Allow enabling experimental ip6tables option with Docker 20.10

[Strayer]

Docker recently gained the option to also set IPv6 NAT rules to forward IPv6 traffic to containers: https://github.com/moby/moby/pull/41622

 

It is relatively easy to enable, but requires enabling experimental features. As a quick test, I added this daemon.json to a Debian VM that has Docker 20.10 installed:

 

{
  "ipv6": true,
  "ip6tables": true,
  "experimental": true,
  "fixed-cidr-v6": "fd00:dead:beef::/48"
}

 

Now Docker will automatically create IPv6 NAT rules to forward ports to containers:

 

root@debian:~# docker run --rm -p 80:80 nginx
[...]
fd15:4ba5:5a2b:1008:c87b:a4ed:6798:4006 - - [11/Apr/2021:12:08:29 +0000] "GET / HTTP/1.1" 200 612 "-" "curl/7.64.1" "-"

 

I'm using IPv6 a lot and would love it to be able to forward IPv6 traffic with this option. Not sure when Unraid would be able to update to 20.10, as 6.9 seems to be using 19.03 for now, but I wanted to add this feature request anyway.

[bonienl]

Unraid 6.10 will have IPv6 support for the docker bridge network.

It uses a private network range and NAT to access the outside world, similar to the IPv4 set up of docker0.

Below an example of a container with a bridge network and IPv6 enabled.

 

/ # ip -6 route
fd17::/64 dev eth0  metric 256 
fe80::/64 dev eth0  metric 256 
multicast ff00::/8 dev eth0  metric 256 
default via fd17::1 dev eth0  metric 1024 
/ # ping -6 google.com
PING google.com (2a00:1450:4009:815::200e): 56 data bytes
64 bytes from 2a00:1450:4009:815::200e: seq=0 ttl=115 time=10.079 ms
64 bytes from 2a00:1450:4009:815::200e: seq=1 ttl=115 time=10.037 ms
64 bytes from 2a00:1450:4009:815::200e: seq=2 ttl=115 time=10.087 ms
^C
--- google.com ping statistics ---
3 packets transmitted, 3 packets received, 0% packet loss
round-trip min/avg/max = 10.037/10.067/10.087 ms