I read in different posts that for better peerformance of cube processing when we are using Oracle as data source, One should use FetchSize = 496 or 2000 in extended proerties.
I configured my data source with FetchSize as both 496 & 2000 (in back to back processing). But this not result in more improvement in performance.
Am I missing out on any thing to have this working properly and processing time is reduced considerably?
madhur_ms