Since I've installed gentoo, all my HDs went crazy. I'm completly at a loss 
when talking of hardware. Here is my disk layout:
hda: mandrake's root, /home
hdb: winXP root
hdc: gentoo's root

Monday, my /home (hda) stopped responding with lots of dma and other hd 
related errors. I buyed a new disk to do some backups and swap the failing 
disk. After plugging the new disk (hdd), everything was fine again. I did my 
backups, but didnt removed the faulty disk.

Tuesday, my / (hdc) stopped responding after a reboot. Every module modprobing 
gave errors. Some directories were gone (/usr/lib/portage for example). I 
mounted the gentoo's cd to get tools back, but /usr/lib/portage was back. 
Next reboot freezed after not finding some modules. I rebooted in mandrake. 
Everything was there but reiserfsck found some errors and the screwed my 
partition. I tought it would be a good idea to clear the disk and reinstall 
from scratch (on hdc again). 36 hours of compilation later, everything was 
fine again. I backuped the whole install on hdd.

Thursday, hdc went wrong again, every parition messed up again. I took hdc out 
and booted on the backup partition of hdd. All clear.

And today, /home (hda), gave me some new errors. First that message: "maison 
kernel: journal-601, buffer write failed" printed on every console. Then 
Input/output error on every access try. Mount, fdisk and hdparm never return.

The BIOS seems happy with each disk. Mandrake never gave me any HD error.
What can I do now? Change my motherboard? throw my disks away?
How can I be sure that the error come from the disks, from the MB's chipset or 
from a gentoo's driver? Could the fact that I told my disks to fo to standby 
mode after a while cause problems?

Is there a tool to tests the whole disks?

Every help would be appreciated. thanks in advance.

-- mathieu

--
[EMAIL PROTECTED] mailing list

Reply via email to