Case Study: Roblox achieves centralized observability and unified tracing with Grafana Cloud

Multiple players, one stack Inside Roblox's centralized observability stack

Roblox, the global gaming platform serving more than 214 million monthly users, needed a way to replace its fragmented, team-by-team observability setup. Different engineering groups were using separate Prometheus, InfluxDB, and custom visualization tools, which made incident response slow, troubleshooting difficult, and global monitoring hard to scale. Roblox turned to Grafana Cloud to help centralize its observability approach.

With Grafana, Roblox built a unified telemetry stack using VictoriaMetrics and Grafana Cloud for visualizations, then added Grafana Cloud Traces to bring metrics, logs, and traces into one place. The result was a single Grafana instance supporting more than 17,000 dashboards, 100% availability for the first three quarters after the migration, and the ability to ingest more than 120 million data points per second.

Roblox

Xiaofeng Han

Engineering Director