Using a single Traefik container for RP across multiple stacks/networks -> 502 Bad Gateway

Traefik Traefik v2
,
1

Here is my big ugly docker compose yaml:
(first one of 7 that I am trying this on)

version: '3.9'

services:
  web:
    image: perseushub/web:latest
    build: ./nginx
    container_name: web
    expose:
      - "80"
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.web.rule=Host(`perseus.acumenus.net`)"
      - "traefik.http.routers.web.entrypoints=websecure"
      - "traefik.http.routers.web.tls.certresolver=myresolver"
      - "traefik.http.services.web.loadbalancer.server.port=80"

  shareddb:
    image: perseushub/shareddb:latest
    build: ./shared-db
    container_name: shareddb
    volumes:
      - shareddb:/data/postgres
    expose:
      - "5432"
    labels:
      - "traefik.enable=true"

  files-manager:
    image: perseushub/files-manager:latest
    build: ./files-manager
    container_name: files-manager
    expose:
      - "10500"
    environment:
      - SPRING_PROFILES_ACTIVE=docker
    depends_on:
      - shareddb
    labels:
      - "traefik.enable=true"

  user:
    image: perseushub/user:latest
    build: ./user
    container_name: user
    environment:
      USER_ENV: Docker
    env_file:
      - user/user-envs.txt
    expose:
      - "5001"
    depends_on:
      - shareddb
    labels:
      - "traefik.enable=true"

  backend:
    image: perseushub/backend:latest
    build: ./perseus-api
    container_name: backend
    environment:
      PERSEUS_ENV: Docker
    expose:
      - "5004"
    depends_on:
      - shareddb
      - files-manager
    labels:
      - "traefik.enable=true"

  frontend:
    image: perseushub/frontend:latest
    build:
      context: ./UI
      args:
        env: prod
    container_name:
      frontend
    expose:
      - "4200"
    labels:
      - "traefik.enable=true"

  white-rabbit:
    image: perseushub/white-rabbit:latest
    build: ../WhiteRabbit
    container_name:
      white-rabbit
    expose:
      - "8000"
    environment:
      - SPRING_PROFILES_ACTIVE=docker
    depends_on:
      - shareddb
      - files-manager
    labels:
      - "traefik.enable=true"

  vocabularydb:
    image: perseushub/vocabularydb:latest
    build: ./vocabulary-db
    container_name: vocabularydb
    healthcheck:
      test: [ "CMD", "pg_isready", "-q", "-d", "vocabulary", "-U", "admin" ]
      timeout: 60s
      interval: 30s
      retries: 10
    volumes:
      - vocabularydb:/data/postgres
    ports:
      - "5431:5432"
    labels:
      - "traefik.enable=true"

  cdm-builder:
    image: perseushub/cdm-builder:latest
    build: ../ETL-CDMBuilder
    container_name:
      cdm-builder
    expose:
      - "9000"
    environment:
      - ASPNETCORE_ENVIRONMENT=Docker
    depends_on:
      - shareddb
      - files-manager
      - vocabularydb
    labels:
      - "traefik.enable=true"

  solr:
    image: perseushub/solr:latest
    build: ./solr
    container_name: solr
    expose:
      - "8983"
    volumes:
      - solr:/var/solr
    depends_on:
      - vocabularydb
    labels:
      - "traefik.enable=true"

  athena:
    image: perseushub/athena:latest
    build: ./athena-api
    container_name: athena
    environment:
      - ATHENA_ENV=Docker
    expose:
      - "5002"
    depends_on:
      - solr
    labels:
      - "traefik.enable=true"

  usagi:
    image: perseushub/usagi:latest
    build: ./usagi-api
    command: python /app/main.py
    container_name: usagi
    environment:
      USAGI_ENV: Docker
    expose:
      - "5003"
    depends_on:
      - shareddb
      - solr
    labels:
      - "traefik.enable=true"

  r-serve:
    image: perseushub/r-serve:latest
    build:
      context: ../DataQualityDashboard/R
      args:
        prop: docker
    container_name:
      r-serve
    expose:
      - "6311"
    depends_on:
      - shareddb
    labels:
      - "traefik.enable=true"

