Advanced Configuration

Best Practice: Use a Docker network

For those who have a few of their upstream services running in Docker on the same Docker host as NPM, here's a trick to secure things a bit better. By creating a custom Docker network, you don't need to publish ports for your upstream services to all of the Docker host's interfaces.

Create a network, ie "scoobydoo":

docker network create scoobydoo

Then add the following to the docker-compose.yml file for both NPM and any other services running on this Docker host:

networks:
  default:
    external:
      name: scoobydoo

Let's look at a Portainer example:

version: '3'
services:

  portainer:
    image: portainer/portainer
    privileged: true
    volumes:
      - './data:/data'
      - '/var/run/docker.sock:/var/run/docker.sock'
    restart: unless-stopped

networks:
  default:
    external:
      name: scoobydoo

Now in the NPM UI you can create a proxy host with portainer as the hostname, and port 9000 as the port. Even though this port isn't listed in the docker-compose file, it's "exposed" by the Portainer Docker image for you and not available on the Docker host outside of this Docker network. The service name is used as the hostname, so make sure your service names are unique when using the same network.

Docker Healthcheck

The Dockerfile that builds this project does not include a HEALTHCHECK but you can opt in to this feature by adding the following to the service in your docker-compose.yml file:

healthcheck:
  test: ["CMD", "/bin/check-health"]
  interval: 10s
  timeout: 3s

Docker Secrets

This image supports the use of Docker secrets to import from file and keep sensitive usernames or passwords from being passed or preserved in plaintext.

You can set any environment variable from a file by appending __FILE (double-underscore FILE) to the environmental variable name.

version: "3.7"

secrets:
  # Secrets are single-line text files where the sole content is the secret
  # Paths in this example assume that secrets are kept in local folder called ".secrets"
  DB_ROOT_PWD:
    file: .secrets/db_root_pwd.txt
  MYSQL_PWD:
    file: .secrets/mysql_pwd.txt

services:
  app:
    image: 'jc21/nginx-proxy-manager:latest'
    restart: unless-stopped
    ports:
      # Public HTTP Port:
      - '80:80'
      # Public HTTPS Port:
      - '443:443'
      # Admin Web Port:
      - '81:81'
    environment:
      # These are the settings to access your db
      DB_MYSQL_HOST: "db"
      DB_MYSQL_PORT: 3306
      DB_MYSQL_USER: "npm"
      # DB_MYSQL_PASSWORD: "npm"  # use secret instead
      DB_MYSQL_PASSWORD__FILE: /run/secrets/MYSQL_PWD
      DB_MYSQL_NAME: "npm"
      # If you would rather use Sqlite uncomment this
      # and remove all DB_MYSQL_* lines above
      # DB_SQLITE_FILE: "/data/database.sqlite"
      # Uncomment this if IPv6 is not enabled on your host
      # DISABLE_IPV6: 'true'
    volumes:
      - ./data:/data
      - ./letsencrypt:/etc/letsencrypt
    secrets:
      - MYSQL_PWD
    depends_on:
      - db
  db:
    image: jc21/mariadb-aria
    restart: unless-stopped
    environment:
      # MYSQL_ROOT_PASSWORD: "npm"  # use secret instead
      MYSQL_ROOT_PASSWORD__FILE: /run/secrets/DB_ROOT_PWD
      MYSQL_DATABASE: "npm"
      MYSQL_USER: "npm"
      # MYSQL_PASSWORD: "npm"  # use secret instead
      MYSQL_PASSWORD__FILE: /run/secrets/MYSQL_PWD
    volumes:
      - ./data/mysql:/var/lib/mysql
    secrets:
      - DB_ROOT_PWD
      - MYSQL_PWD

Disabling IPv6

On some Docker hosts IPv6 may not be enabled. In these cases, the following message may be seen in the log:

Address family not supported by protocol

The easy fix is to add a Docker environment variable to the Nginx Proxy Manager stack:

    environment:
      DISABLE_IPV6: 'true'

Custom Nginx Configurations

If you are a more advanced user, you might be itching for extra Nginx customizability.

NPM has the ability to include different custom configuration snippets in different places.

You can add your custom configuration snippet files at /data/nginx/custom as follow:

/data/nginx/custom/root.conf: Included at the very end of nginx.conf
/data/nginx/custom/http_top.conf: Included at the top of the main http block
/data/nginx/custom/http.conf: Included at the end of the main http block
/data/nginx/custom/stream.conf: Included at the end of the main stream block
/data/nginx/custom/server_proxy.conf: Included at the end of every proxy server block
/data/nginx/custom/server_redirect.conf: Included at the end of every redirection server block
/data/nginx/custom/server_stream.conf: Included at the end of every stream server block
/data/nginx/custom/server_stream_tcp.conf: Included at the end of every TCP stream server block
/data/nginx/custom/server_stream_udp.conf: Included at the end of every UDP stream server block

Every file is optional.

X-FRAME-OPTIONS Header

You can configure the X-FRAME-OPTIONS header value by specifying it as a Docker environment variable. The default if not specified is deny.

  ...
  environment:
    X_FRAME_OPTIONS: "sameorigin"
  ...

OpenID Connect SSO

You can secure any of your proxy hosts with OpenID Connect authentication, providing SSO support from an identity provider like Azure AD or KeyCloak. OpenID Connect support is provided through the lua-resty-openidc library of OpenResty.

You will need a few things to get started with OpenID Connect:

A registered application with your identity provider, they will provide you with a Client ID and a Client Secret. Public OpenID Connect applications (without a client secret) are not yet supported.
A redirect URL to send the users to after they login with the identity provider, this can be any unused URL under the proxy host, like https://<proxy host url>/private/callback, the server will take care of capturing that URL and redirecting you to the proxy host root. You will need to add this URL to the list of allowed redirect URLs for the application you registered with your identity provider.
The well-known discovery endpoint of the identity provider you want to use, this is an URL usually with the form https://<provider URL>/.well-known/openid-configuration.

After you have all this you can proceed to configure the proxy host with OpenID Connect authentication.

You can also add some rudimentary access control through a list of allowed emails in case your identity provider doesn't let you do that, if this option is enabled, any email not on that list will be denied access to the proxied host.

The proxy adds some headers based on the authentication result from the identity provider:

X-OIDC-SUB: The subject identifier, according to the OpenID Coonect spec: A locally unique and never reassigned identifier within the Issuer for the End-User.
X-OIDC-EMAIL: The email of the user that logged in, as specified in the id_token returned from the identity provider. The same value that will be checked for the email whitelist.
X-OIDC-NAME: The user's name claim from the id_token, please note that not all id tokens necessarily contain this claim.

7.3 KiB Raw Blame History