Youtube P2P Caching System

A peer-to-peer decentralized caching system storing RESTful YouTube search results in FreePastry-based distributed hash table (DHT). Java.

Skill

Java, Servlet, FreePastry Distributed Hash Table, Web Services(REST, SOAP), Bekeley DB

Contribution

1. Built a peer-to-peer decentralized caching system storing RESTful YouTube search results in FreePastry-based distributed hash table (DHT).
2. Built a web services based web application that queries and receives REST(JSON)/SOAP messages between client and caching system.

Deployment

To launch P2P ring, you need to use:

    java -cp [lib files] [P2PCache location]
    [the port number on the local machine to which the Pastry node should bind]
    [the IP address of the Pastry bootstrap node]
    [the port number of the Pastry bootstrap node]
    [the port number to which the daemon should bind]
    [the path to the BerkeleyDB database]
    [max maximum storage size]?
    [content expiration time]?

    For example:
    java -cp youtube.jar:target/WEB-INF/lib/* edu/upenn/cis455/youtube/P2PCache 9001 10.12.231.124 9001 10001 /Users/Alantyy/Desktop/database 10 1000

Files

1. Main source files are located in src/edu/upenn/cis455.
2. In /storage folder: DatabaseUtil.java is for Berkeley DB DAO functions, P2PNode.java stores information about PastryNode and YouTubeResult.java for YouTube reults.
3. In /youtube folder: P2P Cache is the main class to launch P2P ring and server, YouTubeSearch.java is the main class to launch servlet web application, NodeFactory.java encapsulates functions to create P2P nodes and ring, P2PApplication.java is used for routing - forwarding requests to specific nodes, YouTubeClient.java is used to search YouTube using google API, P2PDaemonServer.java is the same software running on every P2PNode used to listen to requests from client.
4. Links: view specification.