IT Infrastructure Resource Consumption Prediction via Correlation Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for monitoring and managing IT infrastructure resource consumption are inadequate, as they fail to accurately predict resource anomalies like overload or saturation in real-time, leading to potential malfunctions and requiring administrators to base decisions on past data, without providing recommendations for infrastructure development.
Innovation Solution
A method and device that analyze resource consumption by determining modeling functions and correlation functions to predict future resource usage, allowing for real-time anomaly detection and generation of new sizing values to prevent overloads and saturations, incorporating a system with modules for storage, modeling, correlation calculation, and prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current monitoring methods are used to track resource usage, then administrators can detect resource saturation, but they cannot predict future resource consumption anomalies in real-time
Solution Approach 1:
The system performs preliminary actions by building correlation models between resources in advance and continuously monitoring current consumption patterns. This allows the system to predict future resource saturation before it occurs, enabling proactive rather than reactive management. The correlation models are established beforehand to capture relationships between resources, so when current consumption deviates from expected patterns based on these pre-established correlations, anomalies can be detected early.
Solution Approach 2:
The system implements feedback by continuously comparing actual resource consumption against predicted consumption based on correlation models. When deviations are detected, the system generates alerts and can automatically adjust resource allocation. This closed-loop feedback mechanism enables real-time prediction and response to resource anomalies, transforming the system from passive monitoring to active prediction and management.
2Reliability
If administrators rely on past data to anticipate resource risks, then they can make informed decisions, but they cannot detect anomalies in real-time during load increases
Solution Approach 1:
The system replaces the mechanical manual analysis approach with automated computational modeling. Instead of administrators manually analyzing past data trends, the system uses correlation models and automated algorithms to continuously analyze resource consumption patterns, detect deviations, and predict anomalies in real-time. This substitution of manual mechanical analysis with automated computational systems enables both high reliability and fast response speed simultaneously.
3Ease of operation
If resource saturation occurs sequentially in the system, then one resource can be monitored at a time, but subsequent saturations cannot be detected once the first saturation point is reached
Solution Approach 1:
The system implements universality by creating correlation models that simultaneously monitor multiple resources and their interrelationships. Instead of treating each resource in isolation, the correlation models capture how resources interact and affect each other, allowing the system to detect saturation in any resource while considering the state of all other resources. This multi-functional approach enables continuous monitoring of all resources even when one reaches saturation.
4Measurement precision
If correlation models are used to detect resource errors, then relationships between resources can be identified, but the process is time-consuming and does not allow real-time estimation of resource demand
Solution Approach 1:
The system ensures continuity of useful action by continuously updating and applying correlation models to current resource consumption data in real-time. Rather than performing periodic batch analysis, the system maintains continuous monitoring and prediction operations, constantly comparing actual consumption against model predictions. This continuous operation enables both accurate correlation detection and real-time anomaly estimation without time delays.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method and device for analyzing resource consumption in a computer infrastructure for predicting resource consumption anomalies on a computer device, said method comprising: - a step (200) of determining a plurality of resource consumption modeling functions (FMs); - a step (300) of determining the correlation between said resource consumption modeling functions; - a step (400) of measuring resource consumption, said step comprising measuring the consumption value of a first resource; and - a step (500) of predicting the resource consumption of the computer infrastructure, said prediction step comprising calculating a future consumption value of a resource to be predicted from the consumption value of the first resource and a previously calculated correlation between modeling functions (FMs).