dapsychous
asked on
System crashes when running chkdsk on RAID 0
OK, I'm at my wit's end with this one.
I have a customer that has brought in a computer that we built for them (spiffy gaming machine). He was running an ASUS A8N SLI Deluxe (nForce4 Sli chipset) on a RAID 0 (striped) with twin WD 250GB Sata hard disks.
The problem is this:
Due to an unrelated issue (crappy video drivers that have since been replaced), windows was shut down improperly, naturally, chkdsk came up when the computer booted and attempted to scan as it typically does. During the first phase (verifying files) the computer locks at 7%, then restarts a few seconds later. It is ALWAYS 7%. I can clear the dirty bit with no problem, and we have in the past when it was previously brought in for this same problem, but as soon as he needs to run a chkdsk or windows craps out (as it typically does), we'll be right back where we are now (as we have been four times now).
So far we have:
destroyed and rebuilt the raid array, with a format and reload
low leved formatted the drives using WD Data Lifeguard
COMPLETELY tested the drives using WD data Lifeguard, and Microscope
Replaced the drives
replaced the sata cables
replaced the mainboard
replaced the memory
replaced the video card
removed EVERYTHING non-vital to operation, with new memory and new sata cables
set the speed cap jumpers on the hard disks to cap speed at 150 mb/s
enabled spread-spectrum clocking on both drives
Spent 5 hours on the phone with ASUS, who sent us two replacement motherboards, both of which have exibited the same problem
Attempted to call Microsoft, only to have them demand money for support.
called western digital, who, so for, has not said anything useful
had several heart-to-hearts with google to find others who have had a similar problem.
ran out of ideas.
This customer is driving us absolutely insane. We have had his computer for about three months, now, and he calls us multiple times a day (I probably would, too). We can't just disable the raid controller, because he is adamant that he wants to keep his raid setup. We can't put any more money into this computer, because factoring lost labor time, we have lost hundreds on this deal. We stand behind our machines, but this one has us dumbfounded. ASUS claimed they tested the raid capability and its ability to run a checkdisk, however, it still doesn't work.
Any help with this would be immensly appreaciated.
--[adam]
I have a customer that has brought in a computer that we built for them (spiffy gaming machine). He was running an ASUS A8N SLI Deluxe (nForce4 Sli chipset) on a RAID 0 (striped) with twin WD 250GB Sata hard disks.
The problem is this:
Due to an unrelated issue (crappy video drivers that have since been replaced), windows was shut down improperly, naturally, chkdsk came up when the computer booted and attempted to scan as it typically does. During the first phase (verifying files) the computer locks at 7%, then restarts a few seconds later. It is ALWAYS 7%. I can clear the dirty bit with no problem, and we have in the past when it was previously brought in for this same problem, but as soon as he needs to run a chkdsk or windows craps out (as it typically does), we'll be right back where we are now (as we have been four times now).
So far we have:
destroyed and rebuilt the raid array, with a format and reload
low leved formatted the drives using WD Data Lifeguard
COMPLETELY tested the drives using WD data Lifeguard, and Microscope
Replaced the drives
replaced the sata cables
replaced the mainboard
replaced the memory
replaced the video card
removed EVERYTHING non-vital to operation, with new memory and new sata cables
set the speed cap jumpers on the hard disks to cap speed at 150 mb/s
enabled spread-spectrum clocking on both drives
Spent 5 hours on the phone with ASUS, who sent us two replacement motherboards, both of which have exibited the same problem
Attempted to call Microsoft, only to have them demand money for support.
called western digital, who, so for, has not said anything useful
had several heart-to-hearts with google to find others who have had a similar problem.
ran out of ideas.
This customer is driving us absolutely insane. We have had his computer for about three months, now, and he calls us multiple times a day (I probably would, too). We can't just disable the raid controller, because he is adamant that he wants to keep his raid setup. We can't put any more money into this computer, because factoring lost labor time, we have lost hundreds on this deal. We stand behind our machines, but this one has us dumbfounded. ASUS claimed they tested the raid capability and its ability to run a checkdisk, however, it still doesn't work.
Any help with this would be immensly appreaciated.
--[adam]
ASKER
Well, that's all well and good, but my customer REQUIRES that this be a RAID 0. We have tried to take him off of it, but he always pitches a fit. Additionally, I'm not running chkdsk, it's autorunning at boot because of a flagged dirty bit. Simply clearing the dirty bit won't help, because it just does it again next time an autocheck is scheduled due to a windows glitch or something.
ASKER
plus, he doesn't have important data anyway, its his gaming machine. He keeps his data on his other computer, so this isn't a data issue, it's a speed and whether-or-not-it-works issue.
Have you tried a different machine(non ASUS) or different (non WD)drives?
Sounds like a bug in the firmware on the ASUS board.
If a dirty bit is being set,it sounds as if the file system is not being shutdown properly(which could be a driver and or firmware).
Sounds like a bug in the firmware on the ASUS board.
If a dirty bit is being set,it sounds as if the file system is not being shutdown properly(which could be a driver and or firmware).
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Educate him on the problems with RAID 0 or discontinue support, that is my only other suggestion. Just because someone is ignorant = unaware of the problems with RAID 0 doesnt mean you can solve it. To run RAID 0 for a gaming machine is just about as stupid as you can get -- a nerd with much reading, but little understanding, perhaps?
Clear the dirty flag and rename the chkdsk executable so it can't run anymore :)
I saw this once years ago: NT4 on a compaq 1600 running RAID5. same sort of thing, we ran a chkdsk over a weekend and came back to find it hadn't moved beyond 32%. It was our PDC too (which we then moved of course).
I did a parallel install to it, and then backed up the entire volume (kinda like ghosting the OS without ghost). restored it to an identical bit of kit in the lab and ran chkdsk (our plan was to swap hardware out to fix the issue). Well I'd be darned if the clone didn't exhibit the exact same issue.
Unfortunatley we never did resolve the problem - even running from the parallel install would hang. However, we could assume the issue was related to software rather than hardware.
Are you backing up the customers machine and then restoring after re-jiggin the hardware etc? If you are, maybe try a fresh install and NO restore.
I saw this once years ago: NT4 on a compaq 1600 running RAID5. same sort of thing, we ran a chkdsk over a weekend and came back to find it hadn't moved beyond 32%. It was our PDC too (which we then moved of course).
I did a parallel install to it, and then backed up the entire volume (kinda like ghosting the OS without ghost). restored it to an identical bit of kit in the lab and ran chkdsk (our plan was to swap hardware out to fix the issue). Well I'd be darned if the clone didn't exhibit the exact same issue.
Unfortunatley we never did resolve the problem - even running from the parallel install would hang. However, we could assume the issue was related to software rather than hardware.
Are you backing up the customers machine and then restoring after re-jiggin the hardware etc? If you are, maybe try a fresh install and NO restore.
Recomment you use RAID 1 with 2 disks, and you do not run checkdisk for the drives, let them run as is.