VM Disposition Planning for Moving Object Simulation Load Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing load prediction techniques for moving object simulations rely on history information, which is unreliable due to changing simulation scenarios, making it difficult to accurately predict and manage server loads without historical data.
Innovation Solution
A planning system that generates a migration plan for virtual machines based on moving routes, predicting loads on physical machines and allocating them to divided geographical areas, allowing for proactive server management without relying on history information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If history information is used for load prediction, then prediction accuracy may be improved in stable scenarios, but the system becomes unreliable when simulation scenarios change
Solution Approach 1:
The system performs preliminary actions by predicting virtual machine migration patterns and physical server loads before the simulation actually executes. The load prediction unit calculates expected loads on physical servers based on planned virtual machine migrations, enabling proactive resource allocation and avoiding the need to rely on historical data from previous simulations.
2Reliability
If additional physical servers are provided in advance, then load capacity is ensured, but calculation cost and financial cost increase
Solution Approach 1:
The system changes the parameter of server allocation from static (provisioning servers for maximum expected load) to dynamic (adjusting server allocation based on predicted actual load). The load prediction unit provides accurate load estimates, allowing the system to allocate the minimum necessary server resources while ensuring load capacity, thereby reducing both calculation costs and financial costs.
3Productivity
If virtual machines are migrated frequently to balance load, then resource utilization improves, but system complexity and migration overhead increase
Solution Approach 1:
The system performs preliminary migration planning before virtual machines actually migrate. The migration plan generation unit determines optimal migration targets in advance based on predicted physical server loads, and the load prediction unit provides accurate load forecasts. This preliminary action reduces migration overhead and simplifies migration management by avoiding frequent reactive migrations.
Data Source
AI summary
Provided is a system for planning the disposition of a virtual machine which predicts a load without history information. This planning system includes: a migration plan generating means for generating, on the basis of a moving route planned by a moving object, a virtual machine migrating plan which represents a moving object in a moving object simulation; and a dividing plan generating means for predicting, on the basis of the virtual machine migrating plan, a load of a physical machine where the virtual machine operates, and allocating the physical machine in a district obtained by dividing a geographical area in the moving object simulation.