📈Prometheus

NEW

Open-source systems monitoring and alerting toolkit with a pull-based metrics model and powerful time-series querying.

DESCRIPTION

Prometheus collects metrics from instrumented targets using a pull model over HTTP and stores them as time series. It provides a multidimensional data model, a flexible query language (PromQL), and an Alertmanager for alert routing. With exporters and service discovery, it scales across dynamic environments like Kubernetes. Prometheus underpins SLI measurement and SLO alerting in modern reliability practices.

ALTERNATIVE DEFINITIONS

Prometheus Project

Prometheus is an open-source systems monitoring and alerting toolkit originally built at SoundCloud.

Highlights open-source nature and origin, with focus on monitoring and alerting.

Source

CNCF

Prometheus is an open source systems monitoring and alerting toolkit.

Positions Prometheus within the CNCF ecosystem as a core observability project.

Source

RELATED TERMS

Grafana Alertmanager PromQL OpenMetrics OpenTelemetry Time Series Database Service Level Indicator Service Level Objective Exporters

COMPETING TERMS

InfluxDB Datadog

RELATIONS

• Conforms to: OpenTelemetry
• Narrower: Alertmanager

KNOWLEDGE AREAS

Observability & Telemetry Infrastructure Platform & Tooling

TARGET AUDIENCE

DevOps Engineer Cloud Architect Software Developer MLOps Engineer Data Engineer IT Manager

EXTERNAL SOURCES

• Wikipedia - Prometheus (software)
• Prometheus Project - Best practices for instrumentation
• Prometheus Project - Getting started with Prometheus server
• OpenMetrics - Metrics exposition format used by Prometheus

TOP QUESTIONS

How does Prometheus collect metrics?

It scrapes instrumented targets over HTTP in a pull model. Exporters expose metrics endpoints for third-party systems.

Category: Overview

What is PromQL in Prometheus?

PromQL is a powerful query language for selecting and aggregating time series. It supports labels for multidimensional filtering.

Category: Definition

How does Prometheus support SLOs?

SLIs are computed with PromQL, and alerting uses burn-rate rules routed via Alertmanager to enforce SLO error budgets.

Category: Example

What are common Prometheus pitfalls?

High-cardinality labels explode storage and query costs. Scrape misconfigurations and unbounded metrics can overwhelm resources.

Category: Mistakes

Is Prometheus push-based?

Primarily pull-based. A Pushgateway exists for short-lived jobs, but pull remains preferred for reliability and discovery.

Category: Misunderstanding

How does Prometheus scale?

Use federation, sharding, or long-term storage integrations (e.g., Cortex, Thanos, Mimir) for horizontal scale and retention.

Category: Management

What types of metrics does Prometheus recommend?

Counters, gauges, histograms, and summaries, with emphasis on histograms for latency distributions.

Category: BestPractices

How does Prometheus integrate with Kubernetes?

Service discovery finds pods and services automatically. Exporters and annotations expose metrics endpoints.

Category: Integration

What is Alertmanager?

A component for deduplicating, grouping, and routing alerts to receivers like email, Slack, or PagerDuty.

Category: Definition