Edge Server Selection via Content Allocation Register
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video-on-demand systems face inefficiencies due to the lack of awareness among central servers about the content cached at edge servers, leading to random server selection and increased network latency and resource taxation.
Innovation Solution
Implementing a resource manager with a content allocation register that maintains a popular asset list and edge server list, ranking content by demand and timestamp, to preferentially route requests to edge servers likely to have the requested content cached, thereby optimizing cache efficiency and content delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If random server selection is used in video-on-demand systems, then any server can be selected to deliver content, but network latency increases and cache efficiency decreases
Solution Approach 1:
The system implements feedback by having edge servers report their cached content to the central server, which maintains an updated content allocation register. This feedback loop enables the central server to make informed routing decisions based on real-time cache status, reducing network latency by directing requests to servers that already have the content cached.
Solution Approach 2:
The system performs preliminary action by pre-populating edge server caches with video content before demand occurs. The content allocation register is maintained in advance with information about which content is cached at which edge servers, enabling quick routing decisions without real-time discovery overhead.
2Productivity
If central servers are unaware of edge server cache content, then system simplicity is maintained, but content delivery efficiency deteriorates
Solution Approach 1:
The content allocation register acts as an intermediary data structure that bridges the central server and edge servers. It stores and manages information about content cached at edge servers, allowing the central server to make informed routing decisions without direct real-time communication with each edge server, thus maintaining system simplicity while improving delivery efficiency.
3Loss of energy
If edge servers cache content locally, then network resource taxation is reduced, but cache utilization efficiency decreases without proper routing
Solution Approach 1:
Edge servers provide feedback to the central server about their cache status and content availability. The central server uses this feedback to optimize routing decisions, ensuring that requests are directed to edge servers with cached content, thereby maximizing cache utilization and minimizing network resource taxation.
Solution Approach 2:
The system replaces complex real-time negotiation and discovery mechanisms with a simpler information registration system. Instead of servers continuously advertising their cache status or clients performing complex discovery protocols, the system uses a centralized content allocation register that passively records cache information, simplifying the routing management complexity.
Data Source
AI summary
A video-on-demand system includes a resource manager (501) in communication with a central storage server (502), a back office system (503), and a plurality of edge servers (504). The resource manager can receive (523) a video-on-demand session request 524) requesting delivery of content (529), read (525) a content allocation register (520) that includes a popular asset list (301) with a plurality of content assets ranked (314) by highest demand (314) within a predetermined past usage window (316) and an edge server list (302) indicating which edge servers have delivered each content asset of the popular asset list. The resource manger can then Select (526) a selected edge server (527) from the content allocation register as a function of the popular asset list and the edge server list and cause (528) the selected edge server to deliver (529) the content in response to the video-on-demand session request.