Virtual Environment Power State Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The dynamic demand for virtual machines in virtual desktop environments leads to power inefficiencies due to uneven distribution and unnecessary resource allocation, as existing systems automatically start new virtual machines without optimizing the power state of real machines based on workload demands.
Innovation Solution
A power management system that includes a load identifier to assess virtual machine loads and modify the power state of real machines by turning them on, off, or putting them into sleep mode, ensuring optimal resource allocation and utilization based on current and anticipated workloads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If virtual machines are automatically started to meet user demand, then service availability is improved, but power consumption increases due to unnecessary resource allocation
Solution Approach 1:
The system dynamically adjusts the power state of real machines based on current and predicted virtual machine load requirements. Instead of static resource allocation, the system continuously monitors workload patterns and transitions real machines between powered-on, powered-off, and sleep states to match actual demand, thereby improving service availability while reducing unnecessary power consumption.
Solution Approach 2:
The system performs preliminary actions by predicting future virtual machine load requirements and proactively powering on real machines before they are actually needed. This predictive approach ensures that when users log on, virtual machines are already available, improving service availability while avoiding the need to keep all real machines continuously powered on.
2Reliability
If real machines are kept powered on to ensure virtual machine availability, then service reliability is improved, but energy waste increases
Solution Approach 1:
The system implements periodic monitoring and assessment of virtual machine load requirements across different time periods. By analyzing usage patterns and predicting future needs, the system periodically adjusts real machine power states, keeping them powered on only during periods when virtual machines are likely to be needed, thus reducing energy waste while maintaining availability during high-demand periods.
Solution Approach 2:
The system enables real machines to self-manage their power states based on load assessments. Each real machine can transition to sleep or powered-off states when not needed, and wake or power on when predicted to be required, reducing the need for continuous monitoring and manual intervention while minimizing energy waste.
3Ease of operation
If virtual desktop brokering distributes load equally across all hypervisors, then system simplicity is maintained, but resource utilization efficiency deteriorates
Solution Approach 1:
The system applies local quality by allowing different hypervisors to have different power states and resource availability based on their specific roles and current conditions. Instead of uniform load distribution, the system intelligently routes virtual machine creation requests to hypervisors that are currently powered on and have available capacity, optimizing resource utilization while maintaining operational simplicity through automated decision-making.
Solution Approach 2:
The system transforms the static, equal load distribution model into a dynamic model where load routing decisions are made in real-time based on hypervisor power states and capacity. This dynamic approach automatically adapts to changing conditions, directing new virtual machines to appropriate hypervisors without requiring complex manual configuration, thus improving resource utilization while maintaining ease of operation.
Data Source
AI summary
Systems, methods, and software are described herein for operating a power management system including identifying a virtual machine load in a virtual machine environment, identifying a power state for at least one real machine in the virtual machine environment based on the virtual machine load, and modifying the power state for the at least one real machine.