CDN Network Mapping via Netd Agent Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Content delivery networks (CDNs) face challenges in determining device connectivity and network mapping, especially in dynamic networks like the Internet, where constant topology changes and firewall restrictions limit the ability to measure network connectivity to all clients, leading to incomplete network maps.
Innovation Solution
The implementation of a mechanism that uses netd agent mechanisms to continuously measure network connectivity from servers to known locations, combined with a map builder that updates network maps regularly, and a Least Latent Algorithm that determines optimal name servers by tracking response times, along with CNAME responses to estimate latency behind firewalls, ensures accurate network mapping and server selection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If network connectivity measurement is performed continuously to update network maps in real-time, then network mapping accuracy is improved, but system complexity and resource consumption increase
Solution Approach 1:
The system divides network mapping tasks into segments by assigning different netd agents to measure connectivity to different target groups. Each agent maintains a focused measurement set rather than all-to-all measurements, reducing overall system complexity while maintaining comprehensive coverage through distributed segmentation of the measurement function across multiple agents and target groups.
2Loss of information
If connectivity measurements are performed to all possible client locations, then network map completeness is improved, but measurement capability is limited by firewall restrictions
Solution Approach 1:
Instead of having netd agents attempt to initiate measurements to all possible client locations (which fails due to firewalls), the system inverts the approach by having agents measure connectivity to known target locations and inferring client network positions from these indirect measurements. This reverse measurement strategy bypasses firewall restrictions while still gathering network mapping information.
Solution Approach 2:
The system introduces known target locations as intermediary measurement points between netd agents and clients behind firewalls. By measuring connectivity to these intermediaries and analyzing response characteristics, the system can infer network map information about clients without requiring direct measurement capability through the firewall barrier.
3Loss of time
If netd agents continuously measure network connectivity to maintain real-time network maps, then network connectivity information freshness is improved, but resource consumption increases
Solution Approach 1:
The system implements periodic measurement cycles where netd agents continuously measure connectivity to their assigned target groups at regular intervals. This periodic action maintains fresh network connectivity information for real-time routing decisions while controlling resource consumption through structured, time-bound measurement cycles rather than constant exhaustive probing.
Data Source
AI summary
A computer-implemented method in a content delivery network (CDN) having multiple delivery servers. The CDN delivers content on behalf of at least one content provider. Distinct delivery servers are logically grouped into delivery server groups. One or more CDN name servers are associated with some of the delivery server groups. Network map data are determined using network data determined by the CDN name servers associated with at least some of the deliver server groups. The network data with respect to a CDN name server relative to a resolver is based on an estimated popularity of that CDN name server for that resolver. Responsive to a client request, including a hostname associated with a content provider, at least one CDN name server determines, using network map data, at least one delivery server to process the client request.