Mattermost Error in Self Hosted using Docker + Nginx + Cloudflare

1

Hello.
I’ve used Mattermost docker container for my web app as chat system.
My web application is hosted on EC2 server and this is my docker compose file:

version: '3.9'

services:
  br-db:
    container_name: br-db
    platform: linux/amd64
    image: postgres:15.2-alpine
    ports:
      - "54322:5432"
    restart: always
    env_file: './pgsql/local/.env.local'
    volumes:
      - br-pgsql-data:/var/lib/postgresql/data
    networks:
      - br-net

  br-app:
    container_name: br-app
    build:
      context: ./be
      args:
        - DEV=true
    tty: true
    restart: unless-stopped
    working_dir: /app
    env_file: './be/.env.local'
    ports:
      - "8001:8000"
    environment:
      CONTAINER_ROLE: app
      CONTAINER_ENV: local
    volumes:
      - ./be/source:/app
    depends_on:
      - br-db
    networks:
      - br-net

  br-redis:
    image: redis:6.2-alpine
    container_name: br-redis
    ports:
      - "6379:6379"
    restart: always
    volumes:
      - redisdata:/data
    networks:
      - br-net

  br-celery:
    container_name: br-celery
    build:
      context: ./be
      args:
        - DEV=true
    command: celery -A app worker -l info
    tty: true
    restart: unless-stopped
    working_dir: /app
    env_file: './be/.env.local'
    volumes:
      - ./be/source:/app
    depends_on:
      - br-db
      - br-redis
    networks:
      - br-net

  mattermost:
    platform: linux/amd64
    image: mattermost/mattermost-team-edition:latest
    container_name: mattermost
    ports:
      - "8065:8065"
      - "8067:8067"
    volumes:
      - ./mattermost/local/config:/mattermost/config:rw
    restart: unless-stopped
    depends_on:
      - br-db
      - br-app
    env_file: './mattermost/local/.app.env'
    networks:
      - br-net

  mattermost-db:
    container_name: mattermost-db
    platform: linux/amd64
    image: postgres:15.2-alpine
    ports:
      - "54323:5432"
    restart: always
    env_file: './mattermost/local/.db.env'
    volumes:
      - mattermost-db-data:/var/lib/postgresql/data
    networks:
      - br-net

volumes:
  br-pgsql-data:
  redisdata:
  mattermost-db-data:

networks:
  br-net:
    driver: bridge

And This is my nginx configuration in a file named default.conf.tpl:

server {
    listen 80;

    location / {
        uwsgi_pass br-app:9000;
        include /etc/nginx/uwsgi_params;
        client_max_body_size 50M;
        proxy_read_timeout 300s;
    }

   location ~ /api/v4/websocket$ {
        proxy_set_header Upgrade \$http_upgrade\;
        proxy_set_header Connection "upgrade";
        proxy_set_header Host \$http_host\;
        proxy_set_header X-Real-IP \$remote_addr\;
        proxy_set_header X-Forwarded-For \$proxy_add_x_forwarded_for\;
        proxy_set_header X-Forwarded-Proto \$scheme\;
        proxy_pass http://mattermost:8065;
   }
}
server {
    listen 443;

    location / {
        uwsgi_pass br-app:9000;
        include /etc/nginx/uwsgi_params;
        client_max_body_size 50M;
        proxy_read_timeout 300s;
    }

   location ~ /api/v4/websocket$ {
        proxy_set_header Upgrade \$http_upgrade\;
        proxy_set_header Connection "upgrade";
        proxy_set_header Host \$http_host\;
        proxy_set_header X-Real-IP \$remote_addr\;
        proxy_set_header X-Forwarded-For \$proxy_add_x_forwarded_for\;
        proxy_set_header X-Forwarded-Proto \$scheme\;
        proxy_pass http://mattermost:8065;
   }
}

