Blockchain and Distributed Systems:
Applications based on blockchain technology (such as Bitcoin) are similar to a "distributed database". Although blockchain systems are different from traditional distributed systems, discussions based on the characteristics of distributed systems are inevitable.
In the study of distributed systems, the following classical principles exist:
FLP Impossibility Principle: In a minimal asynchronous model system where the network is reliable but allows node failure (even if there is only one), there is no deterministic consensus algorithm that can solve the consistency problem. The FLP principle actually states that a purely asynchronous system cannot ensure that consistency is completed in a finite amount of time when nodes are allowed to fail. Even under the premise of non-Byzantine errors, algorithms such as Paxos and Raft cannot reach a consensus, but the probability of such a situation in engineering practice is very small.
CAP principle: It is impossible for a distributed computing system to ensure the following three characteristics at the same time: consistency, availability and partition tolerance, and the guarantee for a certain characteristic often needs to be weakened in the design. in,
- Consistency: Any operation should be atomic. Events that occur later can see the results caused by the previous time. Note that this refers to strong consistency.
- Availability: For a limited time, any non-failed node can answer the request.
- Partition tolerance: The network may be partitioned, that is, the communication between nodes is not guaranteed.