  data-quality-dashboard:
    image: perseushub/data-quality-dashboard:latest
    build:
      context: ../DataQualityDashboard
    container_name:
      data-quality-dashboard
    expose:
      - "8001"
    environment:
      - SPRING_PROFILES_ACTIVE=docker
    depends_on:
      - shareddb
      - files-manager
      - r-serve
    labels:
      - "traefik.enable=true"

  swagger:
    image: perseushub/swagger:latest
    build: ./swagger-ui
    container_name: swagger
    ports:
      - "8081:8080"
    labels:
      - "traefik.enable=true"

networks:
    perseus_default:
    traefik-proxy:

volumes:
  shareddb:
  vocabularydb:
  solr:

Here is the docker compose up output:

web                     | 2023/12/01 18:13:05 [error] 19#19: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 172.26.0.17, server: _, request: "GET /favicon.ico HTTP/1.1", upstream: "http://172.17.0.1:4200/favicon.ico", host: "perseus.acumenus.net", referrer: "https://perseus.acumenus.net/"
web                     | 172.26.0.17 - - [01/Dec/2023:18:13:05 +0000] "GET /favicon.ico HTTP/1.1" 502 559 "https://perseus.acumenus.net/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36" "73.175.74.146"
web                     | 172.26.0.17 - - [01/Dec/2023:18:13:05 +0000] "GET / HTTP/1.1" 502 559 "-" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36" "73.175.74.146"
web                     | 2023/12/01 18:13:05 [error] 19#19: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 172.26.0.17, server: _, request: "GET / HTTP/1.1", upstream: "http://172.17.0.1:4200/", host: "perseus.acumenus.net"
web                     | 2023/12/01 18:13:05 [error] 19#19: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 172.26.0.17, server: _, request: "GET /favicon.ico HTTP/1.1", upstream: "http://172.17.0.1:4200/favicon.ico", host: "perseus.acumenus.net", referrer: "https://perseus.acumenus.net/"
web                     | 172.26.0.17 - - [01/Dec/2023:18:13:05 +0000] "GET /favicon.ico HTTP/1.1" 502 559 "https://perseus.acumenus.net/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36" "73.175.74.146"
web                     | 2023/12/01 18:13:05 [error] 19#19: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 172.26.0.17, server: _, request: "GET / HTTP/1.1", upstream: "http://172.17.0.1:4200/", host: "perseus.acumenus.net"
web                     | 172.26.0.17 - - [01/Dec/2023:18:13:05 +0000] "GET / HTTP/1.1" 502 559 "-" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36" "73.175.74.146"
web                     | 2023/12/01 18:13:05 [error] 19#19: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 172.26.0.17, server: _, request: "GET /favicon.ico HTTP/1.1", upstream: "http://172.17.0.1:4200/favicon.ico", host: "perseus.acumenus.net", referrer: "https://perseus.acumenus.net/"
web                     | 172.26.0.17 - - [01/Dec/2023:18:13:05 +0000] "GET /favicon.ico HTTP/1.1" 502 559 "https://perseus.acumenus.net/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36" "73.175.74.146"
backend                 | [2023-12-01 18:17:32,436] INFO: Running job "clear_cache (trigger: interval[0:15:00], next run at: 2023-12-01 18:17:32 UTC)" (scheduled at 2023-12-01 18:17:32.435476+00:00)
backend                 | [2023-12-01 18:17:32,436] INFO: Job "clear_cache (trigger: interval[0:15:00], next run at: 2023-12-01 18:17:32 UTC)" executed successfully

And here the dashboard screenshots of the service in question:

2023-12-01_13-27
2023-12-01_13-271664×1148 75.9 KB

2023-12-01_13-20
2023-12-01_13-201709×1046 67 KB

As the title of this post suggests, this is an attempt to consolidate RP across multiple stacks. This is being built for a non-profit consortium of medical outcomes researchers (about 2000+ around the world) as an application ecosystem demonstrator.

parthenon
parthenon1418×882 30.5 KB

I have made significant progress on this, but tinkering will only get me so far. So any help and guidance would be greatly appreciated.

2

Share your full Traefik static and dynamic config, and docker-compose.yml if used.

