Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 02 Jun 2001 09:48:36 -0400
From:      Bill Moran <wmoran@iowna.com>
To:        rene@xs4all.nl
Cc:        freebsd-questions@FreeBSD.ORG
Subject:   Re: restore fails when trying to run from fixit.flp (freeBSD)
Message-ID:  <3B18EEB4.B5BFE6C7@iowna.com>
References:  <20010602124146.A12549@xs4all.nl>

next in thread | previous in thread | raw e-mail | index | archive | help
It seems to me that you had more than 5G worth of data on your /usr
part. Remake the partitions to match the original sizes (or bigger) and
try the procedure again.

-Bill

rene@xs4all.nl wrote:
> 
> Hi. I've got a 4.3-stable box for which I'm testing a backup procedure.
> 
> I have 4 disks in the box:
>   da0  scsi 8G   (bootdisk)
>   da1  scsi 2G
>   ad0  IDE 10G
>   ad1  IDE 10G
> 
> I have made a single slice of ad0, ad0s1 and newFSed that, then mounted it
> somewhere.
> 
> I have done the following on my running system:
> 
>   dump -0ua /mnt/ad0/l0/da0s1a /
>   dump -0ua /mnt/ad0/l0/da0s1e /var
>   dump -0ua /mnt/ad0/l0/da0s1f /usr
>   disklabel da0 | /mnt/ad0/l0/da0.disklabel.txt
>   dmesg | /mnt/ad0/dmesg.txt
>   cp /etc/fstab /mnt/ad0/fstab
> 
> This produced the correct files without errors in the /mnt/ad0/ structure
> 
> I then used MSDos to format the da0 disk, simulating a complete failure.
> 
> Now, restoration of my volumes I want to have done via the simplest
> procedure possible; I think this is to boot a fixit.flp (after kern.flp
> and mfsroot.flp ofcourse) and start a shell.
> 
> So, in this case, I use the installation menu of mfsroot.flp to re-create
> the partitions/slices (slightly changed in size, decreased 7G /usr to 5G,
> then enlarged / and /var) on the ad0 disk.
> 
> I then newfs'ed those new partitions (da0s1a, da0s1e and da0s1f), and
> mounted them as /mnt/da0s1*.
> 
> Restoring the first two volumes, / and /var, is done without problems. I
> simply use
>   cd /mnt/da0s1a && restore -rf /mnt/ad0/l0/da0s1a
> and the same for da0s1e ofcourse, they both work without any messages and
> re-create files so I guess they work just fine.
> 
> Now, the problem starts with restoring /mnt/da0s1f with
> 
>   cd /mnt/da0s1f && restore -rf /mnt/ad0/l0/da0s1f
> 
> the first set of errors I get is
> 
>   /: write failed, file system is full
> 
> I get this one quite often, but the process continues anyway, and then
> starts outputting a new set of errors:
> 
>   warning: '.' missing from directory ./include/<something>
> 
> or
> 
>   warning: '..' missing from directory ./include/<something>
> 
> Now, a quick look at df output shows me that my / filesystem is indeed
> full. mount shows me that / is in fact /dev/md0c which is a 2.88M ramdisk.
> 
> I thought of remounting / as /dev/da0s1a (since its restored already), but
> ofcourse(?) I can't remount a filesys that's in use...(?)
> 
> I would appreciate some clues here, I never thought that the
> backup-restore procedure of fbsd would be so complicated :((
> 
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe freebsd-questions" in the body of the message

-- 
If a bird in the hand is worth two in the bush,
then what can I get for two hands in the bush?

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3B18EEB4.B5BFE6C7>