Peer-to-Peer Architectures Introduction

Peer-to-peer (P2P) architectures are networked systems where each node in the network is on equal footing with every other node.

The largest and most successful P2P network is the Internet itself. Every machine connected to the network gets an IP address and can interact with any other machine on the Internet so long as it knows that machine’s IP address.

The Web is not a P2P architecture because it sees some machines as clients and some as servers. They are not peers.

Special purpose P2P systems are frequently built on top of the Internet. Such a network is called an overlay network.

Designing a functional P2P system can be difficult because humans think in stories—linearly. And we love centralized systems because they’re easy to understand.

This lesson introduces concepts and algorithms for building practical decentralized peer-to-peer systems.

Lesson Objectives

After completing this lesson, you should be able to

  1. Explain the eight fallacies of distributed computing
  2. Give examples of the CAP theorem and explain tradeoffs
  3. Show how distributed file sharing works and relate it to consensus
  4. Describe distributed discovery and show how a distributed hash table works

Required Reading/Viewing

Additional Resources