Content Delivery Networks

The speed of light is the maximum speed at which all energy, matter, and information can travel

Transcript

1. Content Delivery Networks Sara Rodríguez Cubillas
2. –Albert Einstein “The speed of light is the maximum speed at which all energy, matter, and information can travel”
3. A CDN (content delivery network) is a network of servers located in different parts of a country (or the globe) that stores files to be used by your website visitors.
4. Speed of light and propagation Latency source: “High Performance Browser Networking”, Ilya Grigorik The actual RTT between New York and Sydney, works out to be in the 200-300 milliseconds range.
5. All thing considered that seems pretty fast, right?
6. Waiting for the response • 100–200 milliseconds —> perception of “lag” ! • 300 millisecond —> “sluggish” ! • 1 second —> mental context switch, anything from a daydream to thinking about the next urgent task
7. Waiting for the response • 40% of consumers will wait no more than 3 seconds* ! • CDNs improved website performance by caching static site content ! • The beauty of the virtual network approach: - it works over the existing Internet as is - requires no client software - can easily be adapted to future requirements as the Internet evolves ! * A 2009 Forrester Consulting survey
8. Context: evolution of Internet Applications
9. Context • As the web took off in the 90s, traffic volumes grew and grew. This: ! 1. Concentrated load on popular servers ! 2. Led to congested networks and need to provision more bandwidth ! 3. Gave a poor user experience ! • Idea: – Place popular content near clients – Helps with all three issues above
10. Why are CDNs becoming so popular? • Google has started using page speed as a ranking factor ! • Faster experience for users means happier users who buy more things and click more ads ! • The explosion of tablet and mobile users who depend on speed more so than desktop users who have stabler internet connections ! Some of the more known and reputable CDN companies: • Cloudflare (has free option) • Fastly • Amazon CloudFront • MaxCDN • Akamai • Cachefly
11. Before CDNs • Sending content from the source to 4 users takes 4 x 3 = 12 “network hops” in the example
12. After CDNs • Sending content via replicas takes only 4 + 2 = 6 “network hops” ! • Benefits assuming popular content: – Reduces server load, network load – Improves user experience (PLT)
13. Overall Architecture
14. DNS resolution of site gives different answers to clients – Tell each client the site is the nearest replica (map client IP)
15. Anatomy of a Delivery Network *The Akamai Network: A Platform for High-Performance Internet Applications
16. Ok, but… What´s happens with dynamic content?
17. What is Dynamic content? Stuff that’s not static! ! With web traffic, generally the base HTML -Big deal because it’s blocking -And sometimes a large object —>longer download ! Could be other things too -Ajax calls -API calls
18. Why Is Caching Dynamic Content Difficult? Serving stale pages - Lack of invalidation framework ! Losing visibility: - Analytic/stats - Logging tracking
20. *http://www.webpagetest.org/
24. CDNs and Dynamic content Generally, handling dynamic content has been a matter of transport ! • Optimize from-origin delivery ! • “DSA” (dynamic site acceleration) ! • Middle-mile optimizations ! • Dynamic content delivery/acceleration ! • TCP tweaks
27. Dynamic content We sometimes do micro caching of HTML - Short TTL for HTML content via Cache-control - not really reliable ex: news stories ! ESI (Edge Side Includes) - Partial caching - hard and onerous ! Fastly solution(not everything is dynamic)
28. Service-based Thinking • The CDN is an extension of the app • No longer a black box • Real time integration with the app • Infrastructure as a code - Your content => you need control
29. Control • Programmability - Configuration API - Invalidation API - Instantaneous and real time
30. Caching Control • Granular caching - Dynamic caching - and uncaching! - Flexible cache keys - etc • Ex: Geo-based caching
31. Control at the Edge • Moving app logic to the edge • VCL • Varnish Configuration Language • Script-like configuration for functionality at the edge
32. Visibility • Real time analytics • Network stats • HTTP stats (status codes, etc) • Caching stats (hits, misses,etc) • Stats API • Logging • Real time logs • Streaming to various log destination
33. Summary • The speed of light is fast • The perception of the users is important, google speed ranking too • Distribute the content around the globe => reduce the propagation time • Dynamic content can be handled by CDNs too
34. References • Ilya Grigorik .“High Performance Browser Networking”, O’Really. ! • Rajkumar Buyya · Mukaddim Pathan · Athena Vakali (Eds.)Content Delivery Networks. Springer. • Hooman Beheshti talk: Caching the uncachable, at Velocityconf Europe 2014 • Erik Nygren, Ramesh K. Sitaraman, and Jennifer Sun. "The Akamai Network: A Platform for High-Performance Internet Applications, ACM SIGOPS Operating Systems Review, vol. 44, no. 3, July 2010.”. • Computer Networks Course from the University of Washington by David Wetherall, Arvind Krishnamurthy, John Zahorjan. - https://class.coursera.org/comnetworks-002 • A Beginner's Guide to HTTP Cache Headers- http://www.mobify.com/blog/beginners-guide-to-http-cache-headers/ • Content Delivery Networks - http://www.feedthebot.com/pagespeed/content-delivery-networks.html • Intro to caching and CDNs -https://docs.fastly.com/guides/how-fastly-services-work/intro-to-caching-and-cdns • Leveraging Your CDN to Cache Dynamic Content - http://www.fastly.com/blog/caching-dynamic-content/ • Gang Peng Department of Computer Science,State University of New York at Stony Brook, Stony Brook, NY 11794-4400. CDN: Content Distribution Network