We encountered a similar situation. We have two node SQL clusters with Dell 1955 blades (16 Gb) and an EMC SAN. The first "fix" was getting a driver patch from...