A dispersed network of servers known as a content delivery network (CDN) is able to deliver web material to users effectively. To reduce latency, CDNs cache material on edge servers at points of presence (POP) that are close to end users.
By caching their material at carefully chosen physical nodes around the globe, Azure Content Delivery Network (CDN) provides developers with a global solution for quickly achieving great content to users. By utilising multiple network improvements with CDN POPs, Azure CDN may additionally accelerate dynamic material, which cannot be cached. Route optimization, for instance, to avoid Border Gateway Protocol (BGP).
Using Azure CDN to provide web site assets has several advantages, including:
- better user experience and speed for end users, particularly when utilising applications where loading material requires numerous round-trips.
- Large scalability to better handle sudden high loads, such the beginning of an event for a product launch.
- less traffic is routed to the source server thanks to the dispersion of user queries and the direct serving of content from edge servers.
How does it works?
- A user (Alice) uses a URL with a unique domain name, such as.azureedge.net, to request a file (also known as an asset). This name may be a custom domain or an endpoint hostname. The DNS directs the demand to the POP that offers the best performance, which is typically the POP that really is nearest to the user geographically.
- If none of the POP’s edge servers have the item in their caches, the POP asks the origin server for it. The origin server may be any publicly available web server, an Azure Web App, an Azure Cloud Service, an Azure Storage account, etc.
- An edge server in the POP receives the file back from the origin server.
- The file is cached by an edge server inside the POP, which then sends it back to the requester (Alice). Until the time-to-live (TTL) indicated by its Header fields expires, the file is still cached just on edge server in the POP. The standard TTL is 7 days if the originating server didn’t indicate a different one.
- The same file can then be requested by further users using the same URL as Alice used, as well as by directing them to the same POP.
- The POP edge server delivers the file immediately from the cache if the file’s TTL hasn’t run out. A quicker, more reactive user experience is the result of this approach.
