Prometheus
The CNCF-graduated, pull-based metrics store. Industry default for self-hosted metrics.
| License | Apache-2.0 |
| Repository | https://github.com/prometheus/prometheus |
| GitHub stars (approx) | 56000 |
| Language | Go |
| Self-host deployment | docker, single binary, kubernetes helm |
| Commercial hosted option | no / self-host only |
| Freshness status | Fresh |
| Freshness checked | 2026-04-22 |
| Categories | metrics |
| Replaces SaaS | datadog, new-relic |
Prometheus is a pull-based time-series database designed for reliability and cheap operation. It scrapes /metrics endpoints, stores TSDB, evaluates alerting rules, and forwards alerts to Alertmanager.
Pick it if: you want metrics specifically, and you operate Kubernetes or services that export Prometheus-format metrics.
Skip it if: you need long-term storage at scale (pair with Thanos or Mimir) or you want pushed metrics (look at StatsD gateways or OpenTelemetry Collector).
Deployment: single statically-linked binary. One of the simplest deployments in this directory.