Let’s assume my domain is this: https://domain.com
I’m trying to connect from the client app to this: wss://domain.com/api/v4/websocket
When I see the logs for Nginx, I’m getting 400 error from Mattermost.
And When I see Mattermost docker container logs, This is what I’m seeing:

{"timestamp":"2023-12-21 15:48:07.082 Z","level":"info","msg":"Starting websocket hubs","caller":"platform/web_hub.go:95","number_of_hubs":4}
{"timestamp":"2023-12-21 15:53:36.287 Z","level":"debug","msg":"websocket.NextReader: closing websocket","caller":"platform/web_conn.go:849","user_id":"","conn_id":"ugc36pt8fir6tr3fphzdxb7u3h","error":"websocket: close 1001 (going away)"}
{"timestamp":"2023-12-21 15:53:36.287 Z","level":"debug","msg":"Received HTTP request","caller":"web/handlers.go:173","method":"GET","url":"/api/v4/websocket","request_id":"jgsx4asum7dqxyb996twxnp7ko","user_id":""}

{"timestamp":"2023-12-21 16:09:06.191 Z","level":"debug","msg":"websocket.NextReader: client side closed socket","caller":"platform/web_conn.go:845","user_id":"dunbpz13m3bd7yi3pa3hbuj3hh","conn_id":"rwuerwmspfyo3e7e7pwkk83dew"}
{"timestamp":"2023-12-21 16:09:06.191 Z","level":"debug","msg":"Received HTTP request","caller":"web/handlers.go:173","method":"GET","url":"/api/v4/websocket","request_id":"jymam8uchtgxzjn4pht8r7zknw","user_id":"dunbpz13m3bd7yi3pa3hbuj3hh"}
{"timestamp":"2023-12-21 16:09:18.097 Z","level":"debug","msg":"websocket.NextReader: client side closed socket","caller":"platform/web_conn.go:845","user_id":"dunbpz13m3bd7yi3pa3hbuj3hh","conn_id":"rwuerwmspfyo3e7e7pwkk83dew"}
{"timestamp":"2023-12-21 16:09:18.097 Z","level":"debug","msg":"Received HTTP request","caller":"web/handlers.go:173","method":"GET","url":"/api/v4/websocket","request_id":"6er9kbyjdty1dxqstu116sob9y","user_id":"dunbpz13m3bd7yi3pa3hbuj3hh"}

What is the problem you think ? I’m a little bit confused.

Also I need to mention that the domain itself is on Cloudflare, And I’m using my own certificate for the domain.

2

After struggling with it, I’ve found the problem.

I had a nginx configuration like this:

upstream chatsystem {
    server mattermost:8065;
}

proxy_cache_path /var/cache/nginx levels=1:2 keys_zone=mattermost_cache:10m max_size=3g inactive=120m use_temp_path=off;

server {
    listen 80;
    server_name domain.com;

    location / {
        uwsgi_pass br-app:9000;
        include /etc/nginx/uwsgi_params;
        client_max_body_size 50M;
        proxy_read_timeout 300s;
    }

 location ~ /mattermost/api/v[0-9]+/(users/)?websocket$ {
       proxy_set_header Upgrade '$http_upgrade';
       proxy_set_header Connection "upgrade";
       client_max_body_size 50M;
       proxy_set_header Host '$http_host';
       proxy_set_header X-Real-IP '$remote_addr';
       proxy_set_header X-Forwarded-For '$proxy_add_x_forwarded_for';
       proxy_set_header X-Forwarded-Proto '$scheme';
       proxy_set_header X-Frame-Options SAMEORIGIN;
       proxy_buffers 256 16k;
       proxy_buffer_size 16k;
       client_body_timeout 60;
       send_timeout 300;
       lingering_timeout 5;
       proxy_connect_timeout 90;
       proxy_send_timeout 300;
       proxy_read_timeout 90s;
       proxy_http_version 1.1;
       proxy_pass http://chatsystem;
   }

