Compute Resource Event Mapping for Performance Degradation Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to accurately track and mitigate performance degradation in computing devices due to reliance on single event thresholds, leading to undetected issues and potential system instability, such as increased CPU utilization and crashes.
Innovation Solution
A hybrid tracking approach that utilizes multiple event types associated with different compute resource characteristics to determine severity levels, enabling targeted actions like application termination, decompilation, and resource management to mitigate performance degradation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If single event threshold tracking is used, then system complexity is reduced, but measurement precision and reliability of performance degradation detection deteriorate
Solution Approach 1:
The patent segments the performance degradation detection into multiple independent event types (memory pressure events, CPU utilization events, disk I/O events, network events) each tracked separately with its own threshold. This allows precise detection of different degradation scenarios without requiring a single complex tracking mechanism, resolving the contradiction between simplicity and precision.
Solution Approach 2:
The patent transitions from single-dimension tracking (one threshold) to multi-dimensional tracking by introducing multiple event types and thresholds across different resource dimensions. Each event type represents a separate dimension of performance monitoring, enabling comprehensive detection while maintaining individual event handler simplicity.
2Measurement precision
If multiple event types and thresholds are implemented, then measurement precision of performance degradation improves, but device complexity increases
Solution Approach 1:
The patent divides the complex performance monitoring into separate, manageable event handlers for each event type. Each handler is simple and focused on a specific resource, avoiding the complexity of a unified complex system while achieving comprehensive detection through multiple simple components.
Solution Approach 2:
The patent implements a feedback mechanism where detected events trigger appropriate mitigation actions. The system continuously monitors resource characteristics, compares them against thresholds, and adjusts system behavior accordingly, creating a closed-loop system that maintains precision without requiring overly complex tracking infrastructure.
3Reliability
If performance degradation is not accurately tracked, then system stability is maintained (no false interventions), but harmful factors increase (undetected crashes and lag)
Solution Approach 1:
The patent takes preliminary action by detecting performance degradation events before they lead to system crashes or severe lag. By monitoring resource characteristics continuously and triggering mitigation actions at appropriate thresholds, the system prevents harmful factors from manifesting rather than reacting after damage occurs.
Solution Approach 2:
The patent converts potentially harmful resource exhaustion conditions into beneficial mitigation opportunities. When memory pressure or CPU utilization thresholds are reached, the system automatically terminates processes or adjusts resource allocation, transforming what would be crash conditions into controlled resource management actions that protect system stability.
Data Source
AI summary
Disclosed herein are system, apparatus, article of manufacture, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for mitigating performance degradation of a computing device. In an embodiment, a notification indicative of a characteristic of a compute resource of the device is received. The notification is of a first type of a plurality of types of notifications, and each type of notification of the types of notifications is indicative of a different characteristic of the compute resource. A level of performance degradation, from levels of performance degradation, of the computing device is determined based on a mapping that maps each type of notification of the types of notifications to a corresponding level of performance degradation of the different levels of performance degradation. An action from a plurality of different actions configured to mitigate the performance degradation is performed based on the determined level of performance degradation.