Pre-emptive Container Load-Balancing and Auto-Scaling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Managing computing resources for containerized applications is complex due to varying resource requirements among microservices, making it difficult to track usage, forecast needs, and schedule resources optimally, especially in dynamic environments like cloud and edge deployments.

Innovation Solution

A resource management platform that utilizes telemetry data to track and predict resource usage, attaching this data to container images and using machine learning models to schedule containers on optimal environments, facilitating efficient load balancing, auto-scaling, and placement, while also managing third-party services and bandwidth efficiently.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If containerized applications are used to break monolithic applications into microservices, then resource efficiency is improved, but device complexity increases

Engineering Contradiction:

Improveresource efficiencyVSAvoidmanagement complexity

Core Design Contradiction:

Use of energy by moving objectVSDevice complexity

Solution Approach 1:

The system implements self-service through automated resource tracking, prediction, and allocation. The resource management platform automatically monitors container resource usage, predicts future needs using the trained model, and allocates resources without manual intervention, allowing the containerized system to manage itself despite increased complexity

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system establishes feedback loops by continuously tracking actual resource usage of containers and comparing it with predicted values. The model is retrained periodically with new telemetry data, creating a closed-loop system that adapts to changing conditions and optimizes resource allocation dynamically

Inventive Principle:

Principle #23Feedback

2Adaptability or versatility

If different resource requirements for each microservice are accommodated, then adaptability is improved, but measurement precision deteriorates

Engineering Contradiction:

Improveresource allocation flexibilityVSAvoidresource usage tracking accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system segments resource tracking and prediction at the individual container level. Each container's resource usage is monitored separately, and the machine learning model is trained on container-specific telemetry data, enabling precise measurement of heterogeneous resource requirements while maintaining flexibility in allocation

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system dynamically adjusts resource allocation parameters based on predicted and actual usage patterns. The resource management platform modifies CPU, memory, and other resource allocations in response to changing conditions, maintaining both precision in measurement and flexibility in adaptation

Inventive Principle:

Principle #35Parameter changes

3Device complexity

If manual resource management is performed, then device complexity is reduced, but productivity decreases

Engineering Contradiction:

Improvesystem simplicityVSAvoidresource allocation efficiency

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The resource management platform performs self-service by automatically tracking container resource usage, training prediction models on collected telemetry data, and allocating resources based on model predictions without requiring manual intervention, thereby maintaining high productivity while managing complexity automatically

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system replaces manual mechanical resource management processes with an automated machine learning-based system. The trained model predicts resource needs and triggers automated allocation decisions, substituting human-operated mechanical processes with an intelligent automated system that improves productivity

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

4Measurement precision

If resource usage is tracked in real-time, then measurement precision is improved, but loss of time increases

Engineering Contradiction:

Improveresource usage monitoring accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary action by training the machine learning model in advance using historical telemetry data. The pre-trained model can then quickly predict future resource needs without requiring real-time complex calculations, reducing processing time while maintaining measurement precision through accurate predictions

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11604682B2Pre-emptive container load-balancing, auto-scaling and placement

Publication Date: 2023.03.14 EMC IP HLDG CO LLC

AI summary

A resource usage platform is disclosed. The platform performs preemptive container load balancing, auto scaling, and placement in a computing system. Resource usage data is collected from containers and used to train a model that generates inferences regarding resource usage. The resource usage operations are performed based on the inferences and on environment data such as available resources, service needs, and hardware requirements.