Consensus CFT

Consensus algorithms in distributed systems are protocols that enable multiple computers or nodes within a network to agree on a single data value or decision, ensuring consistency and reliability across the system despite potential failures or malicious behavior of some nodes. These algorithms are foundational for maintaining data integrity and synchrony, especially in environments where nodes operate independently and may experience different states or updates.

Crash Fault Tolerant (CFT) Algorithms:

• Paxos: A family of protocols that achieve consensus despite network delays, node failures, and message losses. Paxos is known for its robustness but is often considered complex to understand and implement.

• Raft: Designed to be more understandable and easier to implement than Paxos, Raft achieves consensus by electing a leader that manages the replication of log entries to other nodes.

Démonstrations avec Zookeeper et Etcd

• https://github.com/heidihoward/distributed-consensus-reading-list
• https://www.geeksforgeeks.org/consensus-algorithms-in-distributed-system/