Subject: Re: kernel panic from piixide or something
To: Manuel Bouyer <email@example.com>
From: Petter Lindqvist <firstname.lastname@example.org>
Date: 03/24/2005 09:47:28
Manuel Bouyer wrote:
>On Wed, Mar 23, 2005 at 08:29:08AM +0100, Petter Lindqvist wrote:
>>>Hum, this is bad. The master disk on this channel has serious problems,
>>>and doesn't respond to anything anymore. If there is a slave drive on the
>>>channel, it's hung too.
>>>If you have any active filesystems on these drives, it may be enouth
>>>to almost hang the box, as any I/O will complete with large delays.
>>Only one disk per channel. Are you sure the disk is b0rken, or could it
>>be driver/chipset problems?
>Intel chipsets are good and well supported (thanks to intel, the docs
>are freely available). It would always be a kernel problem (for example,
>some part of memory being corrupted by something else), but the fact
>that the box didn't crash while the drive was totally unresponsive really
>points to the disk.
>Is it reproductible ?
Haven't got them to reboot the box since I hosed it with quota... Fortunately I was able to remove "quota" from /etc/fstab before it totally died. So now I just wait for them to reboot it for me, or give me a call about the remote console.. (they've forgot to tell me how to access that)