Xen users - Apr 2007 - Mounting RHEL 4 virtual machine disks (dd/mount -o loop)

I seem to have forgotten something:

 

I have some virtual machine image files (created with the dd
if=/dev/zero technique), that have RHEL 4 u 4 installed in them.

 

I would swear that at one time I knew how to mount these image files
using the mount -o loop -t ext3 technique.

 

That doesn''t work.

 

My guess is that this has something to do with the LVM structure inside
the RHEL4 u 4 vm.

 

This question is actually unrelated to Xen, but I''m asking it here as I
learened dd/mount tricks with the qemu installer here.

 

Can someone jog my memory? What''s the magic incantation?

 

 

 

Joe.



_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

On Tue, Apr 17, 2007 at 01:44:57PM -0700, Kraska, Joe A (US SSA)
wrote:
> Can someone jog my memory? What''s the magic incantation?
modprobe loop ? :)

-- 
Marcin Owsiany <marcin@owsiany.pl>              http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216
 
"Every program in development at MIT expands until it can read mail."
                                                              -- Unknown

_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

> > Can someone jog my memory? What''s the magic incantation?
> 
> modprobe loop ? :)
Well. That was helpful... but didn''t solve the problem.
I have one box in which modprobe fatals, but another one
in which it''s okay. Here''s what happens:

[root@blade15 xen-rh64-gold]# modprobe loop
[root@blade15 xen-rh64-gold]# ls
hda  mnt
[root@blade15 xen-rh64-gold]# mount -o loop -t ext3 hda mnt
mount: wrong fs type, bad option, bad superblock on /dev/loop0,
       missing codepage or other error
       In some cases useful info is found in syslog - try
       dmesg | tail  or so

And yet, I would swear that this worked at one time.

The vm installed into "had" is configured with the default
partitioning mechanism, which involves LVM.

Joe.




_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

On 4/17/07, Kraska, Joe A (US SSA) <joe.kraska@baesystems.com>
wrote:
> > > Can someone jog my memory? What''s the magic incantation?
> >
> > modprobe loop ? :)
>
> Well. That was helpful... but didn''t solve the problem.
> I have one box in which modprobe fatals, but another one
> in which it''s okay. Here''s what happens:
>
> [root@blade15 xen-rh64-gold]# modprobe loop
> [root@blade15 xen-rh64-gold]# ls
> hda  mnt
> [root@blade15 xen-rh64-gold]# mount -o loop -t ext3 hda mnt
> mount: wrong fs type, bad option, bad superblock on /dev/loop0,
>        missing codepage or other error
>        In some cases useful info is found in syslog - try
>        dmesg | tail  or so
>
> And yet, I would swear that this worked at one time.
>
> The vm installed into "had" is configured with the default
> partitioning mechanism, which involves LVM.
kpartx ?
http://fedoraproject.org/wiki/FedoraXenQuickstartFC5

jerry

-- 
"Oh joy! Rapture! I''ve got a brain"
-Scarecrow

_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

On Tue, 17 Apr 2007, Kraska, Joe A (US SSA) wrote:
>>> Can someone jog my memory? What''s the magic incantation?
>> modprobe loop ? :)
> Well. That was helpful... but didn''t solve the problem.
...
> [root@blade15 xen-rh64-gold]# mount -o loop -t ext3 hda mnt
> mount: wrong fs type, bad option, bad superblock on /dev/loop0,
you want

losetup /dev/loop1 /dev/hda

then kpartx -av /dev/loop1

(the partitions are in /dev/mapper/loop1px)

then mount /dev/mapper/loop1p2 /mnt

or something like that

...
> The vm installed into "had" is configured with the default
> partitioning mechanism, which involves LVM.
LVM is different... if the above doesn''t work, try running
vgscan/lvscan
after you losetup.



_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

> > [root@blade15 xen-rh64-gold]# mount -o loop -t ext3 hda mnt
> > mount: wrong fs type, bad option, bad superblock on /dev/loop0,
> 
> you want
> 
> losetup /dev/loop1 /dev/hda
> 
> then kpartx -av /dev/loop1
> 
> (the partitions are in /dev/mapper/loop1px)
> 
> then mount /dev/mapper/loop1p2 /mnt
> 
> or something like that
> 
> ...
> 
> > The vm installed into "had" is configured with the default
> > partitioning mechanism, which involves LVM.
> 
> LVM is different... if the above doesn''t work, try running
vgscan/lvscan
> after you losetup.
Mkay. Maybe it wasn''t obvious but "hda" in the above mount
command is
the
Xen image file. It was the file made with dd, into which RHEL44 was
installed
using Qemu. I''m a bit confused in your instructions, because you seem
to
be
forwarding me to a /dev/hda, and don''t reference my file based image at
all.
That I can see.

Sorry about my confusion. My sojourn into Xen has me learning basic
low level linux commands that other gurus take for granted. :)

Joe.



_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

> kpartx ?
> http://fedoraproject.org/wiki/FedoraXenQuickstartFC5
Thank you. Helpful notes. The section saying that lomount cannot
work with LVM volumes suggests to me that mount -o loop would not
either. Perhaps my past mounting successes were only associated
with mounting the dd image file PRIOR to installing the OS with qemu?

Anyway, I''ll try kpartx on the image file, yes.

Thank you.

Joe.


_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

On Tue, 17 Apr 2007, Kraska, Joe A (US SSA) wrote:
> Mkay. Maybe it wasn''t obvious but "hda" in the above
mount command is
> the
> Xen image file. It was the file made with dd, into which RHEL44 was
> installed
> using Qemu. I''m a bit confused in your instructions, because you
seem to
> be
> forwarding me to a /dev/hda, and don''t reference my file based
image at
> all.
> That I can see.
Ahh.  my mistake then.

it''s

losetup /dev/loop1 /path/to/file

now /dev/loop1 (you may need to increment or decrement the 1 to get a 
free loop device)  is a device in the Dom0 that looks a lot like how the 
DomU would see /dev/hda.

so now you go from /dev/loop1  to the sub-partitions.
kpartx -av /dev/loop1
would be how you would do it if it were straight partitions.

with lvm, you need to vgscan it, but from what I understand, vgscan scans 
all devices, including or excluding certain devices in the lvm config 
files.  (on my system, /etc/lvm/lvm.conf specifies a filter= that excludes 
the cdrom device)     You need to somehow get vgscan to read /dev/loop1.

(it is possible that just ''vgscan'' or
''lvscan'' after you do the losetup
will do the job.)

_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users