Proxy Server

A Web proxy is a program that acts as a middleman between a Web browser and an end server. Instead of contacting the end server directly to get a Web page, the browser contacts the proxy, which forwards the request to the end server. When the end server replies to the proxy, the proxy sends the reply to the browser.

Overview

• It is the 7th and the last lab of 15-213: Introduction to Computer Systems.
• The coding style and convention are inspired by suckless coding style.
• It is a multi-threaded program to be capable of handling multiple connections at the same time.
• It can cache web objects by storing copies of the responses it gets from the server to use it to respond immediately to any future connections to the same server. Its cache follows LRU eviction policy.

NOTE: All project specifications are described in ProxyLab write-up.

How does this proxy work?

• The main thread accepts connect requests from clients.
• For each client, the main thread creates a thread and passes the file descriptor of the client to be served.
• All created threads detach themselves after freeing the resources allocated for serving the client.

Program modules and implementation details

proxy.c contains the main function, as well as serve function as described below:

• The main function does the following steps:

  1. Open a listen port to receive connection requests from clients.
  2. Create and initialize a cache to be shared between all worker threads.
  3. Accept and establish a connection with each client.
  4. Create a thread and pass the shared cache and the client file descriptor to it.

• The serve function does the following steps:

  1. Read and parse the request the client sent.

  2. Forward the read request to the server.

  3. Get the response from the server, then forward it to the client.

  4. Free the allocated resources and close the connection with the client.

proxy_server provides the serve function described before with the following three functions:

• It provides serve function with the following three functions:

  • parse_request:

    1. Parse the request line from the HTTP request.
    2. Extract the method, uri and HTTP version from the request line.
    3. Parse the request headers.

  • forward_client_request:

    1. Build the request line that will be sent to the server.
    2. Build the request headers.
    3. Search in the cache for an element that matches the built request line and headers, then return it.
    4. Connect to the server and caches it, then returns it.

  • forward_server_response:

    1. Send the response back to the client.

safe_io:

• It provides safe and re-entrant functions to read and write data to connection sockets.

socket_interface:

• It provides an abstraction layer above the standard socket interface library to create TCP sockets.

• It supports TCP server or client sockets.

• It contains two functions:

  • open_client: establish a client socket connected to a server with the given hostname and port.
  • open_server: creates a listen socket associated with the given port to be ready to receive TCP connection requests.

Requirements

• linux
• git
• gcc
• make

How to test and use it?

1) Download, compile and run the source code as follows:

git clone https://github.com/IslamWalid/proxy_server.git

cd proxy_server

make

./proxy <port>

2) Connect to the proxy and send an HTTP request to the server using:

• telnet:
  • Use the following command to connect to the proxy:

  telnet localhost <port>
  

  • Write an HTTP request line, for ex:

  GET http://www.example.com HTTP/1.0
  

• browser:
  • Change your browser settings to connect to the proxy.
  • Connect to an HTTP website, for ex: http://www.example.com

NOTE: Connect to http websites (ex: websites mentioned here), or use tiny_web_server which I developed to be used in testing this proxy.