Understanding Distributed Theory and Algorithms: Importance, Core Concepts, and Learning Path

This article introduces the significance of distributed theory and algorithms, emphasizing their essential role in high‑level architecture interviews and long‑term career development.

Importance

It addresses common questions such as CAP theorem, Zookeeper leader election, and fault tolerance, highlighting that mastering these concepts boosts core competitiveness.

Is it easy to learn?

Many developers know the basics of CAP and BASE but rarely dive deep into algorithms due to three reasons: perceived complexity, lack of plain‑language resources, and absence of a clear learning path.

Four Foundational Theories

Byzantine Generals Problem

CAP Theory

ACID Theory

BASE Theory

Eight Distributed Protocols and Algorithms

Paxos Algorithm

Raft Algorithm

Consistent Hashing

Gossip Protocol

Quorum NWR Algorithm

FBFT Algorithm

POW Algorithm

ZAB Protocol

How to Learn Efficiently

Choosing the right algorithm for a given scenario requires understanding each algorithm’s trade‑offs between consistency, availability, performance, and fault tolerance.

Four Evaluation Dimensions

The dimensions are Byzantine fault tolerance, consistency, performance, and availability.

Byzantine Fault Tolerance

Describes the untrusted environment model from the Byzantine Generals Problem, distinguishing between Byzantine fault tolerance and non‑Byzantine (fault) tolerance.

Non‑Byzantine algorithms include 2PC, TCC, Paxos, Raft, Gossip, Quorum NWR, and ZAB.

Byzantine algorithms include POW and FBFT.

Consistency

Strong consistency – reads always see the latest write.

Weak consistency – no guarantee that reads see the latest write.

Eventual consistency – reads eventually converge to the latest value.

Strong consistency is often achieved with 2PC or Raft; eventual consistency with Gossip and Quorum NWR; ZAB provides eventual consistency optimized for read performance.

Availability

Availability means the system can respond, though not necessarily with the freshest data. Gossip offers the highest availability, followed by Paxos/Raft/ZAB/Quorum NWR/FBFT/POW, while 2PC and TCC have the lowest.

Performance

Performance correlates with availability; higher availability generally yields better performance. Gossip excels in AP‑type systems, while consensus‑based algorithms have moderate performance, and 2PC/TCC are the slowest.

Learning Roadmap

Lecture 1: Byzantine Generals Problem

Explained using the four roles in the card game "Three Kingdoms Kill".

Lecture 2: CAP, BASE, ACID Theories

Uses 刚 (rigid) and 柔 (flexible) from Tai Chi to illustrate ACID vs. BASE, with CAP balancing the two.

Lecture 3: Paxos Algorithm

Illustrated with characters Zhuge Liang and Pang Tong acting as 提议者 (proposers).

Lecture 4: Raft Algorithm

Demonstrated with animated GIFs showing the leader election process.

Lecture 5: Consistent Hashing

Explained via the story of Han Xin’s troop deployment.

Lecture 6: Gossip Protocol

Compared to a contagious virus to show how information spreads.

Lecture 7: Quorum NWR

Uses the alchemical furnace of Tai Shang Lao Jun as a metaphor for N, W, R parameters.

Lecture 8: POW Algorithm

Connects blockchain and Bitcoin concepts with the story of Zixia Fairy and Supreme Treasure.

A PDF compilation of the eight lectures (over 20,000 words) is available; reply with the keyword 分布式 to receive it.