Machine Learning Circuit Breaker Candidate Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional application management approaches are time-consuming and resource-intensive when identifying circuit breaker patterns in cloud native applications, leading to application downtime.

Innovation Solution

The use of machine learning techniques to determine candidates for circuit breaker patterns in cloud applications by analyzing consumption and production data, training classifiers, and performing automated actions based on the determined composition of application programming interfaces.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional application management approaches are used to identify circuit breaker patterns, then debugging can be performed, but the process is time-consuming and resource-intensive leading to application downtime

Engineering Contradiction:

Improveapplication availabilityVSAvoiddebugging time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary analysis by training machine learning classifiers on consumption-related data before production issues occur. The trained classifiers then automatically analyze production-related data to identify circuit breaker candidates proactively, eliminating the need for time-consuming reactive debugging and reducing application downtime.

Inventive Principle:

Principle #10Preliminary action

2Difficulty of detecting and measuring

If conventional debugging processes are used, then issues can be identified, but the process is resource-intensive

Engineering Contradiction:

Improveissue detection capabilityVSAvoidcomputational resources

Core Design Contradiction:

Difficulty of detecting and measuringVSUse of energy by moving object

Solution Approach 1:

The patent replaces manual, resource-intensive debugging processes with an automated machine learning-based system. The trained classifiers automatically analyze API composition and identify circuit breaker candidates, significantly reducing computational resource consumption while maintaining or improving detection capability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Ease of repair

If reactive issue debugging is performed, then problems can be solved, but application downtime occurs

Engineering Contradiction:

Improveissue resolutionVSAvoidapplication downtime

Core Design Contradiction:

Ease of repairVSDuration of action of moving object

Solution Approach 1:

The system performs preliminary classification and candidate identification before production failures occur. By training classifiers on consumption data and proactively analyzing production data, the system identifies circuit breaker opportunities in advance, enabling preventive rather than reactive repairs and minimizing application downtime.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20230013075A1Determining candidates for circuit breaker patterns in cloud applications using machine learning techniques

Publication Date: 2023.01.19 DELL PROD LP

AI summary

Methods, apparatus, and processor-readable storage media for determining candidates for circuit breaker patterns in cloud applications using machine learning techniques are provided herein. An example computer-implemented method includes obtaining consumption-related data from first cloud-based applications; obtaining production-related data from second cloud-based applications; training at least one machine learning-based classifier using at least a portion of the consumption-related data; determining composition of one or more cloud-based application programming interfaces by processing at least a portion of the production-related data using the at least one trained machine learning-based classifier; determining one or more candidates, within one or more portions of the one or more cloud-based application programming interfaces, for implementation of at least one circuit breaker pattern based at least in part on the determined composition of the one or more cloud-based application programming interfaces; and performing automated actions based at least in part on the determined candidates.