[A web3 concept every day] Quickly understand PAXOS

PAXOS is a typical strongly consistent non-Byzantine consensus algorithm. It is different from the PoW and PoS we talked about earlier. Strong consistency means that all nodes in the system receive consistent results at the same point in time. Non-strong consistency means that it takes a certain amount of time for all data in the system to reach consensus, such as PoW and PoS.

So what is non-Byzantium? Byzantine refers to the "traitor" problem, which is a problem with extremely destructive, unpredictable, and random characteristics, such as hacker attacks. Non-Byzantine refers to problems such as downtime, network delay, and information loss. The Paxos algorithm sets up three nodes: proposer, acceptor and learner. How it works is that the proposer sends a uniquely numbered proposal to the acceptor during the Prepare phase to request a commitment. When the acceptor responds, if they have previously accepted other proposals, they will return the values ​​and numbers of those proposals to the proposer. If the proposer obtains the commitment of more than half of the acceptors, it will enter the Accept phase and send the specific value of the proposal. Once this value is accepted by more than half of the acceptors, consensus is reached on the proposal. The learner is then informed of this decision, which is to ensure that other parts of the system know which proposals were accepted, thus maintaining data consistency across the network.

According to the famous CAP theorem, a distributed system cannot guarantee consistency, availability, and partition tolerance at the same time. All consensus algorithms are based on actual conditions to achieve a balance among the three. The purpose of PAXOS is to ensure strong consistency, which results in sacrificing a certain amount of availability. Different from PoW, PoW achieves eventual consistency, which means that short-term forks can be allowed, while PAXOS will not have forks, because only consistent results can allow the system to continue running. At the same time, its focus is on solving non-Byzantine faults, which will also cause the system to fail to operate normally when it is attacked by malicious attacks. For example, if a malicious proposer keeps issuing new proposals, the entire system will come to a standstill. #Paxos #热门话题 #每日解析