3

The main docker-compose.yml:

version: '3.8'

services:
  traefik:
    image: "docker.io/library/traefik:v2.10.5"
    container_name: "traefik2"
    restart: "always"
    command:
      - "--api.insecure=false"
      - "--providers.docker=true"
      - "--entrypoints.web.address=:80"
      - "--entrypoints.websecure.address=:443"
      - "--certificatesresolvers.myresolver.acme.email=******@mac.com"
      - "--certificatesresolvers.myresolver.acme.storage=/letsencrypt/acme.json"
      - "--certificatesresolvers.myresolver.acme.httpchallenge=true"
      - "--certificatesresolvers.myresolver.acme.httpchallenge.entrypoint=web"
    ports:
      - "80:80"
      - "443:443"
    volumes:
      - "/letsencrypt:/letsencrypt"
      - "/var/run/docker.sock:/var/run/docker.sock:ro"
      - "/portainer/Files/AppData/Config/traefik/config.yml:/etc/traefik/config.yml:ro"
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.traefik-dashboard.rule=Host(`traefik.acumenus.net`)"
      - "traefik.http.routers.traefik-dashboard.entrypoints=websecure"
      - "traefik.http.routers.traefik-dashboard.service=api@internal"
      - "traefik.http.routers.traefik-dashboard.tls.certresolver=myresolver"
      - "traefik.http.routers.traefik-dashboard.middlewares=dashboard-auth"
      - "traefik.http.middlewares.dashboard-auth.basicauth.users=acumenus:*********"

    networks:
      - traefik-proxy

  heimdall:
    image: "linuxserver/heimdall:latest"
    container_name: "heimdall"
    restart: "always"
    environment:
      - PUID=1000
      - PGID=1000
      - TZ=America/New_York # Replace with your timezone
      - APP_NAME=Parthenon
    volumes:
      - "/portainer/Files/AppData/Config/Heimdall:/config"
      - "/portainer/Files/AppData/Config/Heimdall/favicons/favicon-16x16.png:/var/www/localhost/heimdall/public/favicon-16x16.png"
      - "/portainer/Files/AppData/Config/Heimdall/favicons/favicon-32x32.png:/var/www/localhost/heimdall/public/favicon-32x32.png"
      - "/portainer/Files/AppData/Config/Heimdall/favicons/favicon.ico:/var/www/localhost/heimdall/public/favicon.ico"
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.heimdall.rule=Host(`dev.acumenus.net`)"
      - "traefik.http.routers.heimdall.entrypoints=websecure"
      - "traefik.http.routers.heimdall.tls.certresolver=myresolver"
      - "traefik.http.routers.heimdall.middlewares=redirect-to-https"
      - "traefik.http.middlewares.redirect-to-https.redirectscheme.scheme=https"
      - "traefik.http.middlewares.redirect-to-https.redirectscheme.permanent=true"

      - "traefik.http.services.heimdall.loadbalancer.server.port=80"
    networks:
      - traefik-proxy

networks:
  traefik-proxy:

    external: true
    driver: overlay
    attachable: true

This is launched as a portainer stack (as I am testing and have only been in the docker/traefik universe for about 4 months).

I don't have a static or dynamic config, but am open to any/all suggestions!

4

Your other services are not connected to Docker network traefik-proxy.

1 Like
5

Ah. I didn't realize that each service needs to have:

networks:
  - traefik-proxy

I thought that was implicit if the networks were named at the bottom of the docker-compose.yml, and this was on every service:

labels:
  - "traefik.enable=true"

So in the perseus stack I will add networks to all services and report back. Ty.

6

2023-12-02_00-34
2023-12-02_00-341610×1195 283 KB

After adding "traefik-proxy" network to the networks: in the docker-compose.yml (see above), the page loads the favicon (which indicates communication has been established) but then nothing else.

There are a bunch of "Connection Refused" messages etc. I believe there is some confusion regarding the nginx container port, the frontend port and the swagger port. I'm going to reach out to the original developers and try and understand this stack a bit further.

However, from the GitHub, I found this, and I assume it means that nginx is being used as a reverse proxy for all these services in the stack:

2023-12-03_02-23
2023-12-03_02-231179×1358 163 KB

I wonder if it is possible to just replace nginx with traefik as the RP?