Outage Risk Detection via Event Aggregation and Threshold Alerts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Organizations face challenges in determining whether intermittent performance issues with third-party computer services indicate an imminent operation outage, due to the complexity and noise in monitoring systems.
Innovation Solution
An outage risk detection system that ingests events from multiple organizations, rejects events not meeting a predetermined acceptance rate, identifies services with current incidents requiring human interaction, aggregates incident counts across time windows, and generates alerts when the aggregate count surpasses a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system monitors all events from multiple organizations, then the detection coverage is improved, but the false positive rate increases due to common intermittent issues that do not indicate actual outages
Solution Approach 1:
The patent applies local quality by differentiating between types of events based on their characteristics. It specifically identifies events requiring human interaction versus those that can be automatically resolved, and applies different monitoring thresholds and alerting behaviors to each category. This selective approach improves detection accuracy for critical outages while filtering out noise from routine intermittent issues.
Solution Approach 2:
The system performs preliminary action by pre-classifying events into categories based on historical data and event characteristics before they trigger alerts. By预先 identifying which event types are indicative of actual outages versus routine issues, the system prepares filtering rules in advance that reduce false positives when monitoring ongoing events across multiple organizations.
2Reliability
If the system aggregates incident counts across multiple organizations and time windows, then the reliability of outage risk detection is improved, but the complexity of the monitoring system increases
Solution Approach 1:
The patent applies segmentation by dividing the monitoring system into distinct functional components: event ingestion modules, event classification modules, aggregation modules with time window management, and alert generation modules. This modular architecture manages complexity by organizing functions into separate, manageable segments that can be independently configured and maintained while achieving reliable multi-organization monitoring.
Solution Approach 2:
The system introduces another dimension by adding time window aggregation to the monitoring approach. Instead of simply counting incidents across organizations, it aggregates data across multiple time dimensions (e.g., rolling windows, fixed intervals), creating a multi-dimensional view of incident patterns that improves detection reliability without proportionally increasing system complexity.
3Loss of time
If the system processes and analyzes events in real-time, then the response time to detected outages is improved, but the computational resources required increase
Solution Approach 1:
The patent applies partial action by selectively processing only those events that meet predefined criteria for potential outages. Rather than analyzing every event in detail, the system performs lightweight filtering to identify candidate events, then applies more intensive analysis only to those that exhibit characteristics of actual outages. This reduces overall computational resource consumption while maintaining real-time detection capability for critical issues.
Data Source
AI summary
An outage risk detection system ingests events. Each event indicates a condition detected by a monitoring tool within a computing environment. Events that do not meet a pre-determined acceptance rate are rejected by the system. The system monitors the ingested events to identify computer services experiencing incidents. The identified computer services are such that they historically require human intervention to resolve. The system aggregates a count of organizations experiencing service incidents for a particular computer service across multiple time windows. The system generates an outage risk detection alert when the aggregated count surpasses a specified threshold level.