Self-Discovering Throttling via Dual Queue Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional computing resource allocation methods are inefficient, leading to system overload during high demand scenarios like multimedia messaging, as they require excessive overhead to monitor resources and are affected by external factors, resulting in slowed performance.
Innovation Solution
A self-discovering throttling system that uses two queues to track task completion rates and a comparator to adjust the number of connections dynamically, allowing the system to automatically optimize task execution without external intervention, accounting for both internal and external factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional resource allocation methods monitor each computing resource (CPU, bandwidth, memory, databases) to determine system overload, then the system can identify resource bottlenecks, but the monitoring overhead slows down the overwhelmed system further
Solution Approach 1:
The patent extracts the monitoring function from individual resource components (CPU, memory, bandwidth, databases) and consolidates it into a single queue management system. Instead of monitoring each resource separately, the system monitors only the queue depth and task completion rates, eliminating the overhead of multiple resource monitoring operations while maintaining effective overload detection.
Solution Approach 2:
The patent introduces queues as an intermediary layer between incoming tasks and computing resources. The queue management system acts as a mediator that absorbs monitoring overhead and decouples task submission from resource execution, allowing the system to track workload through queue depth without directly monitoring each underlying resource component.
2Productivity
If the system increases the number of connections to handle high demand, then task completion rate improves, but the system becomes overwhelmed by excessive load
Solution Approach 1:
The patent implements a feedback mechanism where the system continuously compares the current task completion rate with the average completion rate. When the current rate exceeds the average by a threshold, the system automatically reduces the number of connections. This closed-loop feedback control enables dynamic adaptation to load conditions, maintaining high productivity while preventing system overload and instability.
3Ease of manufacture
If the system manually configures resource allocation parameters, then initial performance can be optimized, but the system cannot adapt to changing external factors such as available bandwidth and external system capability
Solution Approach 1:
The patent enables the system to self-adjust resource allocation parameters automatically based on observed task completion rates. The queue management system continuously monitors performance metrics and dynamically modifies connection numbers without requiring manual reconfiguration or external input. This self-service capability allows the system to adapt to changing external conditions such as network bandwidth availability and external system capacity while maintaining ease of initial setup.
Data Source
AI summary
A method and apparatus for self-discovering throttling is described. The system includes a first queue to track a current rate of task completion and a second queue to track an average rate of task completion over time. The system further includes a comparator to compare an average of the first queue and an average of the second queue. The system includes a throttle to adjust the number of connections based on the results from the comparator.