Here's another question: Are all the disks on the new array (known to be) working correctly? Is there any chance that one (or more) of the RAID-5 groups is operating one disk short of a full set? On Wed, Feb 27, 2008 at 11:53 AM, David Barbour <david.barbour1@xxxxxxxxx> wrote: > We recently moved our database to a new SAN. Performance has just > tanked. Here's the environment: > AIX5.3L > Oracle 9.2.0.7 > SAN - IBM DS4800 > > We've got 8 filesystems for Oracle data files. Redo, Archive, Undo and > Temp are all on seperate disk/filesystems from the data files. > > All the Oracle datafiles are on RAID5 LUNs with 12 15K RPM 73 (68 usable) > GB drives. SAN Read and Write Caching are both enabled. > > <snip> > -- Cheers, -- Mark Brinsmead Senior DBA, The Pythian Group http://www.pythian.com/blogs