For a while I have had an issue when I restart docker containers and some (intermittent) will have bad gateway when trying to reach trhough traefik. I suspected it is having a docker networking issue. Still testing but I believe I sorted it out. I thought it was the containers that belonged to my external and internal docker networks. Like wordpress needing both external for traefik and internal for database. I think that was wrong afterall and it was any container on the external network. Traefik may be binding to the wrong docker network.
example from docker compose file
blog-1: image: "lscr.io/linuxserver/grav:latest" container_name: "blog-1" depends_on: - traefik restart: always labels: - "traefik.enable=true" - traefik.docker.network=external_network - "traefik.http.routers.blog-1.rule=Host(`blog.example.com`)" - "traefik.http.routers.blog-1.entrypoints=https" - "traefik.http.routers.blog-1.tls.certresolver=le-aws" - "traefik.http.routers.blog-1.tls=true" - "traefik.docker.network=traefik_external_network" volumes: - /stacks/html/blog.example.com/web:/config/ networks: - external_network
networks: external_network: external: false internal_network: internal: falseWARNING: above label we needed the stack name traefik_ not just external_network
docker networks
ubuntu@server01:/stacks/stack01$ docker network lsNETWORK ID NAME DRIVER SCOPE3d7ad1df26f9 bridge bridge localbdb6efaa181b host host localbe6e03dc6ed5 none null local76207085707c traefik_default bridge local2a334976cbb7 traefik_external_network bridge localf1abbe2c4374 traefik_internal_network bridge local