Persistent Bad Gateway error between Traefik and Wordpress

MaggiWuerze · January 16, 2024, 8:38am

Hi, I've been searching this forum and the web for about a week now and finally decided to make a question of my own.

I am running a server with Traefik and various Docker containers for years now, and have several of them available through a domain handled and certified by Traefik.
Now I want to add wordpress to them but can't seem to get it running. I consistenly get a Bad Gateway error when trying to acces the interface through Traefik. When I access it via its local ip and port everything works fine. I've tried about every version of a docker wordpress stack I could find.

My Wordpress docker-compose

version: '3'

services:
   wordpress:
     container_name: wordpress
     depends_on:
       - db
     image: wordpress:latest
     volumes:
       - ./wp_data:/var/www/html
       - ./uploads.ini:/usr/local/etc/php/conf.d/uploads.ini
     restart: on-failure
     environment:
       WORDPRESS_DB_HOST: db:3306
       WORDPRESS_DB_USER: ${MYSQL_USER}
       WORDPRESS_DB_PASSWORD: ${MYSQL_USER_PASSWORD}
     ports:
       - 8889:80
     networks:
       - traefik
       - backend
     labels:
       - traefik.enable=true
       - traefik.http.routers.${PROJECT_NAME}.tls=true
       - traefik.http.routers.${PROJECT_NAME}.tls.certresolver=lets-encrypt
       - traefik.http.routers.${PROJECT_NAME}.entrypoints=websecure
       - traefik.http.routers.${PROJECT_NAME}.rule=Host(`${HOST_NAME}`)
       - traefik.http.services.${PROJECT_NAME}.loadbalancer.server.port=8889

   db:
     container_name: db-blog
     image: mariadb
     volumes:
       - ./db_data:/var/lib/mysql
     restart: on-failure
     networks:
       - backend
     environment:
       MYSQL_ROOT_PASSWORD: ${MYSQL_ROOT_PASSWORD}
       MYSQL_DATABASE: ${MYSQL_DATABASE}
       MYSQL_USER: ${MYSQL_USER}
       MYSQL_PASSWORD: ${MYSQL_USER_PASSWORD}

networks:
  backend:
    external: false
  traefik:
    external: true

My traefik docker-compose:

version: '3.5'
services:
    traefik:
        volumes:
            - '/var/run/docker.sock:/var/run/docker.sock'
            - '/home/daniel/docker/traefik/config/traefik.toml:/traefik.toml'
            - '/home/daniel/docker/traefik/traefik_dynamic.toml:/traefik_dynamic.toml'
            - '/home/daniel/docker/traefik/acme.json:/acme.json'
        ports:
            - '80:80'
            - '443:443'
            - '8081:8080'
        networks:
            - traefik
        container_name: traefik
        image: 'traefik:v2.10.7'
        restart: always

networks:
  traefik:
    name: traefik

traefik-dynamic:

[http.middlewares.simpleAuth.basicAuth]
  users = [
    "admin:password"
  ]

[http.routers.api]
  rule = "Host(`192.168.178.84`,`mediaserver`)"
  entrypoints = ["websecure"]
  middlewares = ["simpleAuth"]
  service = "api@internal"
  [http.routers.api.tls]
        certResolver = "lets-encrypt"

traefik.toml

defaultEntryPoints = ["http","https"]

[entryPoints]
  [entryPoints.web]
    address = ":80"
    [entryPoints.web.http.redirections.entryPoint]
      to = "websecure"
      scheme = "https"
  [entryPoints.websecure]
    address = ":443"

[api]
  dashboard = true
#  insecure = true

[certificatesResolvers.lets-encrypt.acme]
  email = "admin@example.de"
  storage = "acme.json"
  [certificatesResolvers.lets-encrypt.acme.tlsChallenge]

[providers.docker]
  exposedByDefault = false
  watch = true
  network = "traefik"

[providers.file]
  filename = "traefik_dynamic.toml"

wollomatic · January 16, 2024, 9:44am

Hi MaggiWuerze,

please try to remove

     ports:
       - 8889:80

and set the loadbalancer definition to

       - traefik.http.services.${PROJECT_NAME}.loadbalancer.server.port=80

Traefik uses the "internal" ports of the container, not the ports linked to the host system.

MaggiWuerze · January 16, 2024, 12:16pm

That actually did it. Now I'm confused, since I know for a fact hat I also had your version during my previous tests.

Thanks for the effort

MaggiWuerze · January 16, 2024, 1:43pm

Do you maybe also have an idea why Traefik can't get a certificate for the domain?

