Virtual Directed Graphs for Root Cause Prediction in Complex Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In complex systems like information technology infrastructures, determining the root cause of failures is challenging due to numerous contributing factors, making it difficult to form causal correlations between events.
Innovation Solution
A method involving the collection of historical event data to generate undirected and directed graphs, analyzing probabilistic correlations, and reducing the graph to identify the most likely causative event, which allows for real-time recommendations on potential root causes and risk assessment of 'black swan' events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional event correlation methods are used in complex systems, then the analysis process becomes manageable, but the accuracy of identifying root causes deteriorates due to numerous contributing factors
Solution Approach 1:
The patent segments the complex system into a graph structure where events are nodes and correlations are edges. This segmentation transforms the unmanageable complex system into discrete, analyzable units that can be processed individually while maintaining their relational context.
Solution Approach 2:
The patent transitions from traditional linear event analysis to a multi-dimensional graph-based representation. By introducing spatial dimensions (graph nodes and edges) and probabilistic dimensions (correlation strengths), the system can handle complex relationships that cannot be captured in traditional linear analysis.
2Measurement precision
If comprehensive historical event data is collected and analyzed, then the accuracy of root cause determination improves, but the computational time and resources required increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing historical event data to build the correlation graph structure in advance. This includes identifying all possible event correlations and encoding them in the graph before actual root cause analysis is needed, so that when an event occurs, the analysis can proceed quickly by querying the pre-built structure.
Solution Approach 2:
The patent creates a simplified copy or model of the complex system in the form of a correlation graph. This graph copy captures the essential relationships and probabilities without requiring analysis of all raw historical data in real-time, enabling fast querying and root cause identification.
3Reliability
If detailed probabilistic correlations are computed between all events, then the reliability of recommendations improves, but the device complexity and computational requirements worsen
Solution Approach 1:
The patent changes the parameters of the correlation representation by using conditional probabilities and information measures (such as mutual information) to quantify relationships. These parameter transformations allow the system to efficiently compute and compare correlations without requiring exhaustive analysis of all possible event combinations.
Data Source
AI summary
An approach to root cause determination in a complex systems based on monitoring and event data is disclosed. It includes a historical analysis of events with their probabilistic correlations. Applying information measures between the random variables which embody those events one can detect origins of problems and generate real-time recommendations for their locations in a hierarchical system. Estimation of system bottlenecks, as well as the risk of “black swan”-type events are also computed. The processes are based on a statistical processing of a virtual directed graph produced from historical events.