Deterministic Master Node Selection via Suitability Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for selecting a master node in groups of computing nodes are inefficient, particularly when a large number of nodes become unavailable or unresponsive, as they often rely on election algorithms that require majority votes and are prone to failures.
Innovation Solution
A deterministic master node selection process based on suitability values and unique identifiers, using a predefined timing scheme to ensure all nodes select the same master node without relying on election algorithms, even in scenarios where half of the nodes are unavailable.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If election algorithms requiring majority votes are used for master node selection, then the system can handle normal operational scenarios, but the system becomes unreliable when a large number of nodes become unavailable or unresponsive
Solution Approach 1:
The patent changes the selection parameter from vote-based majority to suitability value comparison. Each node calculates a suitability value based on performance metrics (CPU usage, memory availability, network bandwidth), and the node with the highest suitability value becomes master. This parameter change eliminates the need for complex election algorithms and majority voting, providing deterministic selection even when many nodes are unavailable.
Solution Approach 2:
Each computing node independently calculates its own suitability value and compares it with others to determine master node status. The system performs self-selection without requiring external coordination or complex election protocols. Nodes autonomously evaluate their own performance metrics and make deterministic decisions about master node candidacy based on predefined criteria.
2Reliability
If suitability values based on performance metrics are used, then the master node selection becomes deterministic and reliable, but the selection process requires collecting and comparing data from all nodes
Solution Approach 1:
Each computing node pre-calculates its suitability value based on its performance metrics before the selection process begins. Nodes continuously monitor their own resource availability (CPU, memory, network) and maintain up-to-date suitability values. This preliminary action allows for rapid deterministic selection when needed, as the comparison data is already prepared and available locally without requiring real-time data collection from all nodes.
Data Source
AI summary
Selecting a master node in a group of computing nodes. Computing nodes exchange suitability messages, candidate messages, and master messages to indicate the state of each node. A deterministic algorithm based on scores and timings is executed by each computing node to encourage each node to select the same master node.