You are not logged in.

#1 2012-10-10 09:33:43

shetland_breeder
Member
From: England
Registered: 2010-05-17
Posts: 67
Website

Should RAID1 boot degraded?

I can no longer boot from my RAID1 array with one of the drives disconnected.
I know this used to work but I haven't tested it for a couple of years now and I'm wondering whether it's still supposed to work or whether I now have to assemble the arrays manually.

I have three sets of partitions that  assemble as md0 (/) md1 (swap) and md2 (/home) plus several others that are probably not relevant.
My HOOKS line in mkinitcpio.conf is:

HOOKS="base udev autodetect pata scsi sata mdadm_udev resume filesystems"

What I see if I boot with one drive disconnected is:

Booting the kernel.
:: running early hook [udev]
:: running hook [udev]
:: Triggerng events...
:: running hook [resume]
[      2.682471] read-error on swap-device (9:1:0)
:: mounting '/dev/md0' on real root
Segmentation fault

then I get dropped to the shell.

I guess the segfault occurs because the array md0 has not been assembled.

cat /proc/mdstat shows all the arrays are recognised but none have been assembled, eg:

md0: inactive sda1[0](S)

I haven't yet tried assembling by hand - mostly because I've forgotten where I mount the assembled (degraded) array to continue.

Anyone know whether this should work or not or do I have to assemble the arrays by hand now?

Pete

Offline

Board footer

Powered by FluxBB