Re: Swapping activated while RAID reconstruction in progress
I do not subscribe to "debian-isp," so my reply may be rejected.
As far as I know, swapping is turned on when "swapon" is invoked from
/etc/init.d/checkroot.sh or /etc/init.d/mountall.sh. The fix for the bug
I reported was to wrap the "swapon" command in a check for whether
"resync" appears in /proc/mdstat, and it looks to me as if the fix in
Potato should be effective.
Standard Debian rescue disks do not, I think, have kernels with the v0.90
software RAID patches. It is certainly possible to make one, but this is
something of a special situation and, if you are using v0.90 RAID, then
you are outside the scope of what Debian supports out of the box. As a
quick hack, you can replace the kernel on a standard rescue floppy with a
custom kernel, although this could desynchronize the modules; if you do
not need modules to access the disk with that custom kernel, this could be
enough to allow you to do what you want. Once booted on a rescue disk
with a kernel that has v0.90 RAID support, you can conventionally mount
the hard drive partition that usually serves as your root filesystem onto
some convenient place, such as /mnt, and hand-edit configuration files in
/mnt/etc/init.d. Alternatively, you could allow the resync to take place
and complete while booted on the custom kernel, and then reboot normally.
If you are absolutely desperate, you could try booting from the hard drive
using a boot prompt argument which specifies a shell as the grandfather
process instead of init as usual; something like "Linux init=/bin/bash"
ought to do it. This would bypass login and init processing, which would
therefore bypass /etc/init.d and /etc/rc?.d processing where swapping is
ordinarily turned on, allowing your v0.90 RAID resync to complete.
-- Mike
On 2001-01-28 at 23:26 -0600, Sanjeev Gupta wrote:
> Folks,
>
> I refer to bugs #61242 and #61227 (closed). I believe this has hit one of
> our machines.
>
> Is there some way that I can tell lilo to not enable swapping? Or access
> the /etc partition with a standard debian rescue CD?
>
> Thanks. The interesting part is that the machine tracks potato, whwr I
> think this bug has been fixed.
Reply to: