The scale of CDN infrastructure makes a fail-safe system statistically impossible. However, this same scale helps provide record resilience and high availability, allowing 99.9% and 99.999% service level agreements (SLAs).
Typically, commercial CDNs take a “no single mexico phone number library point of failure” approach by carefully staging maintenance cycles and integrating additional hardware and software redundancy. Many also manage internal failover and disaster recovery systems that automatically reroute traffic around downed servers. For additional redundancy, CDN providers also maintain multiple carriers and rely on dedicated out-of-band management channels that allow them to interact with servers in the event of a disaster.
Answerability
With a global network, CDNs continually try to improve responsiveness by measuring the time it takes for configuration changes to take effect across the network.
Remember, even small configuration changes, such as clearing a specific image from the cache or adding an address to a blacklisted IP list, must be pushed across all POPs. The larger and more geographically spread out the network, the longer it takes to do this.
To provide good quality of service to customers, a CDN should be designed with rapid configuration deployment in mind. This is often achieved through a combination of consolidation.
CDN Topology
Distributed CDN
Distributed CDNs use a large number of medium and low capacity POPs that densely populate specific geographic areas. This topology focuses on optimal physical proximity.
