I’m unsure what I did, if anything, but from one day to the next my lxd container won’t boot as long as my routed network for IPv6 is in my container configuration.
I have a config that specifies a eth device for IPv6 traffic.
eth1:
ipv6.address: 2a0a:7000:1337:c:d282:4784:6ad9:7f9d,2a0a:7000:1337:c:be54:955a:33b:cdc2
nictype: routed
parent: enp216s0
type: nic
I restarted my container today after having it run for a week or two (I did restart it before) and all of a sudden it won’t start. It will start if I remove this configuration.
My lxc logs that seem relevant.
lxc tachikoma 20200526112458.397 DEBUG conf - conf.c:lxc_map_ids:2710 - Functional newuidmap and newgidmap binary found
lxc tachikoma 20200526112458.402 TRACE conf - conf.c:lxc_map_ids:2780 - newuidmap wrote mapping “newuidmap 52923 65536 0 1 0 100000 65536”
lxc tachikoma 20200526112458.407 TRACE conf - conf.c:lxc_map_ids:2780 - newgidmap wrote mapping “newgidmap 52923 65536 0 1 0 100000 65536”
lxc tachikoma 20200526112458.407 TRACE conf - conf.c:run_userns_fn:3857 - Calling function “chown_cgroup_wrapper”
lxc tachikoma 20200526112458.407 NOTICE utils - utils.c:lxc_setgroups:1366 - Dropped additional groups
lxc tachikoma 20200526112458.408 WARN cgfsng - cgroups/cgfsng.c:fchowmodat:1455 - No such file or directory - Failed to fchownat(17, memory.oom.group, 65536, 0, AT_EMPTY_PATH | AT_SYMLINK_NOFOLLOW )
lxc tachikoma 20200526112458.409 DEBUG start - start.c:lxc_spawn:1737 - Preserved net namespace via fd 4
lxc tachikoma 20200526112458.409 TRACE start - start.c:lxc_spawn:1744 - Allocated new network namespace id
lxc tachikoma 20200526112458.411 DEBUG network - network.c:instantiate_phys:828 - Instantiated phys “veth11bb0ca8” with ifindex is “136”
lxc tachikoma 20200526112458.412 INFO network - network.c:instantiate_veth:290 - Retrieved mtu 1500 from enp216s0
lxc tachikoma 20200526112458.413 ERROR network - network.c:setup_ipv6_addr_routes:179 - Unknown error -17 - Failed to setup ipv6 address route for network device with eifindex 138
lxc tachikoma 20200526112458.413 ERROR network - network.c:instantiate_veth:430 - Unknown error -17 - Failed to setup ip address routes for network device “vethf11de550”
lxc tachikoma 20200526112458.458 ERROR network - network.c:lxc_create_network_priv:3096 - Unknown error -17 - Failed to create network device
lxc tachikoma 20200526112458.458 ERROR start - start.c:lxc_spawn:1750 - Failed to create the network
lxc tachikoma 20200526112458.460 TRACE network - network.c:lxc_delete_network_priv:3218 - Renamed interface with index 0 from “eth0” to its initial name “veth11bb0ca8”
lxc tachikoma 20200526112458.460 TRACE network - network.c:lxc_delete_network_priv:3227 - Restored interface “veth11bb0ca8” to its initial mtu “1450”
lxc tachikoma 20200526112458.460 DEBUG network - network.c:lxc_delete_network:3693 - Deleted network devices
lxc tachikoma 20200526112458.460 TRACE start - start.c:lxc_serve_state_socket_pair:492 - Sent container state “ABORTING” to 7
lxc tachikoma 20200526112458.460 TRACE start - start.c:lxc_serve_state_clients:427 - Set container state to ABORTING
lxc tachikoma 20200526112458.460 TRACE start - start.c:lxc_serve_state_clients:430 - No state clients registered
lxc tachikoma 20200526112458.460 ERROR lxccontainer - lxccontainer.c:wait_on_daemonized_start:852 - Received container state “ABORTING” instead of “RUNNING”
Does anybody know what could have changed on the server that prevents this from working all of a sudden? The bridged ethernet device still does exist on the host and has the same name. Before I restarted the container IPv6 was working fine.
This is on 4.0.0.