Runtime Event Classification via Hierarchical Catalog
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computer systems face challenges in effectively managing and mitigating events, such as exceptions, during runtime due to the inability of existing systems to classify and respond to these events in a manner that aids administrators and users in determining importance, duplications, and known solutions.
Innovation Solution
A computer system classifies events by comparing their characteristics with a hierarchical classification catalog, creating new classification elements if necessary, and takes actions based on predefined tags and notes associated with matching elements, allowing for nuanced responses and automated recovery mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If existing systems log and monitor events, then event tracking capability is improved, but the ability to classify and respond to events meaningfully deteriorates
Solution Approach 1:
The system segments events into distinct classification categories (e.g., authentication failures, authorization failures, resource exhaustion) based on their characteristics. This segmentation enables meaningful classification and targeted responses, transforming raw event logs into structured, actionable information that improves both classification capability and management efficiency.
Solution Approach 2:
The system introduces an intermediary classification layer between event occurrence and administrative response. This intermediary layer analyzes event characteristics, assigns classifications, and triggers appropriate responses, thereby bridging the gap between raw event data and meaningful administrative action.
2Measurement precision
If the system monitors all event characteristics, then event analysis precision is improved, but system complexity increases
Solution Approach 1:
The system applies local quality by focusing analysis on specific, relevant characteristics of events rather than all possible characteristics. It identifies and monitors only the characteristics that are most useful for classification (such as error codes, event types, frequencies), thereby achieving precise analysis without requiring complex monitoring of every possible event attribute.
Solution Approach 2:
The system changes the parameters of event monitoring from raw, unprocessed event data to classified, summarized characteristics. By transforming events into classified categories with aggregated statistics (frequency, timing patterns), the system achieves precise analysis while reducing the complexity of processing individual event characteristics.
3Speed
If the system responds to every event, then responsiveness is improved, but operational overhead increases
Solution Approach 1:
The system applies partial action by responding only to events that meet specific criteria (e.g., critical classifications, threshold frequencies) rather than every event. This selective response maintains responsiveness to important events while reducing processing resources spent on routine or insignificant events.
Solution Approach 2:
The system implements self-service through automated classification and response mechanisms that operate without continuous human intervention. The classification system automatically analyzes event characteristics, assigns categories, and triggers predefined responses, thereby maintaining fast response speeds while reducing the operational overhead of manual event processing.
Data Source
AI summary
A system may be able classify events that occur during the runtime of applications (e.g., exceptions). The system may receive an indication of the event and may classify the event based on a comparison with elements of a classification data structure. The classification data structure may be a hierarchical data structure, and child elements may inherit characteristics from parent elements. Based on the classification, the system may perform one or more actions, which may be specified by the elements of the data structure. For example, the system may provide notifications to administrators and/or user, may attempt to recover from the event, and/or the like. Each event may be associated with a unique identifier so the user can more easily identify the event to support personnel. The system may include analysis tools to assist administrators in tracking events and identifying which events are most important.