Terminal Resource Request for Edge Node I/O Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Edge nodes in P2SP systems are prone to bottlenecks in I/O performance during video stream downloads, leading to inefficient data retrieval and potential service disruptions.
Innovation Solution
A method where a terminal divides video resources into uniform slices and requests them from both CDN and edge nodes, prioritizing CDN nodes for initial playback and dynamically adjusting edge node requests based on transmission rates to optimize concurrent data retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the client requests data segments from multiple edge nodes simultaneously to increase download rate, then the download speed is improved, but the edge nodes experience I/O performance bottlenecks
Solution Approach 1:
The patent segments the video stream into multiple slices and further divides each slice into multiple data segments. The client requests different data segments from different edge nodes for the same slice, distributing the I/O load across multiple nodes while maintaining high download speed through parallel retrieval operations.
Solution Approach 2:
The patent introduces a hierarchical structure with CDN nodes at the top level and edge nodes at the lower level. By organizing the request architecture into multiple dimensions (CDN level and edge node level), the system can simultaneously achieve high download speeds through parallel requests while avoiding I/O bottlenecks by distributing load across different hierarchical levels.
2Productivity
If the client uses P2P download task with multiple edge nodes, then the download rate is increased, but the edge nodes are prone to bottlenecks in I/O performance
Solution Approach 1:
The CDN node serves as an intermediary between the client and multiple edge nodes. The CDN node receives requests from the client and distributes them to different edge nodes, managing the I/O load distribution and preventing any single edge node from becoming a bottleneck while maintaining high aggregate download rate.
Solution Approach 2:
The patent implements dynamic request management where the client adjusts its downloading strategy based on real-time conditions. The system dynamically selects which edge nodes to request from and which data segments to retrieve, optimizing the balance between download rate and I/O performance under varying network conditions.
3Loss of time
If the client requests data from CDN node in starting stage, then the first screen playing delay is avoided, but the CDN node bears the initial load
Solution Approach 1:
The patent implements preliminary action by having the client first request a certain data amount of small slices from the CDN node in the starting stage before initiating P2P download tasks. This ensures sufficient buffer is loaded for immediate playback, avoiding delay while the CDN node handles only the initial load rather than continuous streaming.
Solution Approach 2:
The system dynamically transitions from CDN-based downloading in the initial stage to P2P-based downloading after the buffer reaches a certain threshold. This dynamic switching strategy ensures fast initial playback while gradually transferring the load to edge nodes, reducing CDN node burden over time.
Data Source
Figure 1~2
Figure 3
AI summary
The present disclosure provides a method for requesting resources and a terminal. The terminal sends a first resource acquisition request for a target resource to each of first edge nodes, determines each of second edge nodes based on each first amount of the target resource returned by each of the first edge nodes, and sends a second resource acquisition request for the target resource to each of the second edge nodes.