[lug] Fedora core 6 and Promise UltraTrak RM8000

[Steve A Hart]

I'm hoping someone can help me figure out what's going on with my setup. 
  I'm still a newbie when it comes to dealing with SCSI RAID problems.

I have a backup server running Fedora core 6 Kernel=2.6.22.9-61.fc6
I have 2 Promise RM8000 raid arrays and 1 Data-Silo DS570 connected to 
an LSILogic SCSI card.

All three arrays have the latest firmware installed.

The SCSI BIOS sees all three arrays with no apparent problems

When I boot using an older kernel (2.6.20-1.2933.fc6) The OS starts up 
normally and I was able to see all three RAIDs in /etc/sysconfig/hwconf. 
  I was also able to run fdisk on all three to create a single linux 
raid partitions on each one.  I was also able to run mkfs.ext3 on all 
three with no problems.  The strange part is that I can only mount the 
DS570.  Both Promise RAIDs complain about superblock errors.  Running 
fsck on the Promises also complain about superblock issues.

Now when I try to boot using the latest FC6 kernel (2.6.22.9-61.fc6) the 
OS won't even start up.  I'll try to post what I'm seeing below but 
since the darn thing does create a log, I had to write this out by hand 
so I'm sure not all the details are here.  This occurs after the kernel 
is selected in GRUB and before the root file system starts:

----------------------------------------------
message c sent on bad reselection
ABORT operation started
SCSI BUS has been reset
control msgout: c
TARGET has been reset
DEVICE reset has complete
Device offlined - not ready after error recovery
rejecting I/O to offline device
READ CAPACITY failed
Result hostbyte = DID_NO_CONNECT drivebyte=DRIVER OK SUGGEST OK
sense not available
Asking for cache data failed
assuming drive cache: write through
----------------------------------------------

This sequence just repeats over and over and the system never even gets 
close to booting up.

Any idea would be GREATLY appreciated!

Steve