On Tue, 24 Feb 2009 15:21:23 -0500
"Barclay, Daniel" <dan...@fgm.com> wrote:

> Eduardo M KALINOWSKI wrote:
> > Barclay, Daniel wrote:
> ...
> >> Since GRUB hasn't loaded the kernel file yet, GRUB can't be using
> >> the kernel and its md driver, and therefore can't be reading the
> >> partition _as_a_RAID_ _volume_ (/dev/mdX), right?
> >>
> >>
> >> So is GRUB just reading the partition directly to get to the file
> >> system?
> >>
> > 
> > GRUB does not know anything about RAID, so I assume this is true.
> 
> That's what I have been thinking, but I just found the message at
> http://www.mail-archive.com/linux-r...@vger.kernel.org/msg09712.html
> that says::
> 
>     ... once grub2 has determined that the intended boot partition is
>     a raid partition, the raid code takes over ... and it scans for
> all the other members of the raid array and utilizes whichever drives
> it needs to in order to complete the boot process.  ... [I]t doesn't
>     need any member of a raid1 array to be perfect[;] it will attempt
> a round robin read on all the sectors and only fail if all drives
> return an error for a given read.
> 
> Is that _just_ for GRUB2 and or does the current GRUB (0.97) in Lenny
> also do that?
> 
> 
> 
> >> ... is GRUB taking advantage of the fact that the RAID metadata is
> >> written at the end of a partition ...
> ...
> >> If so, how reliable is that?
> >>
> >> Should one put /boot on a plain, non-RAID partition on one disk and
> >> somehow (manually or automatically) maintain a backup /boot
> >> partition on 
>  >> the second disk, or is it fine to put /boot on a mirrored
>  >> partition (so maintaining redundancy is automatic) and let GRUB
>  >> read the partition directly?
> >>
> > 
> > Again, while I haven't tried, I've seen several reports that this
> > works. ... So why make things more complicated and not automatic?
> 
> I don't get why you're asking that.  I _am_ trying to avoid the
> complicated and non-automatic solution (trying to check whether the
> simpler solution is reliable).
> 
> 
> 
> Daniel
Hi, Daniel et al

The following is an outline of my setup on a couple of the system disks.

Roughly I have two md devices for the host system, md0 & md1:

j...@host:~$ df -m
Filesystem           1M-blocks      Used Available Use% Mounted on
/dev/md0                  9389      7157      1755  81% /
tmpfs                     4005         1      4005   1% /lib/init/rw
udev                        10         1        10   2% /dev
tmpfs                     4005         0      4005   0% /dev/shm
/dev/dm-0                40318     11628     26642  31% /home
/dev/dm-4                19686     15356      3331  83% /home/jack/XP_VDI
/dev/dm-6                30238      7453     21250  26% /home/jack/suse
/dev/dm-2                 3024        70      2802   3% /tmp
/dev/dm-1                 8064      3426      4229  45% /var
/dev/hda                    90        90         0 100% /media/cdrom0

The only concern I have is that / is marginally small. It's
expandable tho. 
 /boot is just on /mdo. I have run without incident for over a year.
Provides sufficient isolation from my bumbling around...  8-)

FWIW. 

Jack
.


-- 


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to