VM Power Management via Shaving and Pooling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud computing systems face inefficiencies in power resource utilization, with servers often operating below full capacity to maintain reserve power, leading to increased costs and suboptimal usage of existing power resources.
Innovation Solution
A power management system that receives metadata for virtual machines, identifies upcoming limited power events, and implements power shaving actions such as power capping and shedding to maximize power utilization while ensuring service level agreements are met, by grouping virtual machines into pools based on priority and service guarantees.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If server devices operate at or below 60% power capacity to maintain reserve power, then reliability is improved, but power utilization efficiency deteriorates
Solution Approach 1:
The system performs preliminary actions by identifying upcoming limited power events before they occur and proactively shedding non-critical workloads or migrating virtual machines. This allows the system to maintain higher power utilization (over 90% capacity) during normal operation while ensuring service level agreements are met during actual power limitations, resolving the contradiction between reliability and energy efficiency.
Solution Approach 2:
The power management system continuously monitors power capacity, workload priorities, and service level agreement compliance, using this feedback to dynamically adjust power allocation and virtual machine placement. This feedback mechanism enables the system to optimize power utilization while maintaining reliability by adapting to changing conditions in real-time.
2Productivity
If additional servers and datacenters are installed to meet customer demand, then productivity is improved, but cost increases
Solution Approach 1:
The system enables existing power infrastructure to serve multiple functions by dynamically allocating power to different workloads based on priority and availability. During normal operation, power is allocated to maximize utilization; during limited power events, power is reallocated to critical services. This multi-functional approach allows the same physical infrastructure to meet varying demand levels without requiring additional servers or datacenters.
Solution Approach 2:
The power management system dynamically adjusts power allocation, virtual machine placement, and workload distribution based on real-time conditions. This dynamic optimization allows the system to extract maximum productivity from existing power resources, eliminating the need for additional infrastructure investments to meet customer demand.
3Loss of energy
If power shaving actions are implemented to maximize power utilization, then power capacity utilization is improved, but service availability may deteriorate
Solution Approach 1:
The system applies different power management strategies to different workloads based on their priority and service level agreement requirements. Critical workloads receive guaranteed power allocation and are protected from shedding, while non-critical workloads are subject to power shaving and migration. This localized quality approach allows the system to achieve high overall power utilization while maintaining service availability for critical functions.
Solution Approach 2:
The system performs preliminary actions by preemptively migrating or shedding non-critical workloads before limited power events occur. This allows the system to safely maximize power utilization during normal operation knowing that critical services will maintain service availability even during power constraints.
Data Source
AI summary
The present disclosure relates to systems, methods, and computer readable media for enabling server devices to utilize a higher percentage of power resources while maintaining sufficient availability of power resources of a datacenter or other collection of server devices. For example, systems disclosed herein determine and implement power shaving actions based on virtual machine metadata and in accordance with a power shaving policy to facilitate a significantly higher utilization of power resources on a datacenter during normal operation as well as within periods of limited power capacity on various server devices. Systems described herein provide more efficient utilization of power resources while maintaining service availability guarantees for a variety of virtual machines hosted by servers of the datacenter.