HTTP cache @ PUG Rome 03-29-2011
2 likesâ˘1,847 views
HTTP caching involves storing copies of resources near clients to serve future requests faster. Caching can happen locally on a client or through shared proxies. Effective caching requires expiration dates, validation of cached responses, and invalidation of cached responses when content changes. Caching allows servers to scale to many users by offloading work to clients and proxies. The HTTP protocol and technologies like ESI were designed to support caching while handling dynamic content.
HTTP cache @ PUG Rome 03-29-2011
- 1. HTTP cache Alessandro Nadalin March, 29th 2011
- 2. The Web is inconsistent
- 3. get over it
- 4. Why?
- 5. Because it scales to billions of users
- 6. Because it scales to billions of users WWW data traffic won't be possible without being inconsistent
- 7. How does it scales?
- 8. Cache
- 10. Storing data so that future requests for that data can be served faster ( general purpose )
- 11. Process of storing copies of your most accessed resources near to a/some client/s ( the WWW )
- 12. Caching is not saving a webpage, or a fragment of it, on the disk, to serve it to the next client
- 13. Caching is not saving a webpage, or a fragment of it, on the disk, to serve it to the next client This is a caching mechanism
- 16. Can be distributed with a NoSQL database
- 17. Cache implementation can be really funny
- 18. Don't reinvent the wheel
- 19. Goals
- 20. Goals evolve
- 21. Goals evolve loose coupling
- 22. Goals Work less evolve loose coupling
- 23. Evolve Because you want your platform to extensible
- 24. Evolve Because you want your platform to extensible Loose coupling Because you want it to be easy to integrate with, evolve, plug and mantain
- 25. Evolve Because you want your platform to extensible Loose coupling Because you want it to be easy to integrate with, evolve, plug and mantain Work less Because every LoC is bug-prone and our man-day is a hard to scale cost
- 26. enters our Hero #1
- 29. enters our Hero #2
- 32. Goals Work less evolve loose coupling
- 33. Goals Work less evolve loose coupling
- 34. everything is done for us! :) but....
- 35. tmp files, cache tables, procedural crap... mmmmh.... gotta be something better
- 36. Frameworks
- 38. Cache is used for compiling routes, autoloading, ...
- 39. Cache is used for compiling routes, autoloading, ... ...but also for storing the view
- 40. Goals Work less evolve loose coupling
- 41. Goals Work less evolve loose coupling
- 42. at least because we use a framework
- 43. but, probably
- 44. PHP sucks
- 45. Let's call a hero, from the East
- 46. on Rails
- 48. Goals Work less evolve loose coupling
- 49. gosh :-(
- 50. hey, what about the latest products?
- 51. both Symfony2 and Ruby's Rack embrace
- 52. HTTP
- 53. (here comes the nice part)
- 54. Local Shared/proxy Shared/reverse proxy
- 55. Local Shared/proxy Shared/reverse proxy
- 57. Local Shared/proxy Shared/reverse proxy
- 59. Local Shared/proxy Shared/reverse proxy
- 61. on the server side
- 62. Proxy Reverse proxy http://www.codeproject.com/KB/aspnet/ExploringCaching.aspx
- 63. There is something cool about caching softwares.
- 64. Free as free beer
- 65. ...but the server should take advantage of them ( the boring part were you need to write code )
- 66. and here are a few ways to do so, using
- 67. and here are a few ways to do so, using expiration
- 68. and here are a few ways to do so, using expiration validation
- 69. and here are a few ways to do so, using expiration validation invalidation
- 70. and here are a few ways to do so, using expiration validation we will se why invalidation
- 71. Expiration
- 72. GET / HTTP/1.1 Host: www.example.com Expires: 0
- 73. GET / HTTP/1.1 Host: www.example.com Expires: 0
- 74. GET / HTTP/1.1 Host: www.example.com Expires: Tue, 15 Nov 1994 01:00 GMT
- 75. GET / HTTP/1.1 Host: www.example.com Cache-Control: max-age=60, public
- 76. GET / HTTP/1.1 Host: www.example.com Cache-Control: max-age=60, public
- 77. GET / HTTP/1.1 Host: www.example.com Cache-Control: max-age=60, public Cacheable for 60 seconds
- 78. GET / HTTP/1.1 Host: www.example.com Cache-Control: max-age=60, public Cacheable by both local and shared caches
- 79. GET / HTTP/1.1 Host: www.example.com Cache-Control: stale-if-error=600, stale-while-revalidate=600
- 80. GET / HTTP/1.1 Host: www.example.com Cache-Control: stale-if-error=600, stale-while-revalidate=600 fault-tolerant
- 81. GET / HTTP/1.1 Host: www.example.com Cache-Control: stale-if-error=600, stale-while-revalidate=600 available during downtime
- 82. GET / HTTP/1.1 Host: www.example.com Cache-Control: stale-if-error=600, stale-while-revalidate=600 available during revalidation
- 83. Validation
- 84. GET / HTTP/1.1 Host: www.example.com Etag: 1234
- 85. GET / HTTP/1.1 Host: www.example.com Etag: 1234 an identifier for your response
- 86. Conditional requests GET / HTTP/1.1 Host: www.example.com If-None-Match: 1234 the browsers asks you if it has been modified
- 87. Relax
- 88. Calculating an Etag is cheaper than generating a full MVC response
- 89. HTTP/1.1 304 Not Modified
- 90. GET / HTTP/1.1 Host: www.example.com Last-Modified: Tue, 15 Jan 2011 12:00:00 GMT
- 91. GET / HTTP/1.1 Host: www.example.com Last-Modified: Tue, 15 Jan 2011 12:00:00 GMT tell the client about the latest change
- 92. Conditional requests GET / HTTP/1.1 Host: www.example.com If-Modified-Since: Tue, 15 Jan 2011 12:00:00 GMT the client asks you if it has been modified since the last time
- 93. Relax
- 94. Calculating a date is cheaper than retrieving an entire object
- 95. HTTP/1.1 304 Not Modified
- 96. Invalidation
- 97. The web is not meant for invalidating data. Server should not be able to keep clients' state, otherwise they wont scale well. That's why long-polling and endless connections haven't had big success dealing with caching.
- 98. but hey, you say
- 99. HTTP's cache fails when dealing with really dynamic pages, because consumers will always have to hit the origin server, although a part of the page would be cacheable ( header and footer, for example )
- 100. Nope Nope
- 101. ESI was built for that http://www.w3.org/TR/esi-lang
- 102. and hey, Varnish is a reverse proxy implementing what you need of the ESI specification take 2, pay for 1
- 103. So what does HTTP cache is meant to solve?
- 104. Less work
- 105. http://www.flickr.com/photos/snakphotography/5004775320/sizes/o/in/photostream/ because the hard work is delegated to the browser/proxy
- 106. evolve
- 107. because cache is abstracted from the application
- 108. loose coupling
- 109. because caching is bound to the protocol, HTTP, not to your implementation ( Sf, RoR, Django )
- 110. The presentations lacks on technical stuff So here are a few resource for who wants documentation: http://tomayko.com/writings/things-caches-do http://www.slideshare.net/fabpot/caching-on-the-edge http://www.odino.org/301/rest-better-http-cache http://www.w3.org/Protocols/rfc2616/rfc2616-sec13.html
- 111. Deep version of the same talk, about Sf2 cache internals and ESI
- 112. May, 12/14 2011 in Verona Early bid ends on April, 6th
- 113. Thanks! Alessandro Nadalin @_odino_ odino.org
Editor's Notes
- #19: Ne consegue che non dobbiamo reinventare sempre la ruota.