Cloud CPU Utilization Measurement via VM Metrics

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Cloud computing clients face challenges in managing and optimizing resource utilization due to limited visibility and control over underlying physical infrastructure, leading to potential performance issues and increased costs, as they share resources with other consumers.

Innovation Solution

The methods involve determining CPU utilization in a cloud data center by monitoring unhalted CPU cycles, CPU temperature, or artificially bursting resources to assess usage, allowing clients to decide whether to transfer their applications to less burdened physical machines for improved quality of service.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If multiple virtual machines are hosted on shared physical infrastructure, then resource utilization increases and cost decreases, but visibility and control over physical infrastructure usage deteriorates

Engineering Contradiction:

Improveresource utilizationVSAvoidvisibility into physical infrastructure

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent introduces an intermediary measurement mechanism that indirectly observes physical CPU usage through virtual machine performance metrics. By using the virtual machine as a mediator to detect underlying physical resource utilization, the system gains visibility without requiring direct access to physical infrastructure, thus resolving the contradiction between shared resource utilization and visibility.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces direct mechanical monitoring of physical CPU resources with an indirect software-based measurement approach. Instead of physically accessing CPU metrics, the system uses virtual machine performance data and measurement tools to infer physical resource usage, substituting a direct physical measurement system with an indirect computational one.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Loss of energy

If cloud computing clients share physical resources with other consumers, then infrastructure costs decrease, but service quality and performance stability deteriorate

Engineering Contradiction:

Improveinfrastructure costsVSAvoidservice quality

Core Design Contradiction:

Loss of energyVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where clients continuously measure physical CPU utilization through their virtual machines and use this information to make informed decisions about resource allocation and migration. This feedback loop allows clients to monitor service quality metrics and take corrective actions when performance degradation is detected, thus maintaining reliability while benefiting from shared infrastructure.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent enables cloud clients to perform self-service monitoring and assessment of their own resource utilization. By providing tools for clients to independently measure and evaluate their physical infrastructure usage, the system empowers clients to manage their own service quality without requiring additional infrastructure investment, thus maintaining reliability while reducing costs.

Inventive Principle:

Principle #25Self-service

3Adaptability or versatility

If cloud computing clients lack control over physical infrastructure configuration, then deployment flexibility increases, but ability to optimize resource allocation deteriorates

Engineering Contradiction:

Improvedeployment flexibilityVSAvoidcontrol over resource allocation

Core Design Contradiction:

Adaptability or versatilityVSEase of operation

Solution Approach 1:

The patent empowers cloud clients with self-service capabilities to measure, monitor, and optimize their own resource allocation. By providing clients with direct measurement tools and visibility into their physical infrastructure usage, the system enables clients to independently optimize their resource allocation strategies without requiring control over the underlying physical infrastructure configuration, thus maintaining deployment flexibility while improving operational control.

Inventive Principle:

Principle #25Self-service

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

These methods enable cloud computing clients to accurately assess CPU utilization and make informed decisions about resource allocation, ensuring high-quality service and potentially reducing costs by optimizing resource usage.

Implementation Method 1

determining the amount of CPU utilization of a physical machine in a cloud data center hosting a VM of a consumer by determining the temperature of the CPU during a designated or target duration of time based on the correlation between CPU temperature and usage

Methodology Applied

Scientific EffectThermal energy generation from CPU operation: Joule Heating

Data Source

PatentUS9749207B2Methods for measuring physical CPU utilization in a cloud computing infrastructure

Publication Date: 2017.08.29 ACCENTURE GLOBAL SERVICES LTD

AI summary

Novel techniques are provided to determine concurrent hardware resource usage as expressed in activity performed by hardware processors. A cloud computing consumer can verify the level of the quality of service provided by the physical infrastructure of a cloud, thereby allowing the consumer the ability to request a transfer of the hosting physical infrastructure to a less burdened physical machine.