rabbitmq-server/deps/rabbitmq_auth_backend_http

Overview

This plugin provides the ability for your RabbitMQ server to perform authentication (determining who can log in) and authorisation (determining what permissions they have) by making requests to an HTTP server.

As with all authentication plugins, this one requires RabbitMQ server 2.3.1 or later.

Note: it's at an early stage of development, although it's conceptually very simple.

Installing

Install the corresponding .ez files from our Community Plugins page.

Enabling the Plugin

To enable the plugin, set the value of the auth_backends configuration item for the rabbit application to include rabbit_auth_backend_http. auth_backends is a list of authentication providers to try in order.

See the Access Control guide for more information.

To use this backend exclusively, use the following snippet in rabbitmq.conf (currently in master)

auth_backends.1 = http

Or, in the classic config format (rabbitmq.config, prior to 3.7.0) or advanced.config:

[{rabbit, [{auth_backends, [rabbit_auth_backend_http]}]}].

See RabbitMQ Configuration guide for more detail on auth_backends.

Configuring the Plugin

You need to configure the plugin to know which URIs to point at and which HTTP method to use.

Below is a minimal configuration file example.

In rabbitmq.conf (currently RabbitMQ master):

auth_backends.1 = http
auth_http.user_path     = http://some-server/auth/user
auth_http.vhost_path    = http://some-server/auth/vhost
auth_http.resource_path = http://some-server/auth/resource
auth_http.topic_path    = http://some-server/auth/topic

In the classic config format (rabbitmq.config prior to 3.7.0 or advanced.config):

[
  {rabbit, [{auth_backends, [rabbit_auth_backend_http]}]},
  {rabbitmq_auth_backend_http,
   [{http_method,   post},
    {user_path,     "http(s)://some-server/auth/user"},
    {vhost_path,    "http(s)://some-server/auth/vhost"},
    {resource_path, "http(s)://some-server/auth/resource"},
    {topic_path,    "http(s)://some-server/auth/topic"}]}
].

By default http_method configuration is GET for backwards compatibility. It's recommended to use POST requests to avoid credentials logging.

What Must My Web Server Do?

This plugin requires that your web server respond to requests in a certain predefined format. It will make GET (by default) or POST requests against the URIs listed in the configuration file. It will add query string (for GET requests) or a URL-encoded request body (for POST requests) parameters as follows:

user_path

• username - the name of the user
• password - the password provided (may be missing if e.g. rabbitmq-auth-mechanism-ssl is used)

vhost_path

• username - the name of the user
• vhost - the name of the virtual host being accessed
• ip - the client ip address

Note that you cannot create arbitrary virtual hosts using this plugin; you can only determine whether your users can see / access the ones that exist.

resource_path

• username - the name of the user
• vhost - the name of the virtual host containing the resource
• resource - the type of resource (exchange, queue, topic)
• name - the name of the resource
• permission - the access level to the resource (configure, write, read) - see the Access Control guide for their meaning

topic_path

• username - the name of the user
• vhost - the name of the virtual host containing the resource
• resource - the type of resource (topic in this case)
• name - the name of the exchange
• permission - the access level to the resource (write or read)
• routing_key - the routing key of a published message (when the permission is write) or routing key of the queue binding (when the permission is read)

See topic authorisation for more information about topic authorisation.

Your web server should always return HTTP 200 OK, with a body containing:

• deny - deny access to the user / vhost / resource
• allow - allow access to the user / vhost / resource
• allow [list of tags] - (for user_path only) - allow access, and mark the user as an having the tags listed

Using TLS/HTTPS

If your Web server uses HTTPS and certificate verification, you need to configure the plugin to use a CA and client certificate/key pair using the rabbitmq_auth_backend_http.ssl_options config variable:

[
  {rabbit, [{auth_backends, [rabbit_auth_backend_http]}]},
  {rabbitmq_auth_backend_http,
   [{http_method,   post},
    {user_path,     "https://some-server/auth/user"},
    {vhost_path,    "https://some-server/auth/vhost"},
    {resource_path, "https://some-server/auth/resource"},
    {topic_path,    "https://some-server/auth/topic"},
    {ssl_options,
     [{cacertfile, "/path/to/cacert.pem"},
      {certfile,   "/path/to/client/cert.pem"},
      {keyfile,    "/path/to/client/key.pem"},
      {verify,     verify_peer},
      {fail_if_no_peer_cert, true}]}]}
].

It is recommended to use TLS for authentication and enable peer verification.

Debugging

Check the RabbitMQ logs if things don't seem to be working properly. Look for log messages containing "rabbit_auth_backend_http failed".

Example App (in Python)

In examples/rabbitmq_auth_backend_django there's a very simple Django app that can be used for authentication. On Debian / Ubuntu you should be able to run start.sh to launch it after installing the python-django package. It's really not designed to be anything other than an example.

See examples/README for slightly more information.

Building from Source

You can build and install it like any other plugin (see the plugin development guide).

This plugin depends on the Erlang client (just to grab a URI parser).