aio-nextcloud-talk cannot start if ipv6 isn't present

[stro]

Steps to reproduce

1. Disable ipv6 on the host machine, for example, by setting net.ipv6.conf.all.disable_ipv6 = 1 in /etc/sysctl.conf or running sysctl -w net.ipv6.conf.all.disable_ipv6=0
2. Start AIO containers.

Expected behavior

aio-nextcloud-talk starts if it's selected.

Actual behavior

aio-nextcloud-talk fails to start.

Here's the partial log records:

2024-10-05T22:41:55.482339594Z + IPv6_ADDRESS_TALK=

2024-10-05T22:41:56.869331065Z [ERR] [plugins/janus_streaming.c:janus_streaming_create_fd:6000] [rtp-sample] Cannot create socket for audio... 97 (Address family not supported by protocol)
2024-10-05T22:41:56.869360969Z [ERR] [plugins/janus_streaming.c:janus_streaming_create_rtp_source:6231] Can't bind to port 5002 for audio...
2024-10-05T22:41:56.869365238Z [ERR] [plugins/janus_streaming.c:janus_streaming_init:2042] Error creating 'rtp' mountpoint 'rtp-sample'...

2024-10-05T22:41:58.232803096Z 2024-10-05 22:41:58.187771+00:00 [error] CRASH REPORT:
2024-10-05T22:41:58.232854809Z crasher:
2024-10-05T22:41:58.232859476Z initial call: stun_acceptor:init/4
2024-10-05T22:41:58.232863211Z pid: <0.600.0>
2024-10-05T22:41:58.232867166Z registered_name: []
2024-10-05T22:41:58.232870790Z exception exit: eafnosupport
2024-10-05T22:41:58.232874864Z in function stun_acceptor:init/4 (stun_acceptor.erl, line 108)
2024-10-05T22:41:58.232878840Z ancestors: [stun_acceptor_sup,stun_listener_sup,stun_sup,<0.580.0>]
2024-10-05T22:41:58.232882754Z message_queue_len: 0
2024-10-05T22:41:58.232894105Z messages: []
2024-10-05T22:41:58.232897873Z links: [<0.583.0>]
2024-10-05T22:41:58.232901541Z dictionary: []
2024-10-05T22:41:58.232905130Z trap_exit: false
2024-10-05T22:41:58.232908684Z status: running
2024-10-05T22:41:58.232912261Z heap_size: 376
2024-10-05T22:41:58.232915790Z stack_size: 28
2024-10-05T22:41:58.232919326Z reductions: 676
2024-10-05T22:41:58.232922963Z neighbours:
2024-10-05T22:41:58.232926502Z
2024-10-05T22:41:58.233237859Z 2024-10-05 22:41:58.188129+00:00 [critical] Aborting: Cannot start listening on [::]:3478 (udp): address family not supported by protocol family

Other information

Host OS

Linux ***** 6.1.0-23-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.99-1 (2024-07-15) x86_64 GNU/Linux

Output of sudo docker info

Client: Docker Engine - Community
Version: 27.2.1
Context: default
Debug Mode: false
Plugins:
buildx: Docker Buildx (Docker Inc.)
Version: v0.16.2
Path: /usr/libexec/docker/cli-plugins/docker-buildx
compose: Docker Compose (Docker Inc.)
Version: v2.29.2
Path: /usr/libexec/docker/cli-plugins/docker-compose

Server:
Containers: 22
Running: 9
Paused: 0
Stopped: 13
Images: 16
Server Version: 27.2.1
Storage Driver: overlay2
Backing Filesystem: extfs
Supports d_type: true
Using metacopy: false
Native Overlay Diff: true
userxattr: false
Logging Driver: json-file
Cgroup Driver: systemd
Cgroup Version: 2
Plugins:
Volume: local
Network: bridge host ipvlan macvlan null overlay
Log: awslogs fluentd gcplogs gelf journald json-file local splunk syslog
Swarm: inactive
Runtimes: io.containerd.runc.v2 runc
Default Runtime: runc
Init Binary: docker-init
containerd version: 7f7fdf5fed64eb6a7caf99b3e12efcf9d60e311c
runc version: v1.1.14-0-g2c9f560
init version: de40ad0
Security Options:
apparmor
seccomp
Profile: builtin
cgroupns
Kernel Version: 6.1.0-23-amd64
Operating System: Debian GNU/Linux 12 (bookworm)
OSType: linux
Architecture: x86_64
CPUs: 2
Total Memory: 31.15GiB
Name: *******
ID: *******
Docker Root Dir: /var/lib/docker
Debug Mode: false
Experimental: false
Insecure Registries:
127.0.0.0/8
Live Restore Enabled: false

Docker run command or docker-compose file that you used

Other valuable info

It seems that the problem is in configuration file, which checks for ipv4 and acts accordingly but fails to check for ipv6. In the end, here's the result of the configuration:

ea515e75202e:/$ cat /conf/eturnal.yml
eturnal:
listen:
- ip: "::"
port: 3478
transport: udp
- ip: "::"
port: 3478
transport: tcp
log_dir: stdout
log_level: warning
secret: "*******"
relay_ipv4_addr: "172.18.0.2"
blacklist_peers:

• recommended
  whitelist_peers:
• 127.0.0.1
• ::1
• "172.18.0.2"

Fix

If configuration file is modified to replace "::" with "0.0.0.0" and "- ::1" is removed, the container starts correctly. I didn't test the full functionality (see janus_streaming errors above), but I was able to login from a phone app.

It looks like https://github.com/nextcloud/all-in-one/blob/main/Containers/talk/start.sh needs to be corrected to check IPv6_ADDRESS_TALK and adjust the configuration correspondingly, or other methods to detect ipv6 presence to be used, like running "ip a|grep ipv6" or "test -f /proc/net/if_inet6"

[szaimen]

Hi, as far as I know do also other containers have problems if you disable ipv6 in the kernel. Since most containers are based on official containers, you could try to fix these issues upstream as we do not have any intention to hack them. A workaround for you might be to enable ipv6 in the kernel and simply don't enable ipv6 in docker which results in the same but actually works.

[stro]

All non-optional containers work just fine without ipv6 (maybe some optional too). The problem, like I said, in the incorrect configuration. The talk container works too if configuration is updated.

[szaimen]

@Zoey2936 wdyt about this change?

[Zoey2936]

Could be done, but depends on the question if we want to support ipv6 disabled setups

[szaimen]

depends on the question if we want to support ipv6 disabled setups

Yes good question.