Dynamic Network Load Restriction via Stochastic Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network load management systems struggle to accurately set throughput thresholds for backend content sources, leading to either overloading or underloading during request surges, which negatively impacts client experience due to poor performance.
Innovation Solution
The system determines a restriction value based on the supported number of requests by the backend content source, using a stochastic prediction model, to dynamically adjust the threshold for offloading requests to a waiting room, ensuring it aligns with the backend's capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the network offloads too many requests to a waiting room to account for potential surges, then the backend content source is protected from overloading, but the backend content source is underutilized
Solution Approach 1:
The patent implements dynamic threshold adjustment based on real-time backend capacity monitoring. The system continuously adapts the request offloading threshold according to actual backend performance metrics, transitioning from static to dynamic control. This resolves the contradiction by allowing the system to protect the backend during surges while maximizing utilization during normal conditions, as the threshold automatically adjusts to current system state rather than using fixed conservative values
Solution Approach 2:
The system employs feedback mechanisms by monitoring backend content source performance metrics and using this information to adjust offloading decisions. The network analyzes real-time data about backend capacity and request patterns, then dynamically modifies the threshold for offloading requests to the waiting room. This closed-loop control enables the system to balance protection and utilization by responding to actual backend conditions rather than predetermined assumptions
2Productivity
If the network offloads too few requests to the waiting room, then the backend content source is maximally utilized, but poor performance occurs during surges
Solution Approach 1:
The system dynamically adjusts the offloading threshold based on real-time backend capacity and request pattern analysis. During surge conditions, the threshold automatically increases to offload more requests to the waiting room, preventing backend overload and maintaining client experience. During normal conditions, the threshold decreases to maximize backend utilization. This dynamic adaptation resolves the contradiction between maximizing productivity and maintaining reliability during surges
Solution Approach 2:
The system takes preliminary action by proactively offloading requests to the waiting room before the backend becomes overloaded. By monitoring request patterns and backend capacity trends, the system anticipates potential surges and adjusts the offloading threshold in advance, preventing poor client experience before it occurs. This proactive approach maintains reliability while still maximizing utilization during normal operating conditions
3Ease of operation
If a fixed threshold is used for offloading requests, then the system is simple to operate, but it cannot adapt to actual backend capabilities during request surges
Solution Approach 1:
The system implements self-service by automatically monitoring backend content source capacity and autonomously adjusting the offloading threshold without requiring manual configuration or intervention. The network system itself gathers performance metrics, analyzes capacity trends, and dynamically modifies the threshold based on actual backend capabilities and request patterns. This self-adjusting mechanism maintains ease of operation while achieving high adaptability to changing conditions
Solution Approach 2:
The system performs preliminary analysis of backend capacity and request patterns to establish appropriate offloading thresholds before surges occur. By continuously monitoring backend performance metrics and predicting capacity needs, the system prepares adaptive threshold values in advance, enabling automatic adjustment when conditions change. This preliminary action maintains operational simplicity while ensuring the system can adapt to actual backend capabilities during surges
Data Source
AI summary
Systems and methods are provided that relate to imposing a restriction on requests for content by clients to a backend content source. One exemplary method includes generating, by a computing device associated with a backend content source, a restriction value for content of the backend content source, based on a number of supported requests for the content by the backend content source, and transmitting the restriction value to a content delivery network (CDN) interposed between one or more clients associated with the requests and the backend content source. The method then includes routing, by the CDN, a percentage of requests for the content of the backend content source to a waiting room based on the restriction value, thereby permitting the CDN to account for the number of supported requests in offloading requests for the content to the waiting room.