Varnish intro
Download as PPTX, PDF1 like1,095 views
This document discusses using Varnish as a caching solution for Drupal websites. It begins by introducing Varnish and explaining that it is a reverse proxy cache server that can improve Drupal performance. It then describes what content, like anonymous pages and static assets, can be cached in Varnish and what content should not be cached, like logged-in pages. It also discusses how to configure Varnish and the Drupal Varnish module to integrate caching. The document then covers additional Varnish features like caching authenticated users, purging caches, and tools for monitoring caches.
Varnish intro
- 1. Varnish the Drupal Way
- 2. About me • Boyan Borisov • Team Leader @ Propeople • boyanb@propeople.dk • @boyan_borisov • Skype: boian.borisov • http://linkedin.com/in/ boyanborisov
- 3. What is Varnish? • Reverse proxy cache server • ... and more (load balancer, filter, director, ...) • Open Source (BSD), commercially supported • Typically in front of a LAMP stack
- 4. Why do we need Varnish? • Believe or not Drupal is not very fast by default ;) • “Large Scale Drupal” is not only a sales phrase but the present and the future of Drupal • May be this is the best and fastest way to achieve great performance results
- 5. Why do we need Varnish?
- 6. What does Varnish do?
- 7. What can be cached? • Drupal pages for anonymous users • Static content: images, js, css • Other types of static or generated content • Varnish stores a cache entry for each Host + URL
- 8. What should NOT be cached in Varnish? • Responses to POST requests • Responses with a Set-Cookie : header • User-specific content (logged-in page requests) • Very large media files (> 100MB) • Has a Cache-Control: no-cache header
- 9. When the cache will expire? • Response header: Cache-Control: max-age= 300 • If no Cache-Control specified then Varnish uses the Expires date, if specified • If no Cache-Control or Expires header then uses Varnish's default TTL which is 2 minutes but could be changed
- 10. But wait… what about Drupal? • Install varnish • Install and configure the Drupal varnish module - https://drupal.org/project/varnish • Update the default Varnish VCL configuration with a Drupal related one – https://github.com/boyanborisov/varnish-for-drupal – https://fourkitchens.atlassian.net/wiki/display/TECH/ Configure+Varnish+3+for+Drupal+7 – https://drupal.org/node/1427486 • And don’t forget to enable the caching for the anonymous users
- 12. VCL – Varnish Configuration Language • Extend and modify Varnish cache policies • Add behaviors during each phase of delivery • Each Request is processed separately and independently • States are isolated but are related • Return statements exit one state and start another • VCL defaults are ALWAYS appended below your own VCL
- 13. Other VCL features • Add/remove/change HTTP headers • Redirects based on request parameters • Rewrite requests • Retry failed requests • Saint and Grace: Deliver cached content when backend is unresponsive • ACLs: Network-address filtering • Request filtering ESI: Separate cache entries and separate cache policies for each part of a document • Inline C code
- 14. VCL – Subroutines – breaking it down • vcl_init – VCL is loaded, no request yet; VMOD initialization • vcl_recv – Beginning of request, req is in scope • vcl_pipe – Client & backend data passed unaltered • vcl_pass – Request goes to backend and not cached • vcl_hash – call hash_data to add to the hash • vcl_hit – called on request found in the cache • vcl_miss – called on request not found in the cache • vcl_fetch – called on document retrieved from backend • vcl_deliver – called prior to delivery of cached object • vcl_error – called on errors
- 15. Directors • Directors allow you to talk to the backend servers • Directors are a glorified reverse proxy – Allows for certain types of load balancing – Allows for talking to a cluster “A director is a logical group of backend servers clustered together for redundancy. The basic role of the director is to let Varnish choose a backend server amongst several so if one is down another can be used.”
- 16. Varnish realtime tools • varnishstat – realtime status and performance metrics • varnishhist – live frequency plot of response times • varnishtop – live list of hottest files • varnishlog – analyze incoming HTTP headers • Varnish CLI – live admin capabilities
- 17. Drupal specifics • Drupal emits Vary – Cookie,Accept-Encoding • Cookie – Avoids cache collisions between users Accept-Encoding : maintain separate cache entries for compressed content (gzip, deflate) vs. uncompressed content
- 18. ESI – Edge Side Includes • ESI is a small markup language much like SSI (server side includes) to include fragments (or dynamic content for that matter) • Think of it as replacing regions inside of a page as if you were using XHR (AJAX) but single threaded. • Three Statements can be utilized – esi:include – Include a page – esi:remove – Remove content – <!-- esi --> - ESI disabled, execute normally
- 19. How to cache authenticated users? • Authcache - https://drupal.org/project/authcache “Authcache saves the final rendered HTML of a page to serve visitors. A separate cache is created for each user role as defined by the administrator, so some roles can be excluded if necessary. Authcache places priority on serving pages to the visitor as fast as possible. After a page is loaded by the browser, a second HTTP request may be performed via Ajax. This initiates a lightweight DRUPAL_BOOTSTRAP_SESSION that allows SQL queries to be executed , and returns any user-customized data to the page (such as form tokens or default values on a contact form).”
- 20. How to cache authenticated users?
- 21. How to cache authenticated users?
- 22. How to cache authenticated users? “Production graphs from intranet solution developed by Propeople ( around 50 000 users ). Web CPU usage below 1% and varnish hit ratio > 80%. Varnish & Propeople rocks ” Rumen Yordanov Open Source Department manager @Propeople
- 23. Cache invalidation • Drupal invalidate the whole Varnish cache on node/comment/etc save by default • Expire module is a solution - https://drupal.org/project/expire
- 24. Purging and banning • Purging – a purge is what happens when you pick out an object from the cache and discard it along with its variants. Usually a purge is invoked through HTTP with the method PURGE. • Banning – You can think of bans as a sort of a filter on objects already in the cache. You ban certain content from being served from your cache. You can ban content based on any metadata we have. A ban will only work on objects already in the cache, it does not prevent new content from entering the cache or being served.
- 25. Demo
- 26. Questions?