Build high-performance, scalable applications with Mapbox
Our extensive network runs in over 60 locations around the world.
Our edge network is backed by regional origin infrastructure running in nine AWS regions on five continents. Our globally distributed network is important for low latency delivery of cached maps and protects us from failure scenarios like natural disasters.
Our edge network uses Anycast DNS and supports the EDNS-client-subnet extension of DNS, ensuring traffic is served by the closest edge location. Our edge servers and regional origins also use Anycast DNS and latency based routing.
Our web service traffic ranges from thousands to several tens of thousands of requests per second. Mapbox can handle these requests because each regional origin scales independently based on real time traffic.
Fast and stable user experience
Automated monitoring and disaster response to ensure stability.
Our system metrics, dashboard, and alarms help keep Mapbox online
We have a dashboard to monitor system metrics and use front-end health check probes.
All server and application logs are sent to a central logging service.
Alarms are defined on system metrics, system logs, front-end metrics, and application logs.
Mapbox engineers are on call 24/7/365 to receive and respond to alarms.
Our status page shows users our current uptime and if we’re experiencing downtime when it will be remediated.
Failure and disaster recovery
Mapbox is prepared to handle disaster recovery and deal with signs of failure.
We use automated failover on all origins.
We use health check probes around the world by automatically removing unavailable origins and re-routing to healthy ones.
Our databases are redundant and run in multiple regions and data centers.
We take snapshot backups several times a day, automatically verified, and vaulted for safe keeping.
We've taken several measures to keep our platform and data secure