Prometheus Project Intro
summary
Prometheus is a time‑series monitoring ecosystem built as a single Go binary with a local TSDB that stores samples at roughly 1–2 bytes per sample, able to handle millions of active series and over a million samples/sec on big machines. It scrapes metrics over HTTP (instrumentation or exporters like node_exporter), uses a label/dimensional model and PromQL for slicing, evaluates alerting rules locally, and relies on remote systems like Cortex or Thanos for durable, scalable long‑term storage.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 2:43History of Prometheus
- 4:09Motivation
- 4:55Architecture
- 7:38Selling Points
- 8:34Data Model
- 12:23Query Language
- 23:33Alerts
- 31:26Scaling
