Distributed Computing Node Consensus via Execution Trail
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional distributed computing systems have low reliability due to the single-point failure of the control node and potential falsification of computing results, as they lack robust fault tolerance and verification mechanisms.
Innovation Solution
A distributed computing system where multiple execution node computers autonomously select and execute partial processes, record execution results in a shared trail, and form consensus on these results, eliminating the need for a central control node and enhancing fault tolerance and result verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a control node computer is used to manage execution node computers, then centralized control and coordination are achieved, but the system has low reliability due to single-point failure
Solution Approach 1:
The patent extracts the control functionality from a centralized control node and distributes it to execution node computers themselves. Each execution node autonomously determines its own task assignments and manages its execution, eliminating the single-point failure vulnerability of centralized control while maintaining coordinated operation across the distributed system.
Solution Approach 2:
Execution node computers perform self-service by autonomously selecting tasks from the task queue, determining their own task assignments, and managing their execution without external control. This self-organizing behavior eliminates dependency on a central controller and enhances system reliability through distributed decision-making.
2Ease of operation
If a control node computer verifies computing results, then centralized verification is simplified, but the system has low reliability due to potential falsification of results
Solution Approach 1:
The patent implements feedback mechanisms where execution node computers transmit their determined task assignments and computing results back to the system. Other execution nodes verify these results through cross-checking, creating a distributed feedback loop that detects and prevents falsification without requiring centralized verification authority.
Solution Approach 2:
The patent merges the verification function into the execution process itself. Multiple execution nodes simultaneously perform verification of each other's results through distributed cross-checking, combining verification capabilities across the system rather than relying on a single centralized verifier.
3Reliability
If multiple execution node computers operate autonomously, then system reliability is improved through fault tolerance, but coordination and task assignment become more complex
Solution Approach 1:
The patent segments the task management function into individual execution nodes, where each node independently maintains a task queue and autonomously selects tasks. This segmentation eliminates the need for complex centralized coordination while maintaining systematic task distribution through localized decision-making at each node.
Data Source
AI summary
A communication device among a plurality of communication devices is used in a distributed computing system. The distributed computing system executes a target process including a plurality of partial processes by using the plurality of communication devices. The communication device includes a memory and a processor. The memory stores a trail that represents a state of the plurality of partial processes. The processor selects, from among the plurality of partial processes, an uncompleted partial process with a number of equivalent execution results being less than a target number according to the trail. The processor executes the uncompleted partial process selected by the selector. The processor records an execution result obtained by the execution unit in the trail.