Hierarchical Performance Manager for Component-Based Applications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing system and performance management tools are inefficient for highly layered and interconnected component-based applications, as they fail to understand application architecture, business functions, and resource demands due to low-level technical metrics and lack of contextual information, making it difficult to identify root causes of performance issues.
Innovation Solution
A performance manager that accesses and interprets application metadata and execution environment metadata, using a hierarchical entity model to represent complex relationships between application abstractions, components, and resources, conducting inter-component dependency and control flow analysis without access to source code, and collecting metrics from standardized and vendor-specific sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional performance management tools use low-level technical metrics (CPU usage, memory usage, I/O activity), then resource monitoring is achieved, but the ability to understand application architecture and business functions is lost
Solution Approach 1:
The patent implements a hierarchical model where business function metrics are nested within application metrics, which are nested within resource metrics. This multi-level nesting structure allows the system to maintain both high-level business context and low-level resource details simultaneously, resolving the contradiction between measurement precision and information loss.
Solution Approach 2:
The patent adds a new dimension to performance management by introducing application-level and business function-level metrics above the traditional resource-level metrics. This dimensional expansion creates a multi-dimensional view that preserves both resource monitoring precision and application architecture understanding.
2Ease of operation
If code instrumentation is used to provide visibility into component performance, then detailed application architecture knowledge is required, but this approach is only feasible for severe problems and increases complexity
Solution Approach 1:
The patent introduces an intermediary layer (application-level metrics and business function metrics) between resource metrics and business functions. This intermediary provides performance visibility without requiring direct code instrumentation or detailed architecture knowledge, as the intermediary automatically aggregates and presents relevant information.
Solution Approach 2:
The patent creates a virtual model of the application architecture through metadata and hierarchical relationships, rather than requiring physical code instrumentation. This virtual copy allows performance monitoring without the complexity of actual code modification and detailed architecture knowledge.
3Productivity
If application servers share resources and service components across many applications, then resource utilization efficiency is improved, but the one-to-one mapping between hardware and software components is lost
Solution Approach 1:
The patent segments the performance monitoring system into distinct hierarchical levels (resource metrics, application metrics, business function metrics). This segmentation allows each level to be managed independently, resolving the mapping complexity while preserving resource utilization efficiency through the structured hierarchy.
4Quantity of substance
If vast amounts of low-level technical metrics are collected, then detailed performance data is available, but the ability to identify root causes and understand business context is reduced
Solution Approach 1:
The patent extracts and isolates the most relevant metrics at each hierarchical level, separating business function metrics from the vast array of low-level technical metrics. This extraction process maintains data volume while improving root cause identification by presenting only the most relevant information at each level.
Data Source
AI summary
Invention manages underlying application objects that support implemented business functions. A performance manager accesses and interprets application metadata and execution environment metadata, using a hierarchical entity to model complex relationships between application abstractions, components and resources. The performance manager encodes common behavioral and architecture information in models resulting from analysis of component architecture standards, vendor application servers, development tools and administration tools. The performance manager conducts inter-component dependency and control flow analysis, for example by generating control flow graphs from application metadata and performing byte code analysis without access to application source code. The performance manager collects metrics from standardized and vendor-specific sources, instruments interpreters (such as the Java Virtual Machine), and collects and utilizes resource metrics, code performance metrics, configuration parameters, and transaction, failure and lifecycle events.