LXD bridge networking stop working after upgrade to Fedora 32

LXD
1

Hi,

On Fedora 31 I was using cgroup v1 with systemd.unified_cgroup_hierarchy=0 but after upgrade to Fedora 32 this not help anymore.
The problem is that none of containers doesn’t have ipv4 assigned.
I am using default managed lxd bridge with nat.
Any solution?

2

Can you check firewalld isnt causing this issue: https://github.com/lxc/lxd/blob/master/doc/networks.md#allow-dhcp-dns-with-firewalld

1 Like
3

Thank you,
It is working now.

4

I have exactly the same issue (also caused by upgrading to fedora 32). I have followed the instructions and my Centos 8 containers got ipv4. However, I also have 2 Debian 8 containers which remain without ipv4 addresses…

System logs

May 15 22:26:59 dell dnsmasq[36679]: Maximum number of concurrent DNS queries reached (max: 150)
May 15 22:27:01 dell NetworkManager[1239]: <info>  [1589574421.0741] manager: (veth640001bc): new Veth device (/org/freedesktop/NetworkManager/Devices/89)
May 15 22:27:01 dell NetworkManager[1239]: <info>  [1589574421.0760] manager: (vetha7b916ec): new Veth device (/org/freedesktop/NetworkManager/Devices/90)
May 15 22:27:01 dell kernel: lxdbr0: port 8(vetha7b916ec) entered blocking state
May 15 22:27:01 dell kernel: lxdbr0: port 8(vetha7b916ec) entered disabled state
May 15 22:27:01 dell kernel: device vetha7b916ec entered promiscuous mode
May 15 22:27:01 dell audit: ANOM_PROMISCUOUS dev=vetha7b916ec prom=256 old_prom=0 auid=4294967295 uid=0 gid=0 ses=4294967295
May 15 22:27:01 dell systemd-udevd[77724]: ethtool: autonegotiation is unset or enabled, the speed and duplex are not writable.
May 15 22:27:01 dell systemd-udevd[77725]: ethtool: autonegotiation is unset or enabled, the speed and duplex are not writable.
May 15 22:27:01 dell systemd-udevd[77724]: Using default interface naming scheme 'v245'.
May 15 22:27:01 dell systemd-udevd[77725]: Using default interface naming scheme 'v245'.
May 15 22:27:01 dell kernel: eth0: renamed from veth640001bc
May 15 22:27:01 dell kernel: IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
May 15 22:27:01 dell kernel: lxdbr0: port 8(vetha7b916ec) entered blocking state
May 15 22:27:01 dell kernel: lxdbr0: port 8(vetha7b916ec) entered forwarding state
May 15 22:27:01 dell gnome-shell[2378]: Removing a network device that was not added
May 15 22:27:01 dell NetworkManager[1239]: <info>  [1589574421.1978] device (vetha7b916ec): carrier: link connected

Freshly created Debian 8 containers seem to able to get ipv4 addresses without any problems.

Any tips on reviving network in existing Debian 8 containers? :slight_smile: