Load Balancing Device Predicting Server Load Fluctuations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current global server load balancing (GSLB) devices continue to operate spare servers even after load fluctuations have normalized, leading to increased costs due to prolonged power consumption for on-premises servers and metered billing for public cloud servers.
Innovation Solution
A load balancing device with a load prediction/control unit that predicts load fluctuations based on server load information, shortening the operational time of servers when a decrease in load is anticipated, thereby optimizing the number of active servers and reducing operational costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the spare server continues to operate after the rapidly increased traffic returns to a normal state, then the service availability is maintained, but the operational cost increases due to prolonged power consumption and metered billing
Solution Approach 1:
The patent implements dynamic server operation management by continuously monitoring load information and adjusting server operation status based on real-time conditions. The load prediction/control unit dynamically determines whether to operate or stop spare servers based on predicted load fluctuations, transforming the static server operation model into a dynamic one that adapts to changing traffic conditions.
Solution Approach 2:
The patent applies preliminary action by predicting load fluctuations before they occur and proactively adjusting server operation status in advance. The load prediction/control unit analyzes current load information and predicts future load changes, then stops or operates servers accordingly before the actual load change happens, preventing unnecessary operational costs while ensuring service availability when needed.
2Loss of energy
If the spare server is stopped to reduce operational costs, then the energy consumption is reduced, but the service availability may be compromised when load increases occur
Solution Approach 1:
The patent implements feedback mechanisms by continuously collecting load information from servers and using this feedback to adjust server operation status. The load prediction/control unit monitors load changes in real-time and uses this feedback to make informed decisions about server operation, ensuring that servers are stopped only when it is safe to do so and operated when needed for service availability.
Solution Approach 2:
The patent uses preliminary action by predicting load increases before they occur and proactively operating spare servers in advance. The load prediction/control unit analyzes current load trends and predicts future load increases, then operates servers before the actual increase happens, ensuring service availability is maintained while minimizing unnecessary operational time.
3Loss of energy
If load information is continuously monitored to optimize server operation, then the operational cost is reduced, but the system complexity increases
Solution Approach 1:
The patent applies universality by designing the load prediction/control unit to perform multiple functions: collecting load information, analyzing load changes, predicting load fluctuations, and controlling server operation status. This multi-functional approach consolidates what could be separate complex systems into a single integrated unit, reducing overall system complexity while achieving cost optimization.
Solution Approach 2:
The patent implements self-service by enabling the load prediction/control unit to autonomously monitor load information, predict load changes, and make decisions about server operation without external intervention. The system serves itself by automatically adjusting server status based on predicted needs, eliminating the need for manual management and reducing operational complexity.
Data Source
AI summary
A load balancing device which can appropriately control the number of servers in operation according to load is provided. The load balancing device which performs wide area load distribution includes a load prediction/control unit that predicts a load fluctuation of a server on the basis of a change in load information indicating a state of a load on the server and shortens a valid time of the server when the load fluctuation indicates that the load on the server decreases.