Re:lvm+raid1 boot disk: initrd/initramfs? howto?
Hello all,
Just want to add some details:
o given that /, /boot and swap are purely raid1
I can get rid of initrd/initramfs (not very familar with any ;-(
o I don't need to use /etc/mdadm/mdadm.conf (autodetection works fine)
That said, I finaly find a work-around:
---------- Initial header -----------
>From : joel.soete@tiscali.be
To : "debian-hppa" debian-hppa@lists.debian.org
CC :
Date : Fri, 27 Jan 2006 17:15:40 +0100
Subject : lvm+raid1 boot disk: initrd/initramfs? howto?
> Hi all,
>
> Desperating that lvm will ever offer the 'encapsulation' of /boot in lvm like
> hpux does), i am trying to install a more common lvm+raid1 boot disk with
> following scheme:
> boot disk will be build on raid1 + lv, i.e:
> o the / (aka root), /boot, swap will be purely raid1
> (as recommended)
>
> o the rest /home, /var, /tmp will be in LVs over raid1
> (to get benefit of resizing LVM facilities)
>
> ps: raid1 base is still necessary because against announces
> LVM is not yet able to manage its own raid1 stuff ;-(
>
> Additional data (/Develop, /Sources, /DebApt, /MultiCd, ...) will stand
> on LVs over a software raid5 array of 6*4Gb disk (i.e. 5 + 1 spare)
>
[...]
> Starting raid devices: done.
> Setting up LVM Volume Groups...
> Reading all physical volumes. This may take a while...
> No volume groups found
> No volume groups found
> No volume groups found
> Will now check all file systems.
> fsck 1.39-WIP (31-Dec-2005)
> Checking all file systems.
> [/sbin/fsck.ext3 (1) -- /boot] fsck.ext3 -a -C0 /dev/md0
> fsck.ext3: No such file or directory while trying to open /dev/md0
> /dev/md0:
> The superblock could not be read or does not describe a correct ext2
> filesystem. If the device is valid and it really contains an ext2
> filesystem (and not swap or ufs or something else), then the superblock
> is corrupt, and you might try running e2fsck with an alternate superblock:
> e2fsck -b 8193 <device>
>
[...]
>
> fsck died with exit status 8
> File system check failed.
> A log is being saved in /var/log/fsck/checkfs if that location is writable.
> Please repair the file system manually.
> A maintenance shell will now be started.
> CONTROL-D will terminate this shell and resume system boot.
> Give root password for maintenance
> (or type Control-D to continue):
> (DebianUnstable)root@(none):~# cat /proc/mdstat
> Personalities : [linear] [raid0] [raid1] [raid5] [raid4]
> md1 : active raid1 sdc3[1]
> 385984 blocks [2/1] [_U]
>
> md2 : active raid1 sdc5[1]
> 2047872 blocks [2/1] [_U]
>
> md3 : active raid1 sdc6[1]
> 6296448 blocks [2/1] [_U]
>
> md0 : active raid1 sdc2[1]
> 128960 blocks [2/1] [_U]
>
> unused devices: <none>
> (DebianUnstable)root@(none):~# ll /dev/md0
> ls: /dev/md0: No such file or directory
> (
> jso@iatst001:~$
>
> Having no experience at all with this new udev env, I have a lot of questions:
>
> 1/ which could be the reason why this 'Waiting for /dev to be fully
> populated...done (timeout).'
>
(Still have to find the reason of time out)
> 2/ could it be the reason of next pbs (most probably)?
>
The actual pb is that udef didn't populate /dev with md[01234].
The work-around mentioned here (and in other many places):
<http://www.jochem.name/article.php3?id_article=23>
i.e. add following stuff in /etc/udev/links.conf
M md0 b 9 0
M md1 b 9 1
M md2 b 9 2
M md3 b 9 3
M md4 b 9 4
(mmm, fwiw it seems to be a well known pb since a while and not yet fixed
upsteam?)
> 3/ is it possible to continue to use the md mnemonics:
> /dev/md/boot
> /dev/md/swap
> /dev/md/root
> /dev/md/more_root
>
I trust it should be possible by defining persistent names in udev configs but
still have learn it ;-)
Thanks for all,
Joel
---------------------------------------------------------------
A free anti-spam and anti-virus filter on all Scarlet mailboxes
More info on http://www.scarlet.be/
Reply to: