DSXchange

Hi
I tried searching through the forum for accessing Universe tables or querying existing hash files. I even went through the manual on Universe and Unidata. I could successfully get the count of the hash files using the Admin command interface because of the various posts on it. I want to know if the DSN is the same for accessing Universe through the Universe Stage. Any help is greatly appreciated.
Mack
____________
Learn till U Live

What is the release of Universe you are attempting to access? You show DS on Windows, is the Universe server on Unix? Have you configured the ODBC.ini to have the appropriate entries?

How did you succesfully get the count of the "hash files?" What hash files are you talking about?

Are you attempting to point DataStage at its own internal repository? If so, have you read on the forum where we've talked about DS masking out its own files from the Import metadata facility in DS Manager?

Please let us know what you are trying to do, so that we can give a more focused answer.

I have the client and server both on Win XP, I got the count of the hash file by creating a VOC entry by using SETFILE and getting the count through the Admin Command window. The same is true for getting the info through the UV shell. I could also query the hash file successfully. I wanted to get the same thing through the universe stage in the designer and didnt knew what the DSN values should be and If needed to create a new DSN other than the one I created for SQL.
Thanks
Mack

use the default dsn "localuv" when using the UV/ODBC stage to conduct relational type accesses to hash files created by DS. Make sure that you have used the SETFILE command to register these hash files into the "VOC".

Keep in mind that you are simply strapping a SQL interface over what is basically a sequentially accessed file. Unless you put on a Universe DBA hat and optimize the hash file via secondary indices, almost every query you would do against that hash file will be a full table scan. If you are simply doing SQL aggregation, this won't matter. But, if you're doing any kind of ranged lookup, multi-row reference lookup, etc, then you incur full scans for each lookup.

Thanks Ken, will try that :D
______________
Mack