Predictive Anomaly Detection in Distributed Systems

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In distributed information processing systems, anomalies and faults are difficult to detect and isolate, especially in environments with microservices, where causal issues are challenging to identify and resource allocation for detection and remediation is insufficient.

Innovation Solution

An apparatus and method utilizing an unsupervised machine learning model for predictive anomaly detection and fault isolation, combining application-level and hardware-level data to monitor and isolate issues in a multicloud edge platform, employing AI techniques for continuous and scalable monitoring.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If distributed information processing systems are deployed to process vast amounts of data across multiple computing devices, then productivity and data processing capability are improved, but anomaly detection and fault isolation become more difficult

Engineering Contradiction:

Improvedata processing capabilityVSAvoidanomaly detection difficulty

Core Design Contradiction:

ProductivityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent introduces an intermediary system that collects and analyzes telemetry data from multiple distributed computing devices. This intermediary acts as a mediator between the distributed system components and the anomaly detection process, aggregating data from various sources and applying machine learning models to identify anomalies, thereby making detection feasible in distributed environments

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces traditional mechanical or rule-based anomaly detection methods with AI-based machine learning models. These models automatically analyze telemetry data patterns to detect anomalies, substituting manual or simplistic detection mechanisms with intelligent systems capable of handling the complexity of distributed environments

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Device complexity

If traditional monitoring methods are used in distributed systems, then device complexity is reduced, but measurement precision and anomaly detection capability deteriorate

Engineering Contradiction:

Improvesystem simplicityVSAvoidanomaly detection precision

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent changes the parameters of the monitoring system by introducing AI-based analysis capabilities and multi-source telemetry data collection. Instead of using simple threshold-based monitoring, the system employs machine learning models that analyze multiple parameters simultaneously, significantly improving detection precision while maintaining manageable system complexity through automated processes

Inventive Principle:

Principle #35Parameter changes

3Measurement precision

If AI-based predictive anomaly detection is implemented, then anomaly detection precision is improved, but device complexity and resource requirements increase

Engineering Contradiction:

Improveanomaly detection precisionVSAvoidmonitoring system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the monitoring system into modular components: telemetry data collection modules, data aggregation layers, machine learning model components, and fault isolation mechanisms. This segmentation allows the complex AI-based system to be implemented in manageable parts, reducing the burden on individual devices while maintaining overall system precision

Inventive Principle:

Principle #1Segmentation

4Measurement precision

If comprehensive telemetry data collection is performed across distributed devices, then measurement precision is improved, but data volume and processing requirements increase

Engineering Contradiction:

Improvefault detection precisionVSAvoiddata volume

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts only the most relevant telemetry data features needed for anomaly detection using AI-based feature selection. Instead of processing all collected data, the system identifies and extracts key indicators that are most predictive of anomalies, reducing the data volume that requires intensive processing while maintaining detection precision

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20240281359A1Predictive anomaly detection and fault isolation in information processing system environment

Publication Date: 2024.08.22 DELL PROD LP

AI summary

Application monitoring techniques comprising predictive anomaly detection and fault isolation are disclosed for use in an information processing system environment. For example, an apparatus comprises at least one processing device comprising a processor coupled to a memory. The processing device is configured to obtain application-level data generated for an information processing system in accordance with execution of an application and obtain hardware-level data generated for the information processing system in accordance with the execution of the application. The processing device is further configured to utilize an unsupervised machine learning model to predictively detect anomalous behavior in accordance with the execution of the application in the information processing system, based on at least a portion the application-level data and the hardware-level data. The processing device may also initiate fault isolation in addition to predicting anomalous behavior.