Cloud Service Resource Prediction for Cost-Constrained SLA Satisfaction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing service level management systems for cloud services rely on human experience to determine IT resources, leading to cases where the desired service level agreement is not met, resulting in unattainable desired results.
Innovation Solution
A prediction system that acquires candidates influencing the ease of obtaining a desired result in a service, predicts whether the desired result is achieved using a predetermined method, and identifies the candidates likely to achieve the desired result while minimizing costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If IT resources are determined based on human experience, then the service level agreement may not be satisfied, but increasing resources based on experience leads to higher costs
Solution Approach 1:
The patent replaces human experience-based resource determination with an automated prediction system that uses machine learning models and algorithms to predict service level agreement satisfaction. The system automatically processes service definitions, resource configurations, and performance data to generate predictions, eliminating reliance on subjective human judgment and enabling more accurate, data-driven resource allocation decisions.
Solution Approach 2:
The prediction system enables the service management system to automatically evaluate and determine resource requirements without human intervention. The system self-assesses service level agreement satisfaction by processing input data through prediction models and generating recommendations, allowing the system to serve itself in optimizing resource allocation based on objective criteria rather than human experience.
2Reliability
If more IT resources are allocated to ensure service level agreement, then the desired result is more likely to be obtained, but the cost increases
Solution Approach 1:
The patent changes the parameter of resource determination from fixed human experience-based allocation to dynamic prediction-based allocation. The system adjusts resource recommendations by changing key parameters such as processing power, memory, storage, and network bandwidth based on predicted service level agreement satisfaction, enabling optimal resource allocation that minimizes cost while ensuring reliability requirements are met.
Solution Approach 2:
The patent substitutes manual resource allocation decisions with an automated prediction system that uses machine learning models to evaluate the relationship between resource allocation and service level agreement satisfaction. This substitution enables the system to identify the minimum necessary resources to achieve desired reliability outcomes, reducing unnecessary resource allocation and associated costs.
3Ease of operation
If human experience is used to determine service level agreement, then the process is simple, but the desired result is not always obtainable
Solution Approach 1:
The patent replaces simple but unreliable human experience-based determination with an automated prediction system that maintains ease of operation through automation. The system accepts service definitions and resource configurations as input and automatically generates prediction results, preserving operational simplicity while significantly improving reliability through objective, data-driven analysis rather than subjective human judgment.
Solution Approach 2:
The prediction system enables automatic evaluation of service level agreement satisfaction without requiring human expertise or manual analysis. The system independently processes input data, applies prediction models, and generates recommendations, making the complex task of resource optimization accessible and easy to operate while ensuring reliable outcomes through systematic analysis.
Data Source
AI summary
A prediction system, with at least one processor configured to: acquire a plurality of candidates of information influencing an ease of obtaining a desired result in a predetermined service; predict, for each of the plurality of candidates, based on a predetermined prediction method, whether the desired result is obtained; and identify the candidates predicted to obtain the desired result.