Cached Proxy Server

This is a simple Flask app that acts as a caching proxy for GET requests. This is for a take-home interview I did in 2015.

It accepts incoming GET requests and forwards them on their destination using a simple caching layer represented by the ResponseCache class.

I developed this app for an interview task, and I figured I'd use it as an opportunity to mess around with Flask and a few Python modules like requests, urlparse, etc. I wrote it in my last days in Ecuador, so it might be a little sloppy. Give it a try, let me know what you think :)

Requirements

Note: This was developed and tested in an Ubuntu 12.04 virtual machine

• Python (tested with Python 2.7)
• Python Modules
  • flask
  • requests

I recommend you use virtualenv to run the server, so you don't mess with your Python install. To install virtualenv, do:

sudo apt-get install python-virtualenv

To initialize your virtual environment, cd to the directory that contains this repository and do:

virtualenv environment

To install the required Python modules in your virtual environment, do:

environment/bin/pip install flask
environment/bin/pip install requests

Usage

You can edit the configuration of the cache and the port the server runs on by editing conf.py. This file should look like:

CACHE_DURATION_MS   = 10 * 1000
CACHE_SIZE_BYTES    = 100000
CACHE_SIZE_ELEMENTS = 8
LOG_TABLE_MAX_SIZE = 100
PORT = 5000

If you're using virtualenv as described above, run the server by doing:

environment/bin/python server.py

Otherwise just do:

python server.py

Now that the server is running, I recommend you keep two browser tabs open:

• In the first tab, go to localhost:PORT/proxyinfo (port 5000 is default). This shows general information about the cache. It's not that dynamic, so refresh it after every request.
• In the second tab, go to localhost:PORT/proxy/www.google.com (for example) to use the proxy and cache.

Look at the log in the proxyinfo page to see what the cache is doing.

TODO's

• The proxyinfo page is just static HTML (err, it uses Flask's dynamic templating, but you still have to refresh the page for new info). I want to look into Flask's ajax support for making it more dynamic.
• The cache layer is better developed than the proxy layer. I need to do more testing and improvements there.