mike@mx:~$ ip route get default
local 0.0.0.0 dev lo src 127.0.0.1 uid 1000
cache <local>
mike@mx:~$ ip route get 1
1.0.0.0 via 192.168.1.1 dev enp2s0 src 192.168.1.151 uid 1000
cache
mike@mx:~$
Is my revised order of commands above correct - sorry to have to ask, but I just cannot work it out as I don’t fully understand?
mike@mx:~$ ip route show default
default via 192.168.1.1 dev enp2s0 proto dhcp src 192.168.1.151 metric 100
default via 192.168.1.1 dev wlp1s0b1 proto dhcp src 192.168.1.129 metric 600
Thanks for that reassurance - which is very welcome. I get worried when I don’t understand but it’s good that it’s easy to fix, and just to double insure I have taken a root and home snapshot with BTRFS assistant, so should be OK :).
Ok I need to look out for the Blue thumb! Nearly there… some sort of permissions problem on last command. Do I need sudo?
mike@mx:~$ virsh list --all
Id Name State
-----------------------
- win7 shut off
mike@mx:~$ virsh shutdown win7
error: Failed to shutdown domain 'win7'
error: Requested operation is not valid: domain is not running
mike@mx:~$ EDITOR="sed -i -e /bridge=/s/virbr0/br0/" virsh edit win7
Domain 'win7' XML configuration edited.
mike@mx:~$ virsh start win7
error: Failed to start domain 'win7'
error: /usr/libexec/qemu-bridge-helper --br=br0 --fd=25: failed to communicate with bridge helper: stderr=access denied by acl file
: Transport endpoint is not connected
Thanks very much h. Janssen. From Vladislavs tests it looks like I have been only using the wired connection any way. Need time to test, but it is working fast enough at the moment.
I guess fedora uses bandwidth much better than other OS. But I do appeciate your help, and having a viable bridged backup connection may well be worthwhile. So I will try your tests in a few days when I see how the wired connection is working out. So thanks very much for your hard work and research on this.
BTW can I easily delete the test bridge after running your test?
Just type “ip link del bridge1”. But the problem is solved by using the homeplug, which is, I just learned, ethernet protocol so is not affected by the wifi problems. But if wifi is much faster or worse, if you do not have wired available, you can try the ipvtap method. Please note that I’m not sure whether this adapter survives a sleep or hibernate, I’m not in the stage yet that it’s properly implemented.
(Edit: what definitively is wrong is change of MAC address on the wifi adapter, which is cloned into the VM’s XML)
Just for information: tested this ipvtap with two Oracle 8 VMs. you need one ipvtap for each VM. Because the wireless interface, the ipvtaps and the VM interfaces share one MAC address, DHCP will not work. And communication from/to the host system does not work, a “ssh” into the VM address results in a ssh into the host itself.
So the VM’s should get a manually defined IPv4 address which must be set on the ipvtap adapter too.
The VM’s get an IPv6 address and route via router advertisement, because Oracles default is “stable privacy”, the auto-generated addresses are different in the VM’s. Principle is the same: the IPv6 address in the VM has to be set on the ipvtap interface on the host. My router advertizes it’s link-local fe80:: address for routing, and this works, so the VM can be considered really connected to the LAN subnet.