How to Avoid Pitfalls : Scaling Enterprise Search to the Extreme

By on January 16, 2013

An excellent video with pdf presentation by the top architects at Pinterest on how they scaled their stack while facing incredible growth to achieve billions of page views per month. Starting with a group of technologies including ElasticSearch they soon faced a growing set of unstable and failing systems. Marty Weiner of Pinterest explains, “when you’re doubling in scale every month things go nuts … things are always breaking all the time and you have all these people coming to you with fantastic promises ‘Use our system and all you have to do is add a box and you’re done.’” Faced with an increasing number of breakdowns they sat down in January of 2012 and did a full re-architecture of everything they had. They started by asking lots of questions about what actually works and changed their whole system to make it simpler. At this time they moved to Solr 4 and have never looked back. The new architecture gave them an unlimited ability to scale and could finally satisfy their #1 requirement – To keep up with the exponential growth in traffic on

Watch the video | View the slides

Pinterest Growth Rate

Share on LinkedInShare on FacebookTweet about this on Twitter

Related Posts

DuckDuckGo – 1 Million Deep Searches A Day And Growing

Enterprise Social Media: Boris Aleksandrovsky of Yammer

Closing the Knowledge Gap: How Cisco Unlocks Communications

Solr Development Case Study:

Top Posts

Understanding Transaction Logs, Soft Commit and Commit in SolrCloud

Faceted Search with Solr

Nested Queries in Solr

Posted in Case Study, SearchHub, Why Lucene Solr? with tags #Apache #Big data #ElasticSearch #Enterprise Search #Lucene #Lucene/Solr Case Studies #Open Source Search #Performance #Scalability #Solr #Technical Articles