Containers not getting IP address after server reboot after switch to core20

LXD
,
1

On Ubuntu 20.04 snap LXD 4.15 latest/stable. Rebooted server after upgrade to Ubuntu 20.04 - and now container not getting IP address. lxdbr0 is up and running. Containers are using default profile with eth0, nictype bridged, parent lxdbr0 in their configuration.

Read many posts on forum - and tried troubleshooting on my own - some details below:

3: lxdbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
    link/ether 00:16:3e:c3:76:a8 brd ff:ff:ff:ff:ff:ff
    inet 10.139.30.1/24 scope global lxdbr0
       valid_lft forever preferred_lft forever
    inet6 fe80::216:3eff:fec3:76a8/64 scope link 
       valid_lft forever preferred_lft forever

grep lxdbr0 /var/log/syslog

Jun 17 20:27:37 sysk1 systemd-udevd[378]: lxdbr0: Could not generate persistent MAC: No data available
Jun 17 20:27:38 sysk1 kernel: [   14.414603] audit: type=1400 audit(1623958058.017:28): apparmor="STATUS" operation="profile_load" profile="unconfined" name="lxd_dnsmasq-lxdbr0_</var/snap/lxd/common/lxd>" pid=1488 comm="apparmor_parser"
Jun 17 20:27:38 sysk1 lxd.daemon[1419]: t=2021-06-17T20:27:38+0100 lvl=warn msg="Skipping AppArmor for dnsmasq due to raw.dnsmasq being set" driver=bridge name=lxdbr0 network=lxdbr0 project=default
Jun 17 20:27:39 sysk1 kernel: [   15.615735] lxdbr0: port 1(veth1d393f16) entered blocking state
Jun 17 20:27:39 sysk1 kernel: [   15.615738] lxdbr0: port 1(veth1d393f16) entered disabled state
Jun 17 20:27:39 sysk1 kernel: [   16.120292] lxdbr0: port 1(veth1d393f16) entered blocking state
Jun 17 20:27:39 sysk1 kernel: [   16.120295] lxdbr0: port 1(veth1d393f16) entered forwarding state
Jun 17 20:27:39 sysk1 kernel: [   16.120347] IPv6: ADDRCONF(NETDEV_CHANGE): lxdbr0: link becomes ready

sudo lsof -n -i :53 - there seems to be no dnsmasq here. Unclear if I need to install dnsmasq on host or lxd snap includes own dnsmasq packages. Apologies - not a developer by training - learning this for my small business.

COMMAND   PID            USER   FD   TYPE DEVICE SIZE/OFF NODE NAME
systemd-r 961 systemd-resolve   12u  IPv4  19009      0t0  UDP 127.0.0.53:domain 
systemd-r 961 systemd-resolve   13u  IPv4  19010      0t0  TCP 127.0.0.53:domain (LISTEN)

Appreciate any guidance on how to resolve… Thanks.

2

Please check if you’ve been affected by Containers suddenly stopped working since move to core20 snap - No more IP's assigned - #26 by tomp

3

Thank you for the reference. I did comment out the “auth-zone” from raw.dnsmasq using lxc network edit lxdbr0 . Rebooted the server - containers not getting IP address.

lxc network show lxdbr0 - had that line before now does not show up… Is there any other setting I need to look at - do I need to install dnsmasq on the host? Is there a conflict with systemd-resolve listening on same port 53?

sudo ss -ulpn

State      Recv-Q      Send-Q            Local Address:Port           Peer Address:Port     Process                                        
UNCONN     0           0                   10.139.30.1:53                  0.0.0.0:*         users:(("dnsmasq",pid=1537,fd=6))             
UNCONN     0           0                 127.0.0.53%lo:53                  0.0.0.0:*         users:(("systemd-resolve",pid=977,fd=12))     
UNCONN     0           0                0.0.0.0%lxdbr0:67                  0.0.0.0:*         users:(("dnsmasq",pid=1537,fd=4))
4

I’m not sure commenting will be respected by dnsmasq best remove it entirely until you can see the dnsmasq process running with ‘sudo ps aux | grep dnsmasq’

5

Oh actually seems fine (I missed it at first as on my phone as evening here), in that case its probably this issue now: LXD stopped generating firewall rules after switch to core20

6

If you use ufw firewall this is a work around in the meantime Lxd bridge doesn't work with IPv4 and UFW with nftables - #17 by tomp