Case Study: Foursquare achieves double database performance and a scalable, cost-flat hybrid cloud with Equinix

Foursquare - Customer Case Study

Foursquare, a location-based discovery app serving over 45 million users and 1.6 million merchants, outgrew its AWS-only infrastructure as transaction, network and disk performance limits made scaling costly and opaque. The engineering team needed more control, predictable network performance and a platform that could support customized compute and big-data analytics while still using AWS for warehousing and burst capacity.

Foursquare built a private cloud inside Equinix’s Ashburn data center, deploying purpose-built mongoDB clusters and a Hadoop analytics stack and linking them to AWS via AWS Direct Connect. The hybrid design doubled transaction processing while holding costs flat, cut average latency from ~11 ms to ~2.5 ms, sped data transfers fivefold (enabling a 6+ TB nightly sync in six hours instead of days), and delivered seamless migration and multi-cloud flexibility.

Foursquare

Robert Joseph

Director of Site Reliability Engineering