PROMETHEUS-DETAIL

CNCF pull-based monitoring TSDB.

Definition

Prometheus model: Counter, Gauge, Histogram, Summary. Scrape via /metrics endpoint Prometheus exposition format. Recording Rules for pre-aggregation. AlertManager for notifications (PagerDuty, Slack, Email). Native Histograms (sparse) since v2.40. Remote Write for long-term storage (Mimir, Thanos, VictoriaMetrics).

Origin

Created by Matt T. Proud and Julius Volz at SoundCloud in 2012 ; open-sourced 2015 ; CNCF graduation August 2018.

Example in context

A Kubernetes team monitors 5000 pods with Prometheus scraping every 15s and a Slack-integrated AlertManager.

Related terms

• AlertManager — notifications component.