Hide Notes 

Slide 1: How to scale (with ruby on rails) George Palmer george.palmer@gmail.com 3dogsbark.com
Slide 2: Overview • • • • • • • • • • One server Two servers Scaling the database Scaling the web server User clusters Final architecture Caching Cached architecture Links Questions George Palmer 17th February 2007
Slide 3: How you start out Shared Hosting Web Server DB • • • • Shared Hosting One web server and DB on same machine Application designed for one machine Volume of traffic will depend on host George Palmer 17th February 2007
Slide 4: Two servers Web Server DB • • • • Possibly still shared hosting Web server and DB on different machine Minimal changes to code Volume of traffic will depend on whether made it to dedicated machines George Palmer 17th February 2007
Slide 5: Scaling the database (1) Slave Web Server Master DB Slave Slave • DB setup more suited to read intensive applications (MySQL replication) • Should be on dedicated hosts • Minimal changes to code George Palmer 17th February 2007
Slide 6: Scaling the database (2) MySQL Cluster Master DB Web Server Master DB • DB setup more suited to equal read/write applications (MySQL cluster) • Should be on dedicated hosts • Minimal changes to code George Palmer 17th February 2007
Slide 7: Scaling the web server Web Server Worker thread Worker thread Worker thread Worker thread DB Farm • Web Server comprises of “Worker threads” that process work as it comes in George Palmer 17th February 2007
Slide 8: Load balancing App Server Load balancer App Server DB Farm App Server • App Server depends: – Rails (Mongrel, FastCGI) – PHP – J2EE • Some changes to code will be required George Palmer 17th February 2007
Slide 9: The story so far… App Server Master DB Slave Load balancer App Server Slave App Server Slave • App servers continue to scale but the database side is somewhat limited… George Palmer 17th February 2007
Slide 10: User Clusters • For each user registered on the service add a entry to a master database detailing where their user data is stored – UserID – DB Cluster – Basic authorisation details such as username, password, any NLS settings George Palmer 17th February 2007
Slide 11: User Clusters (2) SELECT * FROM users WHERE username=‘Bob’ AND … App Server user_id=91732 db_cluster=2 Master DB User clusters are themselves one of the two database setups outlined earlier User Cluster 1 User Cluster 2 George Palmer 17th February 2007
Slide 12: User Clusters (3) • ID management becomes an issue – Best to use master DB id as user_id in user cluster – If let cluster allocate then make sure use offset and increment (not auto_increment) • Other DBs such as session must reference a user by id and DB cluster • Serious code changes may be required • Will want to have ability to move use users between clusters George Palmer 17th February 2007
Slide 13: The final architecture • As number of app servers grow it’s a good idea to add a database connection manager (eg SQLRelay) • Extract out session, search, translation databases onto own machines • Use MySQL cluster (or equivalent) for any critical database – In replication setup can make a slave a backup master • Add a NFS/SAN for static files George Palmer 17th February 2007
Slide 14: The final architecture (2) NFS/SAN App Server 1 Master DB Master DB App Server 2 Load balancer Session DB DB Connection Manager Search DB NLS DB Master … App Server 50 User Cluster 2 Master User Cluster 1 Slave Slave Slave Slave Slave Slave George Palmer 17th February 2007
Slide 15: Issues • Load balancer and database connection manager are single point of failure – Easy solved • 2PC needed for some operations. For example a user wants to be removed from search database – 2PC not supported in rails • Rails doesn’t support database switching for a given model – Can do explicitly on each request but expensive due to connection establishment overhead – Can get round if using connection manager but a proper solution is required (I may write a gem to do this) George Palmer 17th February 2007
Slide 16: Making the most of your assets • In a lot of web applications a huge % of the hits are read only. Hence the need for caching: – Squid • A reverse-proxy (or webserver accelerator) – Memcached • Distributed memory caching solution George Palmer 17th February 2007
Slide 17: Squid App Server 1 Squid … Not in cache App Server 2 In cache NFS/SAN • Lookup of pages is in memory, storing of files is on disk • Can act also act as a load balancer • Pages can be expired by sending DELETE request to proxy George Palmer 17th February 2007
Slide 18: Memcached Physical Machine Physical Machine App Server Memcached App Server DB Farm Memcached • • Location of data is irrespective of physical machine A really nice simple API – SET – GET – DELETE (Not in memcached) In rails only a fews LOC will make a model cached Also useful for tracking cross machine information – eg dodge user behaviour George Palmer 17th February 2007 • •
Slide 19: Cached Architecture • Introduce Squid – Acts as load balancer (note there are higher performing load balancers) • Introduce memcached – Can go on every machine that has spare memory • Best suited to application servers which have high CPU usage but low memory requirements George Palmer 17th February 2007
Slide 20: Cached architecture App Server 1 M C NFS/SAN Master DB Master DB Squid M App Server 2 C … App Server 50 Session DB DB Connection Manager Search DB NLS DB Master M C User Cluster 1 MC=memcached User Cluster 2 Master Slave Slave Slave Slave Slave Slave George Palmer 17th February 2007
Slide 21: Cached architecture • Wikipedia quote a cache hit rate of 78% for squid and 7% for memcached – So only 15% of hits actually get to the DB!! • Performance is a whole new ball game but we recently gained 15-20% by optimising our rails configuration – But don’t get carried away - at some point the time you spend exceeds the money saved George Palmer 17th February 2007
Slide 22: Cached architecture – 1 machine Physical Machine NFS/SAN App Server 1 Master DB Master DB Session DB App Server 2 Squid Memcached … App Server 5 DB Connection Manager Search DB NLS DB User Cluster Master 1 Slave Slave Slave George Palmer 17th February 2007
Slide 23: How far can it go? • For a truly global application, with millions of users - In order of ease: – Have a cache on each continent – Make user clusters based on user location • Distribute the clusters physically around the world – Introduce app servers on each continent – If you must replicate your site globally then use transaction replication software, eg GoldenGate George Palmer 17th February 2007
Slide 24: Useful Links • http://www.squid-cache.org/ • http://www.danga.com/memcached/ • http://sqlrelay.sourceforge.net/ • http://railsexpress.de/blog/ George Palmer 17th February 2007
Slide 25: Questions? George Palmer 17th February 2007