Le 10/06/2016 18:34, Martin Vietz a écrit : > Hi Antonio, > > On 10.06.2016 12:58, Antonio Terceiro wrote: >> When it worked before, was it using NFS, or rsync? See >> https://github.com/vagrant-libvirt/vagrant-libvirt/issues/419 for a >> discussion on an issue where vagrant chooses a synced folder backend >> non-deterministically. > > It looks, like nfs: > > | vagrant@jessie:~$ mount | grep /vagrant > | 192.168.121.1:/tmp/tmp.FEnloL0pil on /vagrant type > nfs(rw,relatime,vers=3,rsize=32768,wsize=32768,namlen=255,hard,proto=udp,timeo=11,retrans=3,sec=sys,mountaddr=192.168.121.1,mountvers=3,mountport=20048,mountproto=udp,local_lock=none,addr=192.168.121.1) > | vagrant@jessie:~$ cat /etc/debian_version > | 8.4 > > A "vagrant destroy && vagrant up" had the same results. > >> Actually, I wanted to make 9p the default in the libvirt boxes, but it >> seems not all distros have 9p enabled in qemu and/or linux. > > I have tried a debian box with 9p on my fedora workstation a few months > ago. If I correct remember it works, but only unidirectional.
As long as a write in the host is synced is in the guest, it's fine for the standard vagrant uses. (ie I do dev on my host, but my programm is running in the guest) The rsync synced folder backend is also unidirectional.
