There are two possible workarounds/fixes:

a) Ignore failure of creating the new bridge, and just use the existing one.
b) Change the default bridge device from virbr0 to virbr%d. This will make 
libvirt try virbr0, virbr1, etc. until it finds a free one.

a) is useful if libvirt died for some reason without shutting down the
bridge, since newly started vm's (after you start libvirtd again,
obviously) will be attached to the same virtual network as the other
vm's. b) is useful if for some reason the user already had a virbr0
before libvirt came along, as we avoid imposing on his existing network.

Naturally, we should do whatever we can to make sure libvirt never dies
without shutting down the bridge, but if that fails, what do we prefer?
a) causes the least problems, but is only really sane if we consider
"virbr0" to be unique enough (i.e. is it reasonable to expect the user
to *not* have set up such a bridge beforehand?).

-- 
libvirt fails work if virbr0 exists when it is started
https://bugs.launchpad.net/bugs/185598
You received this bug notification because you are a member of Ubuntu
Bugs, which is the bug contact for Ubuntu.

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to