Recommended stripe depth for Oracle on a Raid?

In the next weeks, I will have the seldom opportunity to rebuild the storage containing the DB. All file systems will be reformated.

We are using a CX-400 (Dell/EMC) connected via FibreChannel to a Sun Server; Oracle is 8.1.7 by now but will be 9.2.0.4 after rebuilding the Raid.

Papers I have read say that for Oracle, one should set the stripe depth (the amount of data in KBytes per Disk) according to the product (Oracle block size X DB_FILE_MULTIBLOCK_READ_COUNT), in my case: 8x32=256K.

I just don't understand this theory. That would mean that every I/O in a full table scan would only read from one disk at a time, one disk after the next - bad.

However, somewhere else a rule of thomb is mentioned saying that the total amount of data per I/O (i.e. blocksize x DB_FILE_MULTIBLOCK_READ_COUNT) should correspond to the stripe width. The stripe width is (the stripe depth X the number of disks in the raid group) - we have 8 disks (+1 parity +1 spare). This advice seems more reasonable to me, because every I/O would be serviced by ALL disks in parallel.

So what do you recommend? The number of users is about 5, but they use batch jobs spending hours with DB work.

Thanks a lot
Rick Denoire Received on Wed Oct 29 2003 - 17:00:51 CST