IT Resource Performance Analysis via Statistical Distribution Ranking

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current IT performance monitoring systems face challenges in discovering existing performance problems in IT resources, particularly in large cloud computing infrastructures, as they rely on predictive algorithms based on image analysis and statistical analysis that require significant computational power and are not effective in identifying issues with abnormal distributions or correlations between performance indicators.

Innovation Solution

A method that uses historical data analysis to identify performance problems by ranking IT infrastructure components based on anti-patterns, leveraging advanced statistical analysis and random sampling, without assuming a multivariate normal distribution, and focusing on single performance indicators to create a scorecard for deeper analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If image analysis and statistical analysis algorithms are used for predictive monitoring, then system behavior prediction capability is improved, but computational power requirements increase significantly

Engineering Contradiction:

Improvesystem behavior prediction capabilityVSAvoidcomputational power requirements

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent extracts only the essential statistical features needed for anomaly detection rather than performing complete image analysis. By taking out only the critical measurement data points and their statistical relationships, the system achieves prediction capability with reduced computational overhead, directly resolving the contradiction between reliability and energy consumption.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies partial action by focusing statistical analysis on specific critical parameters and their correlations rather than analyzing all system parameters comprehensively. This selective approach maintains sufficient prediction reliability for critical failures while significantly reducing the computational power required compared to exhaustive analysis methods.

Inventive Principle:

Principle #16Partial or excessive action

2Measurement precision

If time sliding windows are used to analyze evolution over time, then temporal pattern detection is improved, but data processing complexity increases

Engineering Contradiction:

Improvetemporal pattern detectionVSAvoiddata processing complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary statistical processing on measurement data before detailed analysis. By pre-calculating statistical descriptors and organizing data in advance, the system improves temporal pattern detection while reducing the complexity of subsequent processing operations, as the heavy lifting is done in a structured preliminary stage.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent segments the time series data into manageable measurement points with specific statistical descriptors. By dividing the continuous data stream into discrete, analyzable segments with defined statistical properties, the system achieves precise temporal pattern detection while keeping processing complexity manageable through structured segmentation.

Inventive Principle:

Principle #1Segmentation

3Quantity of substance

If image compression techniques are applied for data analysis, then data storage efficiency is improved, but computational power requirements increase and artificial bias is introduced

Engineering Contradiction:

Improvedata storage efficiencyVSAvoidcomputational power requirements

Core Design Contradiction:

Quantity of substanceVSUse of energy by moving object

Solution Approach 1:

The patent changes the parameters of data representation from raw measurement values to statistical descriptors (mean, standard deviation, percentiles). This parameter transformation achieves efficient data storage and processing without requiring compression algorithms, thereby improving storage efficiency while avoiding the computational overhead and artificial bias introduced by compression techniques.

Inventive Principle:

Principle #35Parameter changes

4Ease of manufacture

If multivariate normal distribution assumptions are made for statistical analysis, then analysis simplicity is improved, but detection accuracy for abnormal distributions decreases

Engineering Contradiction:

Improveanalysis simplicityVSAvoiddetection accuracy for abnormal distributions

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent employs dynamic statistical analysis that adapts to the actual distribution characteristics of the data rather than assuming a fixed multivariate normal distribution. By dynamically adjusting the analysis approach based on observed data patterns, the system maintains analytical simplicity while significantly improving detection accuracy for abnormal distributions that deviate from normal assumptions.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS10031829B2Method and system for it resources performance analysis

Publication Date: 2018.07.24 KYNDRYL INC

AI summary

There is disclosed a method of monitoring an infrastructure comprising managed units, the method comprising the steps of: acquiring data associated with a first performance indicator from a first managed unit; determining a first quantized distribution function of at least a subset of pieces of data of the acquired data of the first managed unit; determining if the first quantized distribution function verifies one or a plurality of predefined rules describing particular distribution functions of performance indicators.