Resource Baseline Deviation Detection for Application Health
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current application management systems lack the ability to semantically understand data streams, making it difficult to determine the source or cause of unhealthy application states and differentiate between healthy and unhealthy resource usage in application performance monitoring.
Innovation Solution
The technology generates performance data for computing resources such as CPU, memory, and network bandwidth, establishing a baseline for resource usage and reporting deviations through an interface, allowing users to assess resource health and performance by comparing actual usage to predicted values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If system management products analyze a large number of data streams to determine application health, then the ability to detect abnormal states improves, but the ability to identify the ultimate source or cause of problems deteriorates due to lack of semantic understanding
Solution Approach 1:
The patent introduces a semantic understanding layer as an intermediary between raw data streams and health determination. This layer processes and interprets the semantic meaning of data streams, enabling the system to not only detect abnormalities but also identify their sources and causes by understanding the contextual relationships between different data elements.
Solution Approach 2:
The patent segments the analysis process into distinct stages: data stream collection, semantic understanding processing, baseline comparison, and health determination. By segmenting the workflow and applying semantic analysis at appropriate stages, the system can handle large numbers of data streams while maintaining the ability to trace problems to their specific sources.
2Device complexity
If system management products lack semantic understanding of data, then device complexity is reduced, but measurement precision of application health deteriorates
Solution Approach 1:
The patent performs semantic understanding as a preliminary action before health determination. By pre-processing and interpreting the semantic meaning of data streams in advance, the system builds a foundation of understood relationships and contexts that enables precise health assessment without requiring complex real-time semantic processing during monitoring.
Solution Approach 2:
The system uses historical data and established baselines to automatically determine current health status without requiring complex external analysis. The semantic understanding components leverage previously processed information and self-generated baselines to make precise health assessments, reducing the need for complex external processing.
3Ease of operation
If system management products cannot differentiate between healthy and unhealthy resource usage, then ease of operation is improved, but loss of information about resource health deteriorates
Solution Approach 1:
The patent implements feedback mechanisms that continuously compare current resource usage against established baselines and provide clear health status indications. This feedback loop automatically differentiates between healthy and unhealthy states, presents the information in an operationally simple format, and maintains complete resource health information availability through structured reporting.
Data Source
AI summary
An application monitoring system determines the health of one or more resources used to process a transaction, business application, or other computer process. Performance data is generated in response to monitoring application execution and processed to determine and an actual and baseline value for resource usage data. Resource usage baseline data may be determined from previous resource usage data associated with a resource and particular transaction (a resource-transaction pair). The baseline values are compared to actual values to determine a deviation for the actual value. Deviation information for the time series data can be reported through an interface or some other manner.