Docker: no processes, no containers, for every port: “bind: address already in use”

Question

I get the

docker: Error response from daemon: driver failed programming external connectivity on endpoint beautiful_jang (..): Error starting userland proxy: listen tcp [::]:53007: bind: address already in use.

error for

• any container image
• any port
• with no other docker daemons running
• with no processes listening on any port (apart from mdns and systemd-resolved)
• with iptables -F(lushed) and docker daemon running with or without ipv6 support
• with no meaningful error message in docker logs, even with debug mode

I’m able to start processes that bind to ports nc -l 53007.

I’m running Clear Linux (latest update 39540) on a simple system that has been running docker for the last 5 years.

Here are running systemd services:

(removed because it triggered a spam filter)

I’m really out of ideas here..

Answer 1

The solution was to use 127.0.0.1:53007:53007 instead of 53007:53007 for port definitions. I was told its related to an upgrade the new docker version 24.0.2.