Timeout during connect (likely firewall problem)\n"

wollomatic · January 16, 2024, 6:50pm

After taking a short look, the TLS challenge looks good I think.
Try to set the log level to debug and maybe set the caServer to staging ( https://acme-staging-v02.api.letsencrypt.org/directory ) to not run into some rate limiting while testing

MaggiWuerze · January 17, 2024, 10:18am

I switched to LE staging, here's the log. It now says "Error getting validation data"

Sorry for the late answer, apparently the spam filter found me suspicious

traefik      | time="2024-01-16T19:07:04Z" level=debug msg="Trying to challenge certificate for domain [DOMAIN.de www.DOMAIN.de] found in HostSNI rule" rule="Host(`DOMAIN.de`, `www.DOMAIN.de`)" providerName=lets-encrypt-staging.acme ACME CA="https://acme-staging-v02.api.letsencrypt.org/directory" routerName=wordpress@docker
traefik      | time="2024-01-16T19:07:04Z" level=debug msg="Looking for provided certificate(s) to validate [\"DOMAIN.de\" \"www.DOMAIN.de\"]..." providerName=lets-encrypt-staging.acme ACME CA="https://acme-staging-v02.api.letsencrypt.org/directory" routerName=wordpress@docker rule="Host(`DOMAIN.de`, `www.DOMAIN.de`)"
traefik      | time="2024-01-16T19:07:04Z" level=debug msg="No ACME certificate generation required for domains [\"DOMAIN.de\" \"www.DOMAIN.de\"]." providerName=lets-encrypt-staging.acme ACME CA="https://acme-staging-v02.api.letsencrypt.org/directory" routerName=wordpress@docker rule="Host(`DOMAIN.de`, `www.DOMAIN.de`)"
traefik      | time="2024-01-16T19:07:04Z" level=debug msg="Loading ACME certificates 
traefik      | time="2024-01-16T19:07:05Z" level=debug msg="legolog: [INFO] Deactivating auth: https://acme-staging-v02.api.letsencrypt.org/acme/authz-v3/10649409914"
traefik      | time="2024-01-16T19:07:05Z" level=error msg="Unable to obtain ACME certificate for domains \"DOMAIN.de,www.DOMAIN.de\": unable to generate a certificate for the domains [DOMAIN.de www.DOMAIN.de]: error: one or more domains had a problem:\n[DOMAIN.de] acme: error: 400 :: urn:ietf:params:acme:error:connection :: 2003:d0:bf05:7d00:1dfb:e1ef:23f1:471a: Error getting validation data\n[www.DOMAIN.de] acme: error: 400 :: urn:ietf:params:acme:error:connection :: 2003:d0:bf05:7d00:1dfb:e1ef:23f1:471a: Error getting validation data\n" providerName=lets-encrypt-staging.acme ACME CA="https://acme-staging-v02.api.letsencrypt.org/directory" routerName=wordpress@docker rule="Host(`DOMAIN.de`, `www.DOMAIN.de`)"

MaggiWuerze · January 17, 2024, 3:09pm

Looking at the messages and output by https://letsdebug.net and https://check-your-website.server-daten.de/ it might be an issue with my ipv6 address. I deactivated the privacy extensions on my server and now wait for it to get a new ipv6 address that might work better with public access.

Any input is still welcome though

bluepuma77 · January 17, 2024, 8:29pm

Did you set an AAAA record in DNS for your IPv6?

Is your server and Docker setup IPv6 ready?

MaggiWuerze · January 18, 2024, 8:06am

The AAAA is there and has my servers ipv6 address. What exactly do you mean if docker is ipv6 ready? my server has an ipv6 address and my router allows ipv6 connections on 80 and 443 to my server.

My DNS Setup currently looks like this:

AAAA DOMAIN_A | <-- CNAME *.DOMAIN_A
              | <-- CNAME DOMAIN B <-- CAA
              | <-- CNAME *.DOMAIN_B
              | <-- CAA

Edit:

Ok, I enabled ipv6 for my docker setup now, following this guide: IPv6, Docker(-compose), and Shorewall6/ip6tables

Still no luck with the certificate. Still getting:

traefik      | time="2024-01-18T12:05:36Z" level=debug msg="legolog: [INFO] Deactivating auth: https://acme-staging-v02.api.letsencrypt.org/acme/authz-v3/10681545544"
traefik      | time="2024-01-18T12:05:36Z" level=error msg="Unable to obtain ACME certificate for domains \"DOMAIN.de,www.DOMAIN.de\": unable to generate a certificate for the domains [DOMAIN.de www.DOMAIN.de]: error: one or more domains had a problem:\n[DOMAIN.de] acme: error: 400 :: urn:ietf:params:acme:error:connection :: 2003:d0:bf1c:f800:8cd1:2b9:3e98:de0: Error getting validation data\n" ACME CA="https://acme-staging-v02.api.letsencrypt.org/directory" providerName=lets-encrypt-staging.acme routerName=wordpress@docker rule="Host(`DOMAIN.de`, `www.DOMAIN.de`)"

MaggiWuerze · January 18, 2024, 1:48pm

Ok, I now removed my AAAA record from my dynDNS provider and everything works again.

/thread I guess

system · January 21, 2024, 1:48pm

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.

Topic		Replies	Views
Bad Gateway for WordPress containers behind Traefik reverse proxy in docker-compose Traefik v2 docker	3	1923	September 27, 2022
Wordpress inside Docker with Traefik 2 Traefik v2 docker	1	2127	December 8, 2020
One more docker with identical config leads to Gateway tiemout for that docker Traefik v2 docker	10	989	May 19, 2022
Traefik container not reaching Wordpress as a service Traefik v2 docker , docker-swarm	3	272	March 4, 2024
Traefik 1.7x + Wordpress Woes (Full Logs Inside) [Traefik not working AT ALL on ONLY Wordpress container] Traefik v1 docker	0	942	January 18, 2020

Persistent Bad Gateway error between Traefik and Wordpress

Related Topics