Predictive Autoscaling for Virtualized Resource Groups

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Organizations face challenges in optimizing the use of computing resources provided by service provider networks, as resources often go underutilized due to varying workload demands and cyclical usage patterns, leading to inefficiencies and increased costs.

Innovation Solution

A capacity forecasting and scheduling service is implemented to monitor and predict resource usage patterns, allowing for intelligent allocation of excess resources to other workloads, while ensuring minimal disruption to primary workloads, and providing graphical user interfaces for users to manage and visualize resource utilization.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If computing resources are allocated to multiple workloads with varying demands, then resource utilization efficiency improves, but resource allocation complexity increases

Engineering Contradiction:

Improveresource utilization efficiencyVSAvoidresource allocation complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system implements self-service through automated predictive autoscaling that uses machine learning models to forecast workload demands and automatically adjusts resource allocation without manual intervention. The intelligent agent monitors usage patterns and autonomously provisions or deprovisions resources based on predictions, enabling the system to serve itself and eliminating the need for complex manual resource management.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual resource allocation mechanisms with automated computational systems. Machine learning algorithms substitute for human decision-making in resource provisioning, using historical data and predictive analytics to automatically determine optimal resource levels. This substitution of mechanical/manual processes with automated computational intelligence reduces allocation complexity while improving utilization efficiency.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If resources are provisioned for peak demand, then workload performance is maintained, but resource waste increases during low utilization periods

Engineering Contradiction:

Improveworkload performanceVSAvoidresource waste

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The system implements dynamic resource provisioning that adapts to changing workload demands. Instead of static over-provisioning, the predictive autoscaling mechanism continuously adjusts resource allocation based on forecasted demand, transitioning between different resource levels as conditions change. This dynamic approach maintains performance during peak periods while reducing provisioning during low-demand periods, eliminating the trade-off between reliability and waste.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system performs preliminary actions by predicting future workload demands before they occur. The machine learning models analyze historical patterns and forecast upcoming demand spikes, allowing the system to proactively provision resources in advance of actual need. This preliminary provisioning ensures performance reliability when demand increases while avoiding continuous over-provisioning, as resources are allocated ahead of time based on predictions rather than constant maximum provisioning.

Inventive Principle:

Principle #10Preliminary action

3Measurement precision

If manual resource management is used, then control precision is maintained, but time consumption and operational overhead increase

Engineering Contradiction:

Improvecontrol precisionVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system implements continuous feedback loops where the intelligent agent monitors actual workload performance and resource usage, compares outcomes with predictions, and uses this feedback to refine future resource allocation decisions. Machine learning models continuously learn from observed patterns and adjust their predictions accordingly, maintaining high control precision while operating autonomously. This feedback-driven automation achieves precise resource management without manual intervention, eliminating the trade-off between control precision and time consumption.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS11249810B2Coordinated predictive autoscaling of virtualized resource groups

Publication Date: 2022.02.15 AMAZON TECH INC

AI summary

Techniques are described for optimizing the allocation of computing resources provided by a service provider network—for example, compute resources such as virtual machine (VM) instances, containers, standalone servers, and possibly other types of computing resources—among computing workloads associated with a user or group of users of the service provider network. A service provider network provides various tools and interfaces to help businesses and other organizations optimize the utilization of computing resource pools obtained by the organizations from the service provider network, including the ability to efficiently schedule use of the resources among workloads having varying resource demands, usage patterns, relative priorities, execution deadlines, or combinations thereof. A service provider network further provides various graphical user interfaces (GUIs) to help users visualize and manage the historical and scheduled uses of computing resources by users' workloads according to user preferences.