Prometheus Monitoring System

Prometheus is a free and open-source systems monitoring and alerting toolkit that collects, stores, and queries time-series metrics from instrumented applications and infrastructure. Originally developed at SoundCloud by Julius Volz and Bjorn Rabenstein in 2012, Prometheus became a CNCF graduated project in 2018 (the second after Kubernetes) and is the de facto standard for cloud-native monitoring. Key features: multi-dimensional data model: metrics are stored as time-series identified by a metric name and a set of key-value labels (dimensions), enabling flexible slicing and dicing of data. For example, http_requests_total{method='GET', status='200', handler='/api'}. PromQL: a powerful query language for selecting, aggregating, and computing over time-series data. Supports rate calculations, statistical functions (histogram_quantile for percentiles), arithmetic between metrics, and prediction. Pull model: Prometheus actively scrapes metrics from instrumented targets via HTTP endpoints (/metrics), simplifying deployment and providing built-in health checking. Push model also supported via Pushgateway for short-lived batch jobs. Time-series database: stores metrics locally in a custom compressed format with efficient encoding, handling millions of samples per second. Retention is configurable (default 15 days). Service discovery: automatically discovers scrape targets from Kubernetes, EC2, Consul, DNS, and many other sources. Alerting: the Alertmanager component handles alert routing, deduplication, grouping, and silencing, with notifications via email, Slack, PagerDuty, WebHook, and more. Exporters: hundreds of exporters provide metrics for third-party systems (node_exporter for OS metrics, mysqld_exporter, blackbox_exporter for HTTP/DNS/ICMP probing). Federation for hierarchical setups. Go binary. Apache-2.0.