confused.....So i raised my eye brow.....

Can you tell me how can we create such a custom stages.....

That's neither a short nor simple subject. Why not track down whomever wrote that one of yours and have them tutor you?

There's a section in the Parallel Job Advanced Developer's Guide on writing custom stages. How good are your C++ coding skills?

Hi Ray,

Thanks for the valuable guidance.....also one more thing i would like to know .....do we need to write C++ code in per-record tab only.......
also how to test and debug the rquired code is woking coreect or not...?

Weren't we talking about an already existing stage here?

Code only on the per-record tab? A definite maybe, though I would be very surprised if that were the case. Typically you open the file, declare things and allocate memory in the "before" tab and close the file, clean up and free memory in the "after" tab.

You make your own arrangements about debugging. Maybe you test the code completely externally to DataStage (apart from the transfer, of course), then test in place.