Consensus Algorithms

Ensure agreement among computers (nodes) in a distributed system.

• Agreement: All nodes decide on the same value.
• Termination: Decision-making process ends.
• Validity: Decided value was proposed by at least one node.

• Leader-Based: One node leads decision-making (e.g., Raft, Paxos).
  • Pro: Strong consistency
  • Con: Less available if leader fails
• Quorum-Based: Majority agreement required (e.g., Paxos variants).
  • Pro: Good availability
  • Con: Weaker consistency in network splits
• Proof-of-Work/Stake: Used in blockchains for transaction validation.
  • Pro: Secure, decentralized
  • Con: High energy consumption (PoW), potential centralization (PoS)

• Distributed databases
• Configuration management
• Leader election
• Blockchain
• Cloud computing

Depends on your system's needs for consistency, availability, fault tolerance, performance, and scalability.

see CAP theorem