July 9, 2025 at 5:56 PM, "Mike Larkin" <mlar...@nested.page
mailto:mlar...@nested.page?to=%22Mike%20Larkin%22%20%3Cmlarkin%40nested.page%3E
> wrote:
>
> On Wed, Jul 09, 2025 at 02:42:44PM +0000, Jon Tabor wrote:
>
> >
> > Good day everyone,
> >
> > I recently ran through the process of updating my ancient vmm host server
> > from 7.2 all the way up to 7.7 (stepping through each version using
> > sysupgrade). Everything went extremely well, with the exception of one
> > issue: previously when I started a vmm guest, the tap interface (tap0,
> > tap1, etc) automatically added itself to the bridge (bridge0) so the guests
> > could communicate on the network, but now they no longer do. I have to
> > manually add them using
> >
> > ifconfig bridge0 add tap0
> >
> > every time I start/restart a vmm guest.
> >
> > Any thoughts on why that would be? Here's my vm.conf; not sure what else
> > would be in play here, so let me know if there are other configuration
> > files I should show.
> >
> > vm.conf:
> >
> > socket owner taborj:taborj
> >
> > switch "bridge" {
> > interface bridge0
> > }
> >
> > vm "athos" {
> > memory 10G
> > enable
> > disk /mnt/sd1a/vmware/athos.qcow2
> > disk /mnt/sd1a/vmware/athos_usr.qcow2
> > interface { switch "bridge" }
> > }
> >
> > vm "porthos" {
> > memory 512M
> > enable
> > disk /mnt/sd1a/vmware/porthos.qcow2
> > disk /mnt/sd1a/vmware/porthos2.qcow2
> > interface { switch "bridge" }
> > }
> >
> > Thanks,
> > ------------------
> > Jon Tabor
> > tab...@obsolete.site mailto:tab...@obsolete.site
> >
> I had similar weird problems with bridge(4) long ago and switched to veb(4) at
> the suggestion of some of the other devs. I just toss all the tapN into the
> veb at boot (/etc/hostname.veb0), that way I don't need to worry about them
> not getting added later.
>
> -ml
>
That's a good suggestion, I'll look into it. I've seen a few other places
mention using veb instead of bridge, but I didn't really dive into why.
Thanks,
------------------
Jon Tabor
tab...@obsolete.site mailto:tab...@obsolete.site
