Scaling Management Device for Heterogeneous Hardware Power Optimization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Virtualization platforms fail to optimize resource management across heterogeneous hardware environments, leading to inefficient power consumption due to the lack of consideration for performance and power efficiency differences in scaling decisions.

Innovation Solution

A scaling management device that calculates electric power efficiency characteristics and performance ratios for each hardware type, determining optimal usage rates and resource allocation to minimize power consumption by creating config files that account for these differences.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If autoscaling is implemented using fixed CPU usage rate thresholds, then scaling automation is achieved, but heterogeneous hardware environments with different performance and power efficiency characteristics cannot be optimized

Engineering Contradiction:

Improvescaling automationVSAvoidhardware environment adaptability

Core Design Contradiction:

Extent of automationVSAdaptability or versatility

Solution Approach 1:

The patent applies local quality by creating hardware-specific scaling configurations where each hardware type (HW_A, HW_B) has its own target usage rate and performance ratio. Instead of a single global threshold, the system tailors scaling parameters to local hardware characteristics, allowing HW_A to operate at 90% target usage rate while HW_B operates at 60%, optimizing both automation and hardware-specific adaptability.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system dynamically changes scaling parameters based on hardware type. The target usage rate parameter is adjusted from a fixed value to hardware-specific values (90% for HW_A, 60% for HW_B). The performance ratio parameter is introduced to quantify hardware differences, enabling the scaling controller to adaptively modify scaling decisions according to the specific hardware characteristics being monitored.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If virtual resources are scaled to maximize processing capacity, then system performance is improved, but power consumption increases due to servers operating away from their optimal efficiency points

Engineering Contradiction:

Improveprocessing capacityVSAvoidpower consumption

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

The patent implements dynamics by making scaling decisions adaptive rather than static. The scaling controller continuously monitors usage rates and dynamically adjusts virtual resource allocation based on current hardware performance and power efficiency characteristics. This allows the system to optimize the balance between processing capacity and power consumption in real-time, rather than operating at fixed capacity levels.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system employs feedback mechanisms where the scaling controller receives usage rate information from the monitoring unit, compares it against hardware-specific target usage rates, and adjusts scaling decisions accordingly. This closed-loop feedback ensures that servers operate near their optimal efficiency points (HW_A at 90%, HW_B at 60%) while maintaining required processing capacity, thereby reducing energy loss.

Inventive Principle:

Principle #23Feedback

3Ease of operation

If uniform scaling thresholds are applied across all hardware, then system management is simplified, but heterogeneous hardware performance differences are not accounted for

Engineering Contradiction:

Improvesystem management simplicityVSAvoidhardware performance measurement

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent applies segmentation by dividing the scaling management system into hardware-specific configuration segments. Each hardware type (HW_A, HW_B) has its own scaling configuration file with tailored parameters. The scaling controller selectively applies the appropriate configuration segment based on the monitored hardware type, maintaining management simplicity through modular organization while achieving precise hardware-specific performance measurement and optimization.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS20250004842A1Scaling management device, scaling management method, and program

Publication Date: 2025.01.02 NIPPON TELEGRAPH & TELEPHONE CORP

AI summary

A scaling management device includes an electric power efficiency characteristic calculation unit that calculates electric power efficiency characteristics of hardware and determines a usage rate at which a value of electric power efficiency is highest, a performance ratio calculation unit that measures performance of hardware, identifies hardware having a lowest performance value, and calculates a performance ratio between the identified hardware and other hardware, a virtual resource number calculation unit that calculates a score obtained by multiplying the performance ratio by the usage rate at which the value of electric power efficiency is highest and calculates the number of virtual resources in accordance with the score, a requested resource calculation unit that calculates requested resources of virtual resources, and, a config file creation unit that creates a scaling config file and a resource config file.