Application State Prediction Using Component Temporal Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for monitoring and predicting application states in technology landscapes are typically reactive, computationally wasteful, and prone to false positives, failing to prevent user experiences of malfunctions and requiring excessive time to resolve issues.
Innovation Solution
A system that utilizes unsupervised machine learning to detect temporal patterns in component states from performance metrics, generating standardized and normalized data for supervised machine learning models to predict future application states with high accuracy, enabling proactive issue prevention and prioritization of corrective actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional reactive monitoring techniques are used to track application and component performance, then problems can be identified after occurrence, but the approach is computationally wasteful and generates false positives that reduce accuracy
Solution Approach 1:
The system performs preliminary analysis by detecting temporal patterns in component states before actual application malfunctions occur. By analyzing historical data and identifying predictive patterns, the system can forecast future application states and trigger alerts before failures happen, enabling proactive rather than reactive monitoring.
Solution Approach 2:
The system extracts only the most relevant features and temporal patterns from the vast amount of performance metric data. By filtering and selecting only the critical patterns that predict application failures, the system reduces computational overhead while maintaining high accuracy in malfunction prediction.
2Reliability
If conventional reactive monitoring tracks all components and metrics, then comprehensive coverage is achieved, but computational resources are wasted on components unlikely to affect user experience
Solution Approach 1:
The system applies different monitoring intensities to different components based on their importance to user experience. By analyzing temporal patterns and component relationships, the system identifies which components have the most significant impact on application performance and directs computational resources accordingly, focusing detailed monitoring on critical components while using lighter monitoring for less important ones.
3Speed
If conventional techniques provide frequent alerts on component states, then real-time monitoring is achieved, but false positives increase and reduce operator effectiveness
Solution Approach 1:
The system continuously learns from historical data and refines its pattern recognition capabilities. By feedback-looping through analysis of actual malfunctions and false alerts, the system improves its temporal pattern detection accuracy over time, reducing false positives while maintaining fast response to genuine issues.
Data Source
AI summary
Described systems and techniques enable prediction of a state of an application at a future time, with high levels of accuracy and specificity. Accordingly, operators may be provided with sufficient warning to avert poor user experiences. Unsupervised machine learning techniques may be used to characterize current states of applications and underlying components in a standardized manner. The resulting data effectively provides labelled training data that may then be used by supervised machine learning algorithms to build state prediction models. Resulting state prediction models may then be deployed and used to predict an application state of an application at a specified future time.