M3
Horizontally scalable, distributed time-series database and metrics platform built at Uber for very large-scale monitoring.
Overview
M3 is a distributed metrics platform - storage engine, query layer, and aggregation - designed to handle the write and storage volume of monitoring thousands of services at very large scale, well beyond what a single Prometheus instance handles alone. It exposes a Prometheus-remote-write-compatible interface so existing Prometheus setups can use it as a scalable backend.
M3 is a fit for large organizations whose metrics volume has outgrown a single Prometheus instance and need a purpose-built, horizontally scalable storage layer underneath their existing setup.
- Categories
- Monitoring
- Languages
- Go
- License
- Apache-2.0
Related Projects
Graphite
Apache-2.0Time-series metrics storage and graphing system that has been a foundational building block of monitoring stacks for over a decade.
- Monitoring
InfluxDB
MITPurpose-built time-series database for metrics, events, and real-time analytics at high write volumes.
- Monitoring
Fluentd
Apache-2.0CNCF-graduated log collector that unifies data collection from many sources into a single, structured logging layer.
- Monitoring
Grafana
AGPL-3.0Open source platform for monitoring and observability, letting you query, visualize and alert on metrics.
- Monitoring
Spotted an error? Suggest an edit on GitHub.