Network Event Detection for Near-Real-Time Root Cause Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network monitoring relies heavily on manual analysis, which is time-consuming, resource-intensive, and requires specialized expertise, often leading to delayed response times and increased operational costs due to missed or misdiagnosed issues, especially with user-configured tests providing arbitrary or suboptimal coverage and the presence of noise.
Innovation Solution
A system that automatically detects network events and identifies the root cause in near-real-time using user-defined tests through a multi-step approach of processing, filtering, and grouping test outcomes, enabling cross-layer, cross-test, and cross-account root-cause analysis, independent of user-configured alerts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis of test results is used to pinpoint root causes, then specialized expertise can be applied to diagnose issues, but the process becomes time-consuming and resource-intensive
Solution Approach 1:
The system performs self-diagnosis by automatically analyzing test results and identifying root causes without requiring manual expert intervention. The automated root cause analysis engine processes test outcomes, correlates failures across multiple tests, and determines probable root causes, enabling the system to serve itself in the diagnostic process while maintaining high accuracy.
Solution Approach 2:
The patent replaces the mechanical process of manual expert analysis with an automated computational system. The root cause analysis engine uses algorithmic processing to correlate test failures, analyze patterns, and identify root causes, substituting human expert mechanics with automated computational mechanics that operate faster and without fatigue.
2Adaptability or versatility
If user-configured tests are used for network monitoring, then specific user needs can be addressed, but coverage becomes arbitrary or suboptimal leading to missed issues
Solution Approach 1:
The system dynamically adjusts test configuration based on observed failures and correlations. When test failures occur, the root cause analysis engine automatically determines which additional tests should be run to gather more diagnostic information, making the monitoring system adaptive and self-optimizing rather than static and arbitrarily configured.
Solution Approach 2:
The system implements feedback loops where test results are continuously analyzed and used to inform subsequent testing decisions. The root cause analysis engine uses outcomes from initial tests to determine what additional tests are needed, creating a closed-loop system that learns from results and improves coverage dynamically, ensuring reliable detection without requiring perfect upfront configuration.
3Measurement precision
If extensive manual analysis is performed to distinguish meaningful data from noise, then accurate diagnosis can be achieved, but operational costs increase
Solution Approach 1:
The system applies partial action by selectively analyzing only the most relevant test failures and correlations rather than performing exhaustive manual analysis of all data. The root cause analysis engine identifies and focuses on the subset of test outcomes that are most likely to contain meaningful diagnostic information, avoiding waste of resources on obvious noise while maintaining detection accuracy.
4Ease of operation
If user-configured tests with arbitrary intervals are used, then user preferences are respected, but baselines fail to yield usable data for anomaly detection
Solution Approach 1:
The system dynamically determines appropriate test intervals and baseline parameters based on the specific network conditions and failure patterns observed. Rather than relying on fixed user-configured intervals that may be arbitrary, the system adapts the monitoring rhythm to match actual network behavior, ensuring baselines are of sufficient quality for anomaly detection while still respecting user-defined test configurations.
Data Source
AI summary
In one implementation, a device obtains test results from a plurality of performance monitoring tests performed in a computer network. The device identifies a set of components of the computer network as potential causes of the test results. The device determines that a particular component from among the set of components caused the test results based on its health metrics. The device raises an alert indicative of the particular component having caused the test results.


