SpringTownDBA (8/22/2012)
Some advice for the OP, if you can take down your SAN by slamming it too hard, you should consider reducing your hba queue depth to limit the number of outstanding io requests. you might take a hit on max throughput, but if the san can't handle it anyway, you're not losing much.
Moved it to the new LUN, but I won't know if it helped until this weekend when I run something really I/O intensive.
Will report back...