What is Prometheus?

Short answer

Prometheus is an open-source monitoring and alerting toolkit. It collects metrics via HTTP pull requests, stores them as time-series data, and evaluates alerting rules.

How it works

Prometheus scrapes metrics from instrumented applications and exporters at regular intervals. Data is stored locally in a custom time-series database. PromQL is used to query the data.

Example

scrape_configs:
  - job_name: 'app'
    static_configs:
      - targets: ['app:8080']

Why it matters

Prometheus is the de facto standard for cloud-native monitoring. It integrates with Kubernetes, supports dynamic service discovery, and has a rich ecosystem of exporters.

Tips

• Use recording rules for expensive queries.
• Set appropriate retention periods.
• Use remote storage for long-term data.

Common issues

• High cardinality labels can cause performance issues.
• Scraping failures may be due to network issues.
• Alert fatigue occurs with poorly tuned rules.