m3db/m3

Health v2 · maintenance only

Adoption not part of the score

4895 stars · 465 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

M3 is an open-source distributed metrics platform comprising a purpose-built time series database (M3DB), a metrics aggregator, a Prometheus-compatible query engine supporting PromQL and Graphite, and a Prometheus sidecar. It is designed to ingest and query billions of time series at scale, used in production by companies like Uber, Walmart, and LinkedIn.

Use cases

• store and query Prometheus metrics at large scale
• downsample and aggregate metrics streams
• replace Graphite with a scalable storage backend
• run a centralized metrics platform for thousands of services
• query billions of time series with PromQL
• reduce storage costs for long-term metrics retention

When to choose

• you need Prometheus-compatible long-term storage at very large scale
• you want global metrics aggregation and downsampling
• you need Graphite and PromQL support in one engine
• you're running out of headroom with Thanos or VictoriaMetrics at extreme scale

When to avoid

• you have small-scale monitoring needs where plain Prometheus suffices
• you want a simple single-binary setup with minimal operational overhead
• you need non-time-series data storage
• your team cannot operate a distributed cluster with etcd dependencies

Facets

service · maturity stable

database monitoring search-engine streaming api-framework monitoring databases time-series analytics go self-hosted cloud tsdb prometheus graphite metrics-platform query-engine aggregator sidecar promql devops docker kubernetes linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="m3db/m3")