Adaptive Application Tracing Levels for Runtime Issue Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
IT systems with numerous applications face challenges in managing tracing decisions due to complex dependencies and varying performance factors, leading to runtime issues that affect overall system performance.
Innovation Solution
A method using a Tracing Policy Analyzer that detects application availability and performance metrics, trains a Bayesian optimization model to determine an optimal tracing policy, and adjusts tracing levels using a k-nearest neighbors algorithm to optimize tracing decisions based on collected monitoring data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If tracing is enabled for all applications in complex IT systems, then runtime issue identification capability is improved, but system performance overhead and complexity increase
Solution Approach 1:
The patent implements dynamic tracing level adjustment based on real-time system conditions. The tracing policy analyzer continuously monitors system state and adjusts tracing levels dynamically, transitioning from static all-or-nothing tracing to adaptive tracing that responds to actual runtime needs, thereby reducing unnecessary complexity while maintaining issue identification capability.
Solution Approach 2:
The patent applies different tracing levels to different applications or system components based on their specific needs and current state. Instead of uniform tracing across the entire system, the solution implements localized tracing strategies where only critical components or problematic areas are traced at high levels, reducing overall system complexity while maintaining effective issue identification.
2Measurement precision
If tracing level is increased to capture more runtime issues, then monitoring precision is improved, but system performance and resource consumption worsen
Solution Approach 1:
The patent implements partial tracing by selecting only the most relevant applications, components, or transaction types for detailed tracing based on current system conditions and issue patterns. This partial action approach captures sufficient information for effective issue identification while avoiding the performance penalty of comprehensive full-system tracing.
Solution Approach 2:
The patent dynamically changes tracing parameters such as tracing level, sampling rate, and enabled components based on real-time system conditions. When performance degradation is detected or specific issues are suspected, the system adjusts parameters to increase monitoring precision; when system is stable, parameters are relaxed to minimize performance impact.
3Reliability
If tracing is continuously enabled to monitor system performance, then real-time monitoring capability is improved, but energy consumption and resource usage increase
Solution Approach 1:
The patent implements periodic evaluation of system state and conditional activation of tracing rather than continuous tracing. The system periodically assesses whether tracing is needed based on performance metrics, error rates, or other indicators, and activates or deactivates tracing accordingly, reducing energy consumption while maintaining real-time monitoring capability when needed.
Data Source
AI summary
An embodiment includes detecting an application availability metric and a performance data metric of a system by a Tracing Policy Analyzer, responsive to detecting the application availability metric and the performance data metric, training a Bayesian optimization model of the Tracing Policy Analyzer based on the application availability metric and the performance data metric where the Bayesian optimization model outputs an optimal tracing policy. The embodiment includes detecting the optimal tracing policy by a Tracing Level Determination Module of an application of the system, responsive to detecting the optimal tracing policy, computing a computed tracing level by the Tracing Level Determination Module executing a k-nearest neighbors algorithm based on the optimal tracing policy. The embodiment also includes tracing of the application of the system by a Tracer based on the computed tracing level.