Correlation Analysis for Application Chain Performance Degradation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current monitoring systems for computer infrastructure performance degradation lack real-time correlation analysis of symptoms across application link chains, often requiring repetitive network commands, leading to information loss and precision issues, and rely on administrator intervention and threshold settings.

Innovation Solution

A device with a computer probe that measures and stores resource utilization levels across servers in an application chain during both degradation and normal periods, reducing network congestion and enabling automatic correlation analysis to identify performance issues.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If repetitive commands are sent to monitor server resources, then resource consumption can be detected, but network congestion increases and information accuracy decreases

Engineering Contradiction:
Improveresource consumption detection accuracyVSAvoidnetwork congestion
Core Design Contradiction:
Measurement precisionVSObject-generated harmful factors

Solution Approach 1:

The patent introduces an intermediary agent deployed on each server that collects resource consumption data locally and transmits it to a central server. This intermediary approach eliminates the need for repetitive monitoring commands over the network, as the agent proactively reports data using efficient protocols like SNMP traps or asynchronous callbacks, thereby reducing network congestion while maintaining accurate resource consumption detection.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Loss of information

If command repetition frequency is increased to capture performance degradation, then information accuracy improves, but network load increases

Engineering Contradiction:
Improveperformance degradation information completenessVSAvoidnetwork bandwidth consumption
Core Design Contradiction:
Loss of informationVSLoss of energy

Solution Approach 1:

The monitoring system implements periodic action by configuring the intermediary agent to collect and report resource consumption data at optimized intervals. Instead of continuous or high-frequency command repetition, the agent uses event-driven triggers that activate during performance degradation events, thereby capturing complete information about performance issues while minimizing network bandwidth consumption during normal operation.

Inventive Principle:
Principle #19Periodic action

3Measurement precision

If individual server monitoring is implemented, then resource usage can be tracked, but correlation analysis across application chain is not achieved

Engineering Contradiction:
Improveresource usage measurementVSAvoidcorrelation analysis capability
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges individual server monitoring capabilities into a unified correlation analysis system. The central server receives resource consumption data from multiple intermediary agents across different servers and applications, then performs automated correlation analysis by comparing resource usage patterns, identifying relationships between servers, and detecting performance degradation propagation across the application chain. This combining approach maintains precise resource usage measurement while adding comprehensive correlation analysis capability.

Inventive Principle:
Principle #5Merging (Combining)

4Quantity of substance

If load simulation is used for testing, then performance metrics can be collected, but real-time performance degradation observation is not possible

Engineering Contradiction:
Improveperformance metrics dataVSAvoidreal-time performance monitoring accuracy
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The monitoring system uses self-service by deploying intermediary agents on actual production servers that autonomously collect and report real resource consumption data. Instead of relying on load simulation, the agents continuously monitor actual system performance and transmit authentic data about resource usage, CPU load, memory consumption, and I/O operations. This self-service approach ensures high reliability and accuracy in observing real-time performance degradation events as they occur in the actual production environment.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3163445B1Mechanism for analysing correlation during performance degradation of an application chain
Publication Date: 2021.02.24 BULL SA
  • EP3163445B1 patent drawingFigure 1

AI summary

The present invention relates to a device comprising at least one computer machine and software for implementing a correlation analysis mechanism during performance degradation of an application chain, including a hardware and software arrangement for storing a measurement reference, said measurement reference comprising a hardware and software arrangement enabling the measurement, by consumption probes, of the level of use of each resource on all the servers constituting the application chain during periods of performance degradation, and then the storage in a memory of these levels in the measurement reference, in association with the period, said device being characterized in that it further comprises a hardware and software arrangement for editing a configuration reference, a module for categorizing performance problems according to the measurement and configuration references.