I have a Dell 2950 that has PERC5's in pci2 and pci3, both which have md1000's hanging off them, on a RHEL server. I installed a new PERC6 into pci1, hooked up a new md1000, went into the raid bios and RAID5'd the new drives, but then when RHEL started, it said that volumes that are on my existing PERC5s have a bad magic number in the super-block and want me to do an e2fsck -b and it wouldn't boot.

I then removed the PERC6 from the server, and it started up normally, reporting no errors on my existing volumes.

Any ideas what I did wrong or how I can fix it? I think the new PERC6 is causing a conflict with my current setup and its confusing Linux where certain volumes are truly located