GA-EX58-UD – Intel Raid 0 failed after update to F4 bios

[f_donald]

Hello,

After updating from the F3 to F4 bios I received a RAID “FAIL” prompt on reboot. The second drive on the raid 0 array was no longer a member. I had to delete the array, rebuild and recover from a drive image to get the system up again. It appears stable at present but I have never had this error on any previous MB, and it left me concerned. The system was up and stable for three weeks on the F3 bios prior to the F4 update.

Does anyone know why this would happen and/or has anyone experienced the same problem?

Should I be concerned with the bios build or the motherboard?

Thanks…

 

[bilbat]

Welcome to the early adopter's blues... The i7 is new, the x58 is new, the whole three channel DDR-3 access scheme is new: you WILL have problems - just hope none of 'em are deep enough in the hardware to require a board swap as matters develop...

 

[jitpublisher]

Er, when you update your BIOS, everything goes back to the default settings, which means your RAID controller was likely turned off, all of your boot settings and such were probably defaulted back for a single drive.
If you would have simply went back into the BIOS, turned the controller back on, reset your boot preferences-entered the RAID BIOS, pointed it back to the original drives in RAID, it should have found and re-established the array perfectly fine.

 

[f_donald]

I build my new system alongside the old until bugs are worked out, but the notion of a board swap would certainly be undesirable.

Relative to the BIOS update, I flashed from the BIOS update utility, reset the system BIOS to the previous settings and rebooted to find the drive on Ch0 dropped as a member in the RAID BIOS. No variation from many previous BIOS updates on RAID configured systems, but I have never lost a drive until this BIOS flash.

How do I point the RAID BIOS back to original drives in RAID when one drive is dropped, without losing data? I could not figure out how to reestablish the array without deleting the one that failed.

Thanks for the input and help.

 

[michaelt84]

I had the same problem running a raid 10 array ... here's the answer I received from Intel:

Thank you for contacting Intel Technical Support.

Once the drives are marked as "Missing", there is no way to rejoin these. You will need to create the RAID volume again from scratch.

When performing a BIOS update, the SATA Controller mode should be checked prior to attempting the first bootup.

Sincerely,

Carlos B.
Intel Technical Support

My mistake was not resetting the raid array before the first reboot. Looks like once its gone, its gone.

 

[bilbat]

If I were in your situation (which I wouldn't be in a million years - I believe, in most case, in letting other people do the beta testing...), here's what I would do: invest in some good drive imaging s'ware, like Ghost, and image anything on a RAI0, so you can restore it at will. I use BootIt_NG as my boot manager, and keep all my OSs & swaps on RAID0s - BootIt it has drive imaging built in, and I keep a 1.5T non-RAID just for backups and images. My RAID0s seem stable as a table, but RAID0 just makes me nervous - once they're imaged, I can quit taking tranquilizers...

 

[f_donald]

I agree and you can stop taking tranquilizers... I have actually been running Ghost going back to when it was Powerquest Drive Image. I image to a 1TB internal on the new system, keep current copies on an external 2TB RAID, and store most data on yet another three drives. This is how I recovered from the first fiasco and several subsequent replications. I can replicate the BIOS degrading the RAID array with F3 to F4, but not with F4 to F3, or F4 to F4. Gigabyte support denied any problems but advised that I stay with the F3 BIOS despite the obvious S3 issues. It may also relate to external drives that were attached at the time but I am to tired to crash the RAID again. In all, they had me reset the BIOS old school by pulling the memory and CMOC battery and reverting to F3, The battery latch is of course conveniently located under the video card. The technician also told me to unplug all external USB and SATA drives as they can cause this MB to freeze or hang at boot, and this assertion was very true in my case. At present, my old and fully functional system is conveniently located on a table to the right of the new system, and the new system appears to be running well with F3 and does not hang if the external USB drives are unplugged or off when I boot. The RAID failure is still a bit disconcerting...

 

[steveru621]

http://forums.tweaktown.com/f69/bios-flashing-how-qflash-guide-27576/

I plan using this process when I go to F4 this weekend. I have RAID0, and all these posts are making me question if I need F4.

