Computing Power Node Load Reallocation for Dynamic Server Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The issue of unbalanced computing power load distribution among multiple servers in a network, which occurs due to dynamic changes in the number and usage of computing power nodes, leading to inefficiencies and potential overload or underutilization.
Innovation Solution
A method for computing power load balancing that involves determining load adjustment amounts based on the load and weight of each computing power node and its associated nodes, adjusting loads accordingly to achieve dynamic balancing across the network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple servers provide computing power for multiple users, then the network can serve more users with open computing power, but unbalanced computing power load occurs among the servers
Solution Approach 1:
The patent implements a feedback mechanism where the management module continuously monitors the load status of each server and uses this information to dynamically adjust task allocation. The system calculates load balancing metrics based on real-time server states and redistributes computing tasks to achieve balanced load distribution across all servers, thereby resolving the load balance problem while maintaining open computing power access.
2Adaptability or versatility
If computing power nodes are dynamically added or removed, then the system can adapt to changing demands, but load distribution becomes unbalanced
Solution Approach 1:
The patent employs dynamic load balancing strategies that automatically adjust to changing system conditions. When computing power nodes are added or removed, the management module detects these changes and recalculates optimal task distributions. The system dynamically reassigns computing tasks to maintain balanced load distribution, ensuring high resource utilization efficiency regardless of node fluctuations.
3Device complexity
If load balancing is not implemented, then the system structure remains simple, but servers experience overload or underutilization
Solution Approach 1:
The patent integrates load balancing functionality into an existing computing power management system without requiring a complete system redesign. The management module performs multiple functions including task allocation, load monitoring, and dynamic task redistribution within a unified framework. This multi-functional approach maintains relative system simplicity while significantly improving server stability and preventing overload or underutilization conditions.
Data Source
AI summary
A computing power load balancing method includes: obtaining a load and a weight of a computing power node, and loads and weights of associated computing power nodes of the computing power node; determining a load adjustment amount of the computing power node and load adjustment amounts of the associated computing power nodes based on the load and the weight of the computing power node and the loads and the weights of the associated computing power nodes; and adjusting the load of the computing power node based on the load adjustment amount of the computing power node and the load adjustment amounts of the associated computing power nodes.