DSXchange

Having just had the first cup of enterprise....

We always did server job lookups using hashed files. We found that lookups from ODBC stages seemed to do full table scans - or take much longer.

What is the best practise for lookups in enterprise? We have varying volumes - from less than 100 rows to billing data to financial data to call data records.

I see some of my colleagues use ODBC connectors through a lookup stage, but we have been asked to hit source system tables only once. In addition, some of these tables may still need to be used in server job lookups.

Sooooo, any ideas for best practices for our environment?

If you just create a default lookup the way you would have frmo a server job, the lookup will be loaded into the DataStage job and be run from memory, you have to explicitly code a "sparse" lookup so that it does singleton queries the way server jobs used to.

Thanks. I thunk it through and tiggered something out...

Best practise for doing a lookup into a lookup into a lookup in enterprise?