 

[bilbat]

I like GB MOBOs, but if I have one complaint (besides, of course, that they seem to run their web-site off an old Apple IIc w/a whole 16K of RAM), it's their lackadaisical attitude toward keeping BIOS updated, and addressing known problems promptly - I've been running a beta BIOS forever, as the previous 'ready-for-prime-time' issue doesn't support E0 stepping of the 9550; it runs OK, I guess, but has various little wierdnesses that I know are just BIOS oddities - wish they'd GTST, and properly support otherwise (with a couple well-known exceptions) their otherwise great boards...

 

[Guest]

I Just fixed the same problem on a P35DQ6

with the util below

http://www.cgsecurity.org/wiki/TestDisk

It's Free

I had a XP stand alone drive also on the system as well as a 1TB raid 0

its a dos util can be run I assume from a floppy or usb stick

I had the same problem running a raid 10 array ... here's the answer I received from Intel:

Thank you for contacting Intel Technical Support.

Once the drives are marked as "Missing", there is no way to rejoin these. You will need to create the RAID volume again from scratch.

When performing a BIOS update, the SATA Controller mode should be checked prior to attempting the first bootup.

Sincerely,

Carlos B.
Intel Technical Support

My mistake was not resetting the raid array before the first reboot. Looks like once its gone, its gone.

 

[bilbat]

...once its gone, its gone.

Verily!

Here's a trick:

I use several RAID arrays, and the Intel RAID BIOS has a lag in the reboot while it connects and verifies the drives; on any single reboot, it's no big deal, but when 'BIOS tweaking' it gets really annoying. What I do, after a BIOS flash, or when tweaking, is to leave the RAID off in the BIOS (which speeds up the reboots enormously), but set the first boot drive to the DVD, and pop in a bootable copy of MemTest86+; that way, if I 'miss' the <DEL> to get into the BIOS on the reboot, the DVD 'catches' it, and the system never gets to the hard drives to screw them up...

Better safe than sorry!

 

[frontier2]

I had same problem.
I did so with Windows XP

1. Recreate the Array of Hard Disk same as they were before lost info;
2. Insert Windows XP Installation CD, and open the shell;
3. When load, if you need, insert the floppy disk with SATA RAID Drivers;
4. At this point you should find your last installation of windows;
5. now try repair the boot, fixboot, fixmbr, refix the installation of windows.

I did that when my BIOS needed to reset, i did not do with update BIOS.
Now i've bought new CPU and need do that, I hope it will work, but before to do that, i think i will save data first.

 

[bilbat]

i think i will save data first.

That is certainly the best (and easiest) policy! Treal all RAID0's as 'volatile' storage - as if they can be 'gone' at any second... I use six boot partitions on two pair of RAID0 arrays; when I am 'happy' with how one is booting and running, I 'imageback' each individual partition to another location (my 'main' working system, 7Ult64, to two locations - a RAID1 array and a 'backup' single drive!), so that when the next linux install trashes a RAID0, it's just an hour or two's work to restore 'working' partitions! [:isamuelson:8]

 

[frontier2]

i think i will save data first.

That is certainly the best (and easiest) policy! Treal all RAID0's as 'volatile' storage - as if they can be 'gone' at any second... I use six boot partitions on two pair of RAID0 arrays; when I am 'happy' with how one is booting and running, I 'imageback' each individual partition to another location (my 'main' working system, 7Ult64, to two locations - a RAID1 array and a 'backup' single drive!), so that when the next linux install trashes a RAID0, it's just an hour or two's work to restore 'working' partitions! [:isamuelson:8]

Yes man!!!
I use the RAID STRIPING to work fast and play, infact games are installed there.
My works are in desktop temporanly, but all my documents are in other NOT RAID Hard Disk, but I need to backup the SATA RAID Hard Disk however, many programs, movies, musics, games are there, if I think to reinstall All i get crazy !
So, i take an Hard Disk of 1TB, copy secure there, update BIOS, if my procedure works, ok, else I just copy the Backup data to new RAID STRIPPING i create.

I've the AMD ATHLON 64 X2 3800+ 2Ghz of 65watt, and now I'm waiting for AMD ATHLON 64 6000+ 3Ghz 2MB cache but 125watt with Zalman cooler CNPS 8700 LED and "soft" video card called ATI SAPHIRE RADEON HD5850 1GB TOXIC.
For the Video Card i don't need upgrade the BIOS, but for the CPU I need due my bios (of mainboard M2N-E ) is outdate 0205 and I need 0602.