   location /mattermost {
       client_max_body_size 50M;
       proxy_set_header Connection "";
       proxy_set_header Host '$http_host';
       proxy_set_header X-Real-IP '$remote_addr';
       proxy_set_header X-Forwarded-For '$proxy_add_x_forwarded_for';
       proxy_set_header X-Forwarded-Proto '$scheme';
       proxy_set_header X-Frame-Options SAMEORIGIN;
       proxy_buffers 256 16k;
       proxy_buffer_size 16k;
       proxy_read_timeout 600s;
       proxy_cache mattermost_cache;
       proxy_cache_revalidate on;
       proxy_cache_min_uses 2;
       proxy_cache_use_stale timeout;
       proxy_cache_lock on;
       proxy_http_version 1.1;
       proxy_pass http://chatsystem;
   }
}

This configuration was in a .tpl file “default.conf.tpl” and it was copying to container using this docker entrypoint:

#!/bin/sh

set -e

envsubst < /etc/nginx/default.conf.tpl > /etc/nginx/conf.d/default.conf
nginx -g 'daemon off;'

When It was parsing the default.conf.tpl it was ignoring the nginx variables like $host.

I’ve changed my nginx configuration file from default.conf.tpl to default.conf and Now it is working fine.

Here is the final nginx configuration file:

upstream chatsystem {
    server mattermost:8065;
}

proxy_cache_path /var/cache/nginx levels=1:2 keys_zone=mattermost_cache:10m max_size=3g inactive=120m use_temp_path=off;

server {
    listen 80;
    server_name domain.com;

    location / {
        uwsgi_pass br-app:9000;
        include /etc/nginx/uwsgi_params;
        client_max_body_size 50M;
        proxy_read_timeout 300s;
    }

 location ~ /mattermost/api/v[0-9]+/(users/)?websocket$ {
       proxy_set_header Upgrade $http_upgrade;
       proxy_set_header Connection "upgrade";
       client_max_body_size 50M;
       proxy_set_header Host $http_host;
       proxy_set_header X-Real-IP $remote_addr;
       proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
       proxy_set_header X-Forwarded-Proto $scheme;
       proxy_set_header X-Frame-Options SAMEORIGIN;
       proxy_buffers 256 16k;
       proxy_buffer_size 16k;
       client_body_timeout 60;
       send_timeout 300;
       lingering_timeout 5;
       proxy_connect_timeout 90;
       proxy_send_timeout 300;
       proxy_read_timeout 90s;
       proxy_http_version 1.1;
       proxy_pass http://chatsystem;
   }

   location /mattermost {
       client_max_body_size 50M;
       proxy_set_header Connection "";
       proxy_set_header Host $http_host;
       proxy_set_header X-Real-IP $remote_addr;
       proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
       proxy_set_header X-Forwarded-Proto $scheme;
       proxy_set_header X-Frame-Options SAMEORIGIN;
       proxy_buffers 256 16k;
       proxy_buffer_size 16k;
       proxy_read_timeout 600s;
       proxy_cache mattermost_cache;
       proxy_cache_revalidate on;
       proxy_cache_min_uses 2;
       proxy_cache_use_stale timeout;
       proxy_cache_lock on;
       proxy_http_version 1.1;
       proxy_pass http://chatsystem;
   }
}

And this is the final Dockerfile:

FROM nginxinc/nginx-unprivileged:1.25.3-alpine
LABEL maintainer="x@example.com"

COPY default.conf /etc/nginx/conf.d

COPY ./uwsgi_params /etc/nginx/uwsgi_params

COPY run.sh /run.sh

USER root

RUN chown nginx:nginx /etc/nginx/conf.d/default.conf

RUN chmod +x /run.sh

USER nginx

CMD ["/run.sh"]

And this is the entrypoint:

#!/bin/sh

set -e
nginx -g 'daemon off;'

Note: During my misconfiguration I was getting different kinds of errors from “malformed host header” or “missing host header” or “websocket errors”.

1 Like
3

Thank you so much for following up with the solution there, MJ!