anxolerd/reverse_forwarding

Description

This service is for accepting requests from remote services on your local machine during debuging. Common case - testing integration with payment providers.

How it works

Application creates 2 processes and instance of multiprocessing.Array for communication between processes. In first of them app establishes ssh connection with service serveo.net for reverse forwarding. Then we parse remote origin, given us by servo.net and save it into created array. In the other process app for sending remote origin and redirecting requests is launched.

Integration

By default service is running on localhost:9090. In your main application you must get remote origin from the service to send it to needed remote application which can send request to your local app. Also, during getting remote origin, you acknowledge the service, that it has forward requests from remote to concrete local app and only requests on concrete url. To get remote origin for your local app, send POST request to the reverse forvarding service on url /proxify_link with required body parametr url on which you want to accept remote requests.
Here is example of implementation on Python:

import json
import requests
from urlparse import urlparse

class RemoteUrlForwarder(object):
    _url = 'http://localhost:9090/proxify_link'  # url to reverse forwarding service

    def _post(self, url_to_forward):
        try:
            response = requests.post(
                url=self._url,
                data=json.dumps({'url': url_to_forward}),
            )
            return json.loads(response.content)
        except Exception:
            return {}

    def _is_configured(self):
        return self._url is not None

    def get_remote_link_to_localhost(self, url_to_forward):
        if not self._is_configured():
            return url_to_forward

        data = self._post(url_to_forward)
        remote_origin = data.get('proxy_hostname')
        if not remote_origin:
            return url_to_forward

        parsed = urlparse(url_to_forward)
        local_origin = '%s://%s' % (parsed.scheme, parsed.netloc)
        return '%s%s' % (remote_origin, url_to_forward[len(local_origin):])

Running

0. Install pyenv.

If you have pyenv on your computer, skip this step. Otherwise, follow next:

• Execute in terminal

  curl -L https://raw.githubusercontent.com/pyenv/pyenv-installer/master/bin/pyenv-installer | bash

• Add lines in ~/.bashrc:

  export PATH="~/.pyenv/bin:$PATH"
eval "$(pyenv init -)"
eval "$(pyenv virtualenv-init -)"

• Execute in terminal:

  exec bash

• Execute in terminal:

  pyenv update

1. Create virtualenv:

Create virtualenv using pyenv:

• Install appropriate version of python:

  pyenv install 3.6.4

• Create virtualenv named reverse_forwarding:

  pyenv virtualenv 3.6.4 reverse_forwarding

2. Clone project:

• Execute being in your workspace directory:

  git clone https://github.com/landrew31/reverse_forwarding.git

3. Run service:

Run beeing in project path.

• Launch virtualenv if needed:

  pyenv local reverse_forwarding

• Run:

  python3 app.py

This will create ssh tunnel to your local machine from serveo.net service and
use the obtained url to generate external urls.

As an alternative, you can use any reverse proxy you want in combination with
this utility. To do so first run the reverse proxy and obtain the external url.
For example with ngrok reverse proxy you may use the following command:

>`ngrok http 9090  # or whatever else port you want`

After that run reverse_forwarding app with obtained external url as commandline
parameter:

>`python3 app.py --url http://external.url.example.com`