Anomaly Detection Model for Data Center Telemetry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Anomaly detection in data center operations is hindered by the overwhelming number of variables to monitor, making it difficult to identify important variables and understand root causes, leading to potential malfunctions and financial losses.
Innovation Solution
An extra-exogenous anomaly detection model that categorizes system variables and combines them to focus on relevant ones, using a derivative representation of configuration and workload variables to define normality and detect anomalies, thereby improving accuracy and interpretability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all variables are monitored to ensure comprehensive anomaly detection, then detection reliability is improved, but device complexity and difficulty of analysis increase significantly
Solution Approach 1:
The patent segments the monitoring system into two distinct components: (1) a derivative representation module that transforms the second set of variables into a condensed first set of variables, and (2) an anomaly detection module that operates on the reduced variable set. This segmentation allows comprehensive monitoring capabilities while reducing the complexity of the detection process by working with transformed, aggregated variables rather than raw individual variables.
Solution Approach 2:
The patent introduces a derivative representation as an intermediary between the raw second set of variables and the anomaly detection process. This intermediary transforms complex multi-dimensional variables into a reduced first set of variables that capture essential patterns, thereby maintaining detection reliability while reducing the complexity burden on the monitoring system.
2Measurement precision
If all variables are monitored to identify root causes, then measurement precision is improved, but difficulty of detecting and measuring increases
Solution Approach 1:
The patent segments variables into two functional groups: a second set of variables used to create derivative representations and a first set of transformed variables used for actual anomaly detection. This segmentation enables precise anomaly identification by focusing computational efforts on the reduced first set of variables while the second set provides contextual information through the derivative transformation process.
Solution Approach 2:
The patent applies parameter changes by transforming the second set of variables into a derivative representation that produces the first set of variables. This transformation changes the parameter space from raw operational variables to aggregated statistical variables, making anomaly detection more precise while reducing the difficulty of measurement and analysis.
3Loss of information
If comprehensive variable monitoring is implemented, then information completeness is improved, but loss of information through data complexity increases
Solution Approach 1:
The patent extracts essential information from the second set of variables by creating a derivative representation that yields the first set of variables. This extraction process removes redundant and less important information while preserving the critical patterns needed for anomaly detection, thereby maintaining information completeness while reducing the overall data volume that needs to be processed.
Data Source
AI summary
Facilitating detection of anomalies of a target entity is provided herein. A system can comprise a processor and a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations. The operations can comprise training a model on a first set of variables that are constrained by a second set of variables. The second set of variables can characterize elements of a defined entity. The first set of variables can define a normality of the defined entity. The operations also can comprise employing the model to identify expected parameters and unexpected parameters associated with the defined entity to at least a defined level of confidence.