Auto-Adjusting Thresholds for System Metric Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer system monitoring technologies face issues with inaccurate or dynamically changing metric thresholds, leading to false notifications due to the need for manual configuration, which is time-consuming and error-prone.
Innovation Solution
Implementing an auto-adjusting threshold system that dynamically calculates new thresholds based on historical metric values, using a baseline value and maximum deviation to ensure accurate notifications over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual threshold configuration is used, then system administrators can control alert notifications, but false notifications occur due to inaccurate or changing thresholds
Solution Approach 1:
The system automatically adjusts thresholds based on historical metric data and system behavior patterns without requiring manual administrator intervention. The auto-adjust mechanism learns from past data to dynamically optimize threshold values, eliminating the need for continuous manual configuration while improving notification accuracy.
Solution Approach 2:
The system implements a feedback loop where notification data and metric history are continuously analyzed to refine and adjust thresholds. This closed-loop approach allows the system to learn from past notifications and system behavior, automatically optimizing thresholds to reduce false alerts while maintaining operational awareness.
2Productivity
If manual threshold configuration is used, then administrators can set initial thresholds, but the process is time-consuming and error-prone
Solution Approach 1:
The system performs preliminary analysis of historical metric data to pre-calculate optimal threshold values before they are needed for monitoring. By proactively establishing thresholds based on past system behavior, the system eliminates time-consuming manual configuration while ensuring accurate, data-driven threshold settings are ready for immediate use.
Solution Approach 2:
The auto-adjust mechanism autonomously handles the entire threshold configuration process by analyzing historical data and generating optimized thresholds without administrator involvement. This self-service approach dramatically reduces both the time required for setup and the potential for human error in threshold configuration.
3Adaptability or versatility
If fixed thresholds are used, then notification rules remain stable, but they do not adapt to changing system conditions
Solution Approach 1:
The system transitions from static, fixed thresholds to dynamic thresholds that automatically adjust based on changing system conditions and historical patterns. The auto-adjust mechanism continuously monitors metric behavior and modifies thresholds in real-time, enabling the system to adapt to evolving operational patterns while maintaining stability through data-driven consistency.
Solution Approach 2:
The system uses feedback from historical notification data and metric trends to continuously refine thresholds. This feedback loop ensures that thresholds remain consistent with actual system behavior while adapting to changes over time, balancing adaptability with operational stability through continuous learning from past performance.
Data Source
AI summary
Automatically setting a new threshold for a computer metric by detecting a first threshold for the computer metric, continuously taking measurements of the value of the computer metric at a predetermined rate, calculating the average of the measurements of the value of the computer metric over a predetermined time period, and calculating a second threshold for the computer metric based on the average of the values of the measurements. Calculating the second threshold may comprise establishing a base value for the threshold, establishing a maximum deviation of the threshold, and determining the average of the value of the measurements with respect to the base value and the maximum deviation.