Need benchmarks for LARGE db's on Unix (10's of Gig)

Hi, We are scoping/designing a system's integration project and need numbers that show how Unix performs with a LARGE database. Information on Teradata would also be helpful. The current debate is where to put the database: on the Unix box or on the Mainframe? Obviously a high-end (mainframe equivalent) Unix machine must be involved. I'm interested in benchmarks for large databases - in the range of 10's of gigabytes. If you have worked with any databases of this size on Unix and could provide any of the following, it would be greatly appreciated: Platform: Database Size: Performance: Aids/Tools to tune: Any other information would be helpful as well. ADVthanksANCE. Kevin Dombroski -- ================================================================================ Kevin M. Dombroski | "One good thing about music, Cambridge Technology Partners | When it hits, you feel no pain...." ================================================================================