Case Study: Uber achieves 8.53x more cost-effective metrics storage with CNCF Prometheus and M3

How Uber is monitoring 4,000 microservices with its open sourced Prometheus platform

Uber faced a significant challenge in monitoring its 4,000 proprietary microservices and numerous open source systems, as it had outgrown its use of Graphite and Nagios. The setup was hard to maintain at scale and struggled with the volume of metrics generated. They sought a solution from the Cloud Native Computing Foundation ecosystem.

The solution involved implementing the Prometheus monitoring system, and Uber also built and open sourced the M3 platform to serve as a scalable store for Prometheus metrics. With this solution from the Cloud Native Computing Foundation, Uber's storage costs became 8.53x more cost effective, setting up new monitoring systems was 4x faster, and operational maintenance burden was reduced by 16.67x.

Uber

Rob Skillington

Technical Lead for Metrics and Systems Monitoring