I run multiple docker swarm instances on NixOS but I never got this kind of error.
Its seems related to your network config, on my side I only used network manager or static interfaces with networking.interfaces. Maybe the IP you use when you run docker swarm init --advertise-addr <IP> has an impact.
I’m a beginner with both NixOS and Docker so I might be missing something obvious.
This is a new VPS with a NixOS 24.11 install. Could I break down the steps somehow to know what exactly fails? Do I need to add any specific networking configuration?
Docker Registry seems to be running fine at the same time:
services.dockerRegistry.enabled = true
I’ve tried
docker swarm leave -f
docker prune -a
docker swarm init # no advertise-addr
virtualisation.docker = {
enable = true;
autoPrune.enable = true;
autoPrune.flags = [ "-af" "--filter='label!=avoid-prune'" ]; # this label filter is for a special use case
liveRestore = false; # this should be put for swarm but I don't think it's related to your issue
};