Matirix RAID Issues

Pyongmu · May 1, 2008, 09:03 PM

I have been running RAID0/ RAID5 combo on 4 400GB NS Seagate disks using the Matrix RAID feature of the ICH9R chipset for about a year now. The os and the progs are installed in the RAID0 and all valuable data in the RAID5. All was fine until recently when I began receiving reports of disk failure during the boot I can still boot to windows so is obvious the RAID0 portion of the "bad disk" is still working. In windows, the Intel Matrix Storage Console reports the RAID0 is functioning normally whereas the Raid5 is degraded.
I ran Seatools on the compromised disk but it failed to identify any problems. Any ideas?

Judas · May 2, 2008, 09:49 AM

Did seatools detect that SMART had been tripped?

Try doing a repair of the degraded raid 5 array.

I'd also suggest making sure you backup your critical files asap before proceeding with any fix at all.

Pyongmu · May 2, 2008, 05:31 PM

According to Seatools, SMART has not been tripped and furthermore there is nothing wrong with the disk (passed the long test 3 times).

Repairing the degraded RAID was the first option I sought but Intel Matrix does not appear to have that function as an option as far as I can determine: nothing in the BIOS or in the Windows GUI.

I am hoping to avoid having to reset the RAID0 since it is working fine. I am getting ready to reset the RAID5 as soon as the backups are complete to see if that helps.

Judas · May 2, 2008, 05:40 PM

My fingers are crossed for you...

Pyongmu · May 2, 2008, 05:54 PM

Thank you, Judas.

PangingJr · May 3, 2008, 01:52 AM

once you finish backing up your files, and before you delete and recreate the RAID array and volumes, try this first, this may not help in your case but try it.

turn off the computer, unplug and replug all SATA cables from both ends of the cable, CLR_CMOS,
and then restart the computer and go into BIOS setup (no need to go into the Intel Matrix Storage Manager OROM page), and reset the Integrated Peripherals page for the SATA RAID/AHCI Mode [RAID] and SATA Port0-3 Native Mode [Enabled] and the rest as you usually set them, but do not overclock yet, and then Save & Exit the setup.

if the CLR_CMOS doesn't help, you can try reflashing BIOS with the same or a different version of BIOS.

I had that worked for me a few months ago after some bad overclocking attempts, then got a few, failed member drives and some Windows important boot files is missing. it doesn't seems to work, but it did work.

just a note, about the QFlash utility that exist on Gigabyte's boards, if you've never noticed this (small) problem before.

in some version of BIOS's... if your BIOS and IMSM OROM has been set to use RAID and if you load BIOS optimal default setting before you perform the BIOS reflash and you also use an USB drive to store your BIOS file and you also use the QFlash utility for the BIOS flashing, then when you use the End key at the system boots to get to the QFlash utility, and you find you are not able to select the USB drive... or sometimes, your keyboard stops working all of sudden in the QFlash utility page.

i find that i can fix this by, make sure that you reset the Integrated Peripherals page, for the SATA RAID/AHCI Mode to [RAID] before you enter the QFlash utility.

Pyongmu · May 3, 2008, 07:27 AM

Appreciate your input Panging. Unfortunately, by the time I read your post, I had already deleted and recreated the RAID volumes. But I will keep your suggestions in my notes for future use should I experience similar problems again.

Anycase, deleting and recreating the RAID array and volumes made the problem go away.

PangingJr · May 3, 2008, 07:56 AM

i suspected that i might be too slow to make the first comment on this one.
sorry that i wait until you talked about backing up your files in the RAID-5, lol.
i didn't want to suggest something that i'm not even sure it will help you and then it may even destroy all your important files in the volumes. believe it or not i've never killed my RAID drives, i've been doing countless (bad) overclocks. so try it next time.