Qconn

Scaling Pinterest

Scaling Pinterest

Location: 
Grand Ballroom A
Time: 
Monday, 5:20pm - 6:10pm
Abstract: 

Pinterest grew to one of the world's largest social networks in just a few years. The first year and half was a scalability rocket ship. We had to grow the architecture hyper fast without much sleep, and had the opportunity to try lots of things, and make LOTS of mistakes before starting to get things under control.

 

Stop by and ask anything.

 

We'll give a quick overview of our architecture, some of the new systems we're building (Pinball, Frontdoor), and talk about some of the tech we use / used for databases (MySQL, MongoDB, Casandra, etc), caching (Memcache, Redis), logging (Flume, Kafka), map reduce (EMR, Qubole, Redshift), logic (Python, Java, Go, Nutcracker), load balancing / HA (haproxy, nginx, Varnish, ZooKeeper), server management (Puppet), and others. We'll keep the presentation relatively short and open the floor for any and all questions.

Yash.Nelapati's picture
Yash Nelapati is an infrastructure engineer at Pinterest where he focusses on scalability, capacity planning and architecure. Prior to pinterest he was into web development and rapidly prototyping UI.
Marty.Weiner's picture
Marty joined Pinterest in early 2011 as the 2nd engineer. Previously worked at Azul Systems as a VM engineer focused on building/improving the JIT compilers in HotSpot.