I'm working with the MS SQL DBA's to try and optimize our storage for them. We are not having any specific problems right now, but we expect exponential DB growth on some of our larger databases.
We are running several tests that simulate production data to measure IO and Throughput under different scenarios:
RAID: RAID 5, RAID1+0
Partition format sizes: (64Kb, 8Kb, 4Kb, etc.)
Disk Offset settings: (Changing from an offset of 63 to 64)
Cache Settings, etc.
I'm seeing inconsistent results and believe it's related to the DMX and CX700 caching data. I'd like to clear the cache between each test so I can get a true measure of each change...
Any suggestions?
We are running several tests that simulate production data to measure IO and Throughput under different scenarios:
RAID: RAID 5, RAID1+0
Partition format sizes: (64Kb, 8Kb, 4Kb, etc.)
Disk Offset settings: (Changing from an offset of 63 to 64)
Cache Settings, etc.
I'm seeing inconsistent results and believe it's related to the DMX and CX700 caching data. I'd like to clear the cache between each test so I can get a true measure of each change...
Any suggestions?