Textbook of Linux

← Software

VictoriaMetrics

A fast, cost-efficient time-series database compatible with Prometheus.

VictoriaMetrics speaks the Prometheus wire and query languages — its servers can scrape Prometheus exporters directly, and its endpoint accepts the same /api/v1/query and remote-write requests Prometheus uses — but typically delivers substantially better resource efficiency at scale: less RAM per active series, better disk compression, and faster query latency on large-cardinality metric stores.

The project was started in 2018 by Aliaksandr Valialkin, who had previously worked on the fasthttp HTTP library and other high-performance Go projects. The core engine is open source under the Apache-2.0 licence; the Cluster and Enterprise versions add scale-out and operational features.

VictoriaMetrics has gained significant adoption among organisations with large Prometheus deployments that hit single-node Prometheus's resource ceiling. The cluster version's scale-out architecture (separate vmstorage, vmselect, and vminsert components) handles many millions of active series across nodes. It is also commonly used as the long-term storage back-end behind Grafana dashboards.

Download from https://github.com/VictoriaMetrics/VictoriaMetrics/releases
Or via Docker:
  docker run -p 8428:8428 victoriametrics/victoria-metrics:latest