From Check-Ins to Location Intelligence: How Foursquare Powers Global Data Decisions with Fastly’s Edge Cloud Platform
Key Results
100% of edge logs streamed for real-time observability
Faster issue detection and troubleshooting at global scale
Eliminated maintenance of homegrown logging tools
Improved performance and reliability for enterprise customers
Scaled globally with developer-first edge infrastructure
The challenge
Foursquare has evolved over the last decade from a consumer-facing app to a B2B platform offering APIs and data services to enterprises. This transition raised the stakes for performance and reliability, where every slow response or failed request could have significant contractual implications.
Maintaining Service Level Indicators (SLIs) became essential, but existing internal application logs generated by Kubernetes and microservices were insufficient. These logs provided insight into server-side failures but didn’t illustrate the end-user experience or differentiate between internal issues and external factors like internet conditions.
To resolve this, Foursquare needed complete telemetry into their platform’s traffic and user behaviors, while avoiding the maintenance burden of their legacy log-parsing solutions.
The solution
Foursquare uses Fastly’s Edge Cloud Platform not only for content delivery but as a foundation for their observability strategy. By streaming 100% of their edge logs, Foursquare gains comprehensive, high-fidelity insights into every request, error, and latency fluctuation worldwide. These logs are streamed directly into Observe, a third-party observability platform designed for faster search, correlation, and cost-efficient operations. With this setup, Foursquare can visualize platform health and troubleshoot issues in real time.
Troubleshooting Made Simple
Previously reliant on homegrown log tailing and parsing tools, which required significant upkeep, Foursquare transitioned to Fastly’s real-time log streaming integrated with Observe. This switch eliminated the need for maintaining their internal tools.
"Nowadays, we just send the full firehose of data straight into Observe," said Rob Joseph, VP of Infrastructure at Foursquare. "It lets us visualize and aggregate however we want without maintaining any of our own software."
This integration empowers the team to swiftly detect issues, such as regressions tied to new code deployments. "Fastly logs help us instantly identify if latency spikes or error rates increase," Joseph noted.
Operational Efficiency Through Developer-Centric Tools
Foursquare also maximizes efficiency by leveraging Fastly’s developer-friendly features. They manage over 60 service configurations with an infrastructure-as-code approach, deploying VCL changes directly via API instead of relying on a graphical user interface.
In addition, Foursquare utilizes Fastly Image Optimizer to manage the vast volume of user-generated images from their Swarm app and venue databases. "We used to handle image resizing internally, which was a huge pain. Moving that function to Fastly’s edge was a massive win for us," shared Joseph. "It just works."
Key Takeaway
Fastly has been instrumental in Foursquare’s journey from a consumer app to a global data platform, evolving from a caching layer to a critical tool for business success. By streaming 100% of edge logs into their observability stack, Foursquare ensures it meets the demanding expectations of enterprise clients.
"We wanted to improve the performance for users far from our data centers," Joseph explained. "When we scale up or down, Fastly scales with us—it’s not even a consideration now."
By pairing Fastly’s edge capabilities and high-cardinality logs with robust observability from Observe, Foursquare continues to scale globally without sacrificing reliability or performance.