Consensus Node Removal Through Byzantine Voting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed systems face performance degradation due to faulty nodes that hinder consensus processes, leading to inefficient operations and the need for null votes, which disrupt the normal functioning of the network.
Innovation Solution
A method to identify and remove abnormal nodes by analyzing historical proposal records, transmitting node identifiers in votes across consensus nodes, and achieving consensus on the removal through a Byzantine consensus protocol, ensuring only normal nodes participate in the consensus.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If faulty nodes are allowed to remain in the distributed system, then the system maintains its original node composition and structure, but the consensus process efficiency deteriorates and performance degrades due to null votes and disrupted operations
Solution Approach 1:
The patent extracts faulty nodes from the distributed system by identifying them through historical proposal record analysis and removing them through a structured voting process. This extraction eliminates the harmful impact of faulty nodes on consensus efficiency while maintaining system reliability through the Byzantine consensus protocol.
Solution Approach 2:
The patent performs preliminary identification of faulty nodes by analyzing historical proposal records before they can significantly degrade system performance. The node identifier is embedded in votes in advance, preparing the system for efficient removal through the voting process before consensus efficiency is severely impacted.
2Reliability
If historical proposal records are analyzed to identify abnormal nodes, then faulty nodes can be detected and removed, but the complexity of the consensus process increases due to additional analysis and voting steps
Solution Approach 1:
The patent uses historical proposal records as feedback to identify faulty nodes. By analyzing past proposal generation patterns and embedding node identifiers in votes, the system creates a feedback mechanism that automatically detects and removes faulty nodes through the consensus process, improving identification accuracy without requiring external intervention.
Solution Approach 2:
The distributed system performs self-diagnosis and self-purification by having nodes analyze historical records and vote on faulty node removal. The system serves itself by automatically identifying and removing faulty nodes through the existing consensus mechanism, reducing the need for external management while maintaining reliability.
3Reliability
If node identifiers are transmitted in votes across consensus nodes, then faulty nodes can be identified and removed through consensus, but the communication overhead and time required for consensus increases
Solution Approach 1:
The patent merges the faulty node identification process with the existing Byzantine consensus voting process. By embedding node identifiers directly in the votes during the proposal, prevote, and precommit stages, the system combines identification and removal functions into a single consensus process, reducing communication overhead and time compared to separate identification and removal procedures.
4Measurement precision
If the total number of proposal generations reaches a threshold, then faulty nodes can be identified based on proposal generation counts, but the system requires accumulation of historical data which delays immediate detection
Solution Approach 1:
The patent performs preliminary accumulation of proposal generation counts for each node, storing this data in a node proposal map. When the total proposal count reaches a threshold, the system can quickly identify faulty nodes by comparing individual node counts against the accumulated historical data, enabling accurate detection without requiring continuous real-time analysis.
Data Source
AI summary
A node consensus method includes selecting, an abnormal node meeting a set node anomaly condition from a plurality of the consensus nodes in response to a target proposal being obtained and a locally recorded total number of proposal generations reaching a total number threshold; transmitting a first vote in a prevote stage to other consensus nodes and receiving one or more first votes in the prevote stage that are transmitted by the other consensus nodes, the first vote transmitted by a first consensus node carrying a node identifier of the abnormal node; transmitting a second vote in a precommit stage to the other consensus nodes, and receiving one or more second votes in the precommit stage that are transmitted by the other consensus nodes; and determining that a consensus is reached on the target proposal and deleting the abnormal node from the plurality of consensus nodes.