Anomaly Detection Check Scheduling in Computing Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The cumulative effect of running checks to detect anomalies in computing systems degrades system performance by consuming resources, while not scheduling checks leads to system integrity degradation due to undetected glitches.
Innovation Solution
Anomaly detection module schedules checks based on an optimal frequency calculated from the average time between detected anomalies and runtime, minimizing costs associated with executing and not executing checks, using a scheduler module to balance resource usage and anomaly detection efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If checks are run frequently to detect anomalies, then system integrity is improved, but system performance degrades due to resource consumption
Solution Approach 1:
The patent implements dynamic check scheduling where the frequency of executing checks is adjusted based on system state. The scheduler monitors anomaly detection effectiveness and resource consumption, then dynamically modifies check intervals to optimize between detecting glitches and preserving system performance for other applications.
Solution Approach 2:
The patent changes the temporal parameter of check execution by calculating optimal intervals based on the average time between anomalies. Instead of fixed frequent checks, the system adjusts check frequency as a variable parameter, extending intervals when system performance is critical and reducing them when anomalies are detected, thereby resolving the contradiction between reliability and productivity.
2Productivity
If checks are not scheduled to preserve system performance, then productivity is improved, but system integrity degrades due to undetected glitches
Solution Approach 1:
The scheduler implements feedback mechanisms by monitoring the outcomes of check executions and the timing of detected anomalies. This feedback information is used to adjust future check scheduling decisions, ensuring that system performance is maintained while still detecting glitches effectively. The feedback loop allows the system to learn from past performance and optimize future check intervals.
3Reliability
If all checks are executed continuously, then anomaly detection coverage is improved, but resource consumption increases to 100% of system capacity
Solution Approach 1:
Instead of executing all checks continuously (excessive action), the patent applies partial action by selectively scheduling checks based on system state and anomaly probability. The scheduler determines the optimal subset of checks to execute at any given time, avoiding unnecessary resource consumption while maintaining adequate anomaly detection coverage through strategic check selection and timing.
Data Source
AI summary
In an example embodiment, a method is provided for scheduling a check to detect anomalies in a computing system. An average time between the anomalies that are detectable by the check is identified and additionally, a runtime of the check is identified. A frequency of the check is then calculated based on the average time between the anomalies and the runtime of the check, and execution of the check may be scheduled based on the calculated frequency.