Nfs mount 'gone' after vm's start

I have had a nfs-ganesha mount provision install images for vm's, and recently I have the mount stall after starting a vm. (even when it does not access an iso image on the nfs mount) I have the host ip on a macvtap of the connected interface. The vm is also having an macvtap on the same interface to allow host-guest communication. I am mounting the nfs with these options. nfs4 ro,vers=4,proto=tcp,soft,nodev,_netdev,noatime,timeo=30,x-systemd.device -timeout=30,x-systemd.mount-timeout=30,x-systemd.automount,x-systemd.idl e-timeout=30 0 0 The problem is fixed by doing a simple 'umount -l' Anyone seen something similar? I suspect this could be related by the macvtap being created, but I cannot remember seeing this mount fail before.