D20 Blue Screen loading Windows 7-64bit using Intel controller and Marvell enabled

Question

Hi, Help,
Hardware:
D20, 4158, Four 500GB Seagate SATA drives configured as a RAID10 1TB virtual disk (partitioned into five logical drives).
Goal:
Recover or repair a degraded RAID array.
I want to be able to launch Windows 7-64bit using the Intel controller and have it access the virtual disk of the degraded RAID array attached to the Marvell controller.  Subsequently, I want to use the Marvell MRU utility to repair/recover the situation.
Situation/Issue:
Things ran well for two years and two months until yesterday.  Now, when booting I see:
1962: No operating system found. Press F1 to repeat boot sequence.
If I use the Windows 7 installation CD and load the Marvell driver, the attempt to recover the operating system reports that the MBR is corrupt and cannot be repaired.
If I boot with the Marvell controller disabled in the D20 BIOS, using a Windows 7 installation (with Marvell drivers) attached to the Intel controller, everything works except (of course) I cannot access the RAID drives.
If I reboot with the Marvell controller enabled, using the same Windows 7 installation via Intel controller, I see a "Starting Windows" announcement followed by a flash of the blue screen of death.
The BSOD message is unspecific, no error code, just...
"A problem has been detected and Windows has been shut down to prevent damage to your computer.
If this is the first time you've seen this Stop error screen,
restart your computer..."
Additional information:
If I access the Marvell BIOS (ctrl-M) and look at Devices, one of the four Seagate drives has status FREE instead of status ASSIGNED.  If I run the Marvell BIOS verify disk utility diagnosic on that FREE status disk, I get a fault error after about 30 minutes of scanning.  So, one Seagate disk is sour and somehow the MBR was corrupted at the same time.
History:
Two years ago, during the intial configuration, I could access the Marvell RAID disks from this Intel-controller-booted Windows 7 installation, but my fuzzy memory is that I had to change something in the D20 BIOS, such as the AHCI setting or similar.
Status:
Today, using the Intel controller to boot Windows 7 with the Marvell controller enabled is just BSOD.  What am I missing or failing to do or set?
Thank you in advance for any help and ideas.
  -John
Solved!
Go to Solution.

Answer 1

Update:
I replaced the 500GB Seagate disk with a new 500GB Seagate.  I ran the Marvell BIOS verify disk utility on the new disk drive and it passed with no errors.  Following the steps given in the Marvell MRU online help, I rebuilt the virtual disk.  Upon completion (which took a little more than an hour to complete), I rebooted.  Before the Seagate disk went south yesterday, the Marvell virtual disk contained one primary partition (C, and four NTFS drives (D:, E:, F:, and Z.  Now it contains three primary NTFS drives (C:, F:, and Z.  An extended partition contains D: (an unformatted volume of size 199.30MB) and an unused space of 199.30GB.
Event log:
Looking at the event log, Windows reports that the disk in question raised a warning yesterday.  Three messages share the same timestamp (all from mv64xx (a Marvell process):
"Physical disk 16 is plugged out."
"Virtual disk 0 is degraded."
"Physical disk 16 is plugged in."
Eight seconds later, UserPnp logs:
"Driver Management has concluded the process to add Service disk for Device Instance ID SCSI\DISK&VEN_SEAGATE&PROD_ST3500418AS&REV_CC38\5&248BBE23&0&000100 with the following status: 0"
One second later, UserPnp logs:
"Driver Management concluded the process to install driver FileRepository\disk.inf_amd64_neutral_10ce25bbc5a9cc43\disk.inf for Device Instance ID SCSI\DISK&VEN_SEAGATE&PROD_ST3500418AS&REV_CC38\5&248BBE23&0&000100 with the following status: 0x0."
Questions:
Is it possible that UserPnp corrupted the Marvell virtual disk after an intermittent physical disk unplugged/plugged-in event that happened when the Seagate hardware went failed.  It looks to me like my D: and E: are toast despite RAID 10.
Any ideas?  Is there anyway to recover the Marvell virtual disk configuration to the state prior to the hardware failure?  If not, what good is RAID 10?
My apologies for the long posting, but I am an old software guy who does not understand the intricacies of RAID and disk management.
Thank you in advance for any leads toward a solution,
  -John