now really solved Re: [Xen-users] Error: Device 0 (vif) could not be connected. Backend device not found.

[Jens Nachtigall <nachtigall@xxxxxx>]

Am Montag, 19. Juni 2006 06:29 schrieb pak333@xxxxxxxxxxx:
> I am having the same problem, however on RHEL 2.6.16.
> I have tried setting up the bridge manually but it does now work.
> Can someone give some clues on what we should do to solve this problem in
> RHEL

Please ignore my previous mail, here is how to do it without modifying 
/etc/network/interfaces:

I read /usr/share/doc/xen-utils-3.0/README.Debian.gz

> * About networking:
>
>    By default Xen modifies your networking configuration, creating a
> bridge. To avoid breaking a machine's connection to the network the debian
> package doesn't touch the network configuration unless requested. On the
> other hand xen needs to connect the guest domains somewhere, in order for
> them to have networking, and expects that the xen bridge exists. You should
> hand-tune your networking configuration by editing /etc/xen/xend-config.sxp
> and possibly /etc/network/interfaces to decide how to connect your domains
> to the network.

Maybe this is the same on Red Hat? I had to comment out 
#(network-script network-dummy)
and change it to 
(network-script network-bridge)

> * About loop devices:
>
>    If you plan hosting virtual domains with file backed block devices (ie.
> the ones xen-tools creates by default) be careful about two issues:
>
>    1. Maximum number of loop devices
>       By default the loop driver supports a maximum of 8 loop devices. Of
>       course since every xen domain uses at least two (one for the data and
> one for the swap) this number is absolutely insufficient. You should
> increase it by adding a file named local-loop in /etc/modprobe.d containing
> the string "options loop max_loop=128", if the loop driver is compiled as a
> module, or by appending the string max_loop=128 to your kernel parameters
> if the driver is in-kernel. Of course you can increase or decrease the
> number 128 as you see fit.

I created this file with options loop max_loop=128. Do a "ls -l /dev/loop*", if 
there 
are only a few loop devices and /var/log/daemon.log says
logger: /etc/xen/scripts/block: Failed to find an unused loop device

Then you should do this as well as what comes here:

>
>    2. Driver loading (only if loop is compiled as a module)
>       Normally the loop driver gets loaded when the first loop device is
>       accessed. When using udev, though, the loop devices get created only
>       after the driver gets loaded. This means that xen will fail if the
> loop driver is not already loaded when it tries to start a file-backed
> virtual domain.  To fix this just add "loop" in your /etc/modules file,
> thus forcing it to be loaded at boot time.


This could all also be necessary on RH, so give it a try. Maybe RH also has 
such a README?

HTH,
Jens

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