Kvm network issues (virt-man)

English Virtualization
1

I’m a little stumped on this – I’ve got a (Windows 10) VM configuration using macvtap/libvirt/kvm/virsh/virt-man that I’ve migrated from 15.2 – I can boot the VM, it can receive an IP address from the DHCP server; however, other Windows boxes can’t ping it, nor can they connect via RDP.

pinging out of the machine results in about 57% packet loss, and it cannot rdp out.

The strange thing is, if I try to connect to the VM from the dhcp server (Also LEAP 15.4), that works.

I’m a bit stumped as to where to look next for troubleshooting.

2

I would compare network packet captures om guest, host and the ping target. This may give some hint where and in which direction packets get lost.

Stab in the dark - it may be IP address conflict.

3

I’ve made progress. If I make 10 pings, I can see 20 icmp packets (request and reply) on the physical adapter its bridged to, but only 15 when I check the vtap adapter (10 requests, 5 replies.) Is there some sort of latency tolerance somewhere that’s dropping packets between the physical adapter and the virtual adapter?

It still doesn’t make sense, but at least I seem to have isolated where my issue lies.

4

UPDATE: I was bridging to a bond, if I destroy the bond, and just use a single port, everything works as expected.

So now I just need to know why it won’t distribute traffic properly across the bond.