Intelligent Logging System Dynamic Threshold Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing logging systems rely on static logging levels that are predetermined, which may not provide useful information for root cause analysis when a system becomes unhealthy or stops operating, as the issues have already occurred by the time additional logging is initiated.
Innovation Solution
An intelligent logging system that dynamically adjusts logging levels based on real-time monitoring of critical processes, automatically changing log levels without operator intervention, by monitoring process parameters and thresholds stored in a non-transitory machine-readable storage medium.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If static logging levels are used, then system operation is simple and predictable, but useful information for root cause analysis is lost when system becomes unhealthy
Solution Approach 1:
The logging system transitions from static to dynamic by automatically adjusting logging levels based on real-time process health monitoring. The system monitors critical processes and dynamically changes logging verbosity without manual intervention, ensuring detailed logs are captured only when needed (when processes become unhealthy) rather than continuously or statically.
Solution Approach 2:
The system implements feedback mechanisms by monitoring process parameters and using this information to automatically adjust logging levels. When process health indicators indicate deterioration, the system responds by increasing logging detail, creating a closed-loop system that adapts logging behavior based on actual system state.
2Loss of information
If logging level is increased to capture more details, then root cause analysis capability is improved, but system performance and resource usage deteriorate
Solution Approach 1:
The system dynamically adjusts logging intensity based on actual need. During normal operation, logging remains at minimal levels to preserve system performance. When process health monitoring detects issues, the system automatically increases logging detail only for affected processes, capturing necessary diagnostic information without continuously burdening system resources.
Solution Approach 2:
Instead of uniformly increasing logging across the entire system, the patent applies detailed logging locally only to specific critical processes that show signs of health deterioration. This targeted approach ensures diagnostic information is captured where needed while minimizing the performance impact on the broader system.
3Loss of information
If operator intervention is required to change logging levels, then logging control is precise, but response time to capture error information is delayed
Solution Approach 1:
The logging system becomes self-managing by automatically monitoring process health and adjusting its own logging levels without operator intervention. When critical processes show signs of failure, the system autonomously increases logging detail, ensuring error information is captured immediately at the source rather than waiting for manual detection and configuration changes.
Solution Approach 2:
The system performs preliminary monitoring of process health indicators continuously and is prepared to immediately increase logging detail when thresholds are breached. This preliminary preparation ensures that when errors occur, the system can instantly capture relevant information without waiting for operator awareness or manual intervention.
Data Source
AI summary
Examples relate to intelligent logging in a system. One example enables monitoring a set of critical processes of the system, responsive to a first process parameter of a first critical process exceeding a corresponding first parameter threshold, changing a first process log level associated with the critical process from a first log level to a second log level; and logging information related to the first critical process by: obtaining a second set of information associated with the second log level, wherein the second set of information is different from a first set of information associated with the first log level.