DoorDash's Journey from StatsD to Prometheus - Scaling to 10 Million Metrics per Second

Explore DoorDash's migration from StatsD to Prometheus in this informative conference talk. Discover best practices and lessons learned as Emma Wang and Ben Raskin share their experience scaling to 10 million metrics per second. Learn about the challenges of migrating to Prometheus metrics, including exposing metrics from CI/CD, batch jobs, and short-running tasks. Gain insights into efficiently using histograms at scale across multiple teams and transitioning from centrally aggregated percentile-driven latency metrics. Understand the complexities of adopting Prometheus and PromQL in large-scale environments and how to overcome potential roadblocks in becoming fully instrumented with Prometheus metrics.