Transaction Path Node Event Relevance Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern web applications face difficulties in tracing the root cause of poor transaction performance due to complex reports and alerts that do not adequately indicate the cause of issues, and existing methods for troubleshooting slow transactions often dilute significant performance aberrations in aggregations or fail to account for the prevalence of transaction properties.
Innovation Solution
A method that identifies transaction-path nodes as problem nodes based on end-to-end response times, determines event types, and infers event-relevance weights from both abstract and concrete probabilistic graphical models to identify the most-relevant events, facilitating a comparative analysis of transaction properties and their correlation with performance outcomes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If typical reports and alerts are generated to monitor transactions, then transaction monitoring coverage is provided, but the reports become complex and fail to adequately indicate root cause
Solution Approach 1:
The patent extracts only the most relevant events and transaction properties from the complex monitoring data using probabilistic relevance weights. Instead of presenting all monitoring information, the system identifies and extracts specifically those events with highest relevance to performance degradation, thereby reducing information loss while avoiding complex report structures.
Solution Approach 2:
The patent introduces probabilistic graphical models as an intermediary layer between raw transaction data and final reports. These models compute relevance weights that mediate between complex monitoring data and simplified root cause indicators, allowing the system to maintain comprehensive monitoring coverage while presenting simplified, actionable information to users.
2Measurement precision
If aggregation is used to analyze transaction performance, then overall performance trends are visible, but significant performance aberrations are diluted
Solution Approach 1:
The patent applies local quality by computing relevance weights specifically for events and properties that exhibit abnormal behavior. Instead of uniform aggregation, the system identifies local deviations from normal performance patterns and assigns higher relevance weights to these anomalous events, thereby preserving precision in detecting performance aberrations while maintaining analysis efficiency through targeted rather than exhaustive examination.
Solution Approach 2:
The patent uses partial action by focusing computational resources on analyzing only those events and transaction properties that show signs of performance degradation. Rather than aggregating and analyzing all transaction data equally, the system applies excessive analytical action selectively to suspicious events identified through relevance weighting, thereby maintaining high measurement precision for aberrations while improving overall analysis efficiency.
3Loss of information
If all events are monitored to ensure comprehensive coverage, then complete visibility is achieved, but identifying relevant events becomes difficult
Solution Approach 1:
The patent changes the parameter of event selection from binary (monitored/not monitored) to continuous (relevance weight). By computing probabilistic relevance weights for all monitored events based on their correlation with performance degradation, the system maintains comprehensive monitoring coverage while transforming the complexity of identifying relevant events into a straightforward ranking process based on weight values.
Data Source
AI summary
In one embodiment, a method includes identifying at least one transaction-path node as a problem node based, at least in part, on an analysis of end-to-end response times for a group of transactions. The method further includes determining one or more event types for the at least one transaction-path node. Also, the method includes, for each of the one or more event types, inferring a first event-relevance weight from an abstract model. The method also includes, for each of the one or more event types, inferring a second event-relevance weight from a concrete model. Furthermore, the method includes, for each of the one or more event types, determining an event relevance based, at least in part, on the first event-relevance weight and the second event-relevance weight. Additionally, the method includes identifying most-relevant events among a set of active events based, at least in part, on the determined event relevance.