pah57
asked on
How do I know if it's my motherboard, hard drive or something else that's died?
My server died a couple of days ago - I had trouble logging in (something about respawning too quickly) and ended up rebooting it. At first the BIOS saw all three HDDs (2*IDE + 1*SATA) but Fedora wouldn't boot. Reboot #2 had the BIOS mis-reporting one of the IDE drives - instead of Maxtor it showed "Mgfsgr". 3rd reboot - no HDDs at all!
I should probably include my h/w specs before I go too far. It's a home-built tower system; mobo is an ASUS P5GD1-VM; CPU is a liquid-cooled P4 540 Socket 775 HyperThreading @ 3.2GHz. (Mobo+CPU+power ~1yr old.) The cooling system and fans were running fine, although one HDD did feel warmer than usual.
I left the server turned off for about half an hour, to ensure it wasn't a heat issue, and powering it up resulted in the same mis-reporting then nothing detected by the BIOS.
I've tried disconnecting/removing all the drives and re-connecting the IDE drives individually (the SATA drive isn't bootable), but the BIOS didn't recognise either of them.
Fortunately a local store had a 200Gb IDE HDD on sale that day(!) so I picked one up and installed it (as the only HDD in the server) - same scenario! (This makes me think it's the mobo.)
I've tried the server's boot drive in a spare PC (older mobo/CPU) but it wouldn't boot from it. I tried the new HDD in the spare PC (BIOS saw it OK) but when I tried installing a new o/s (Damn Small Linux) on it, the install hung. Tried install again, this time with (known to be good) Win2k CD - hung again.
Could the problem be my motherboard? How can I tell? Could a mobo problem have screwed up ALL my drives?
Any ideas? How can I go about troubleshooting this further? I don't have any compatible components (e.g. mobo/CPU) other than a HDD, but I really don't want to use my one working drive in it in case I screw it up too! and buying replacements in the hope they'll fix it (i.e. without knowing what's really broken) isn't an option.
I should probably include my h/w specs before I go too far. It's a home-built tower system; mobo is an ASUS P5GD1-VM; CPU is a liquid-cooled P4 540 Socket 775 HyperThreading @ 3.2GHz. (Mobo+CPU+power ~1yr old.) The cooling system and fans were running fine, although one HDD did feel warmer than usual.
I left the server turned off for about half an hour, to ensure it wasn't a heat issue, and powering it up resulted in the same mis-reporting then nothing detected by the BIOS.
I've tried disconnecting/removing all the drives and re-connecting the IDE drives individually (the SATA drive isn't bootable), but the BIOS didn't recognise either of them.
Fortunately a local store had a 200Gb IDE HDD on sale that day(!) so I picked one up and installed it (as the only HDD in the server) - same scenario! (This makes me think it's the mobo.)
I've tried the server's boot drive in a spare PC (older mobo/CPU) but it wouldn't boot from it. I tried the new HDD in the spare PC (BIOS saw it OK) but when I tried installing a new o/s (Damn Small Linux) on it, the install hung. Tried install again, this time with (known to be good) Win2k CD - hung again.
Could the problem be my motherboard? How can I tell? Could a mobo problem have screwed up ALL my drives?
Any ideas? How can I go about troubleshooting this further? I don't have any compatible components (e.g. mobo/CPU) other than a HDD, but I really don't want to use my one working drive in it in case I screw it up too! and buying replacements in the hope they'll fix it (i.e. without knowing what's really broken) isn't an option.
PCBONEZ is right the PSU could cause such problems.
Try a different PSU just to check..
Try Ultimate Boot CD ( UBCD ). it has a lot of test on it you can run.
http://www.ultimatebootcd.com/download.html
Try a different PSU just to check..
Try Ultimate Boot CD ( UBCD ). it has a lot of test on it you can run.
http://www.ultimatebootcd.com/download.html
it can be your motherboard. i had once a system that came in with a burn mark on the disk; when i connected another disk, it burned it right away - so yes, a motherboard CAN damage a disk.
i would first make sure that your spare pc works again, in order to be able to use it as test system for the other parts : disk, cd drives, ram ...
ijust hope that the fault the mobo creates on the disk is not causing the disk to burn the mobo too . .
i would first make sure that your spare pc works again, in order to be able to use it as test system for the other parts : disk, cd drives, ram ...
ijust hope that the fault the mobo creates on the disk is not causing the disk to burn the mobo too . .
I'd vote for the motherboard being the issue. When you get inconsistent detections like that, that's going to be one of your top culprits. I'd personally say go ahead and replace it myself.
ASKER
A friend (who runs his own IT tech support company) suggested upgrading the BIOS, which I did yesterday and it *seems* to have fixed the problem!
No hardware changes required, just a bit of a fight to get it to boot from a USB memory stick and then the BIOS upgrade went smoothly.
One of the HDDs appears to be fine; one is mostly OK but a couple of the Linux partitions have an fstype of "unknown" instead of ext3; the third drive (which of course was the one with all my backups on it) looks like it's been wiped.
No hardware changes required, just a bit of a fight to get it to boot from a USB memory stick and then the BIOS upgrade went smoothly.
One of the HDDs appears to be fine; one is mostly OK but a couple of the Linux partitions have an fstype of "unknown" instead of ext3; the third drive (which of course was the one with all my backups on it) looks like it's been wiped.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
The first things in common are the motherboard and PSU.
The first thing I'd check is the PSU.
- That not only powers the drives it powers the board.
- A real bad PSU problem could have damaged drives (and/ot the board).