Prometheus Alerting Rules Pack

Prometheus Alerting Rules Pack

Over 40 production-tested alerting rules organized into focused files by concern. Covers Node Exporter system metrics, cAdvisor container metrics, and uptime checks. Drop into any existing Prometheus setup and tune thresholds as needed.

What Is Included

• CPU Rules — High CPU sustained over 5 minutes, steal time alerts for VM environments
• Memory Rules — High memory usage, out-of-memory risk, swap exhaustion
• Disk Rules — Disk filling within 4 hours (predictive), over 85% usage, inode exhaustion
• Network Rules — High error rate, packet drop rate, interface saturation
• Container Rules — Container restart loop, CPU throttling, high memory per container
• Uptime Rules — Instance down, scrape target down, Prometheus job missing

How It Works

• Copy the alerts/ directory alongside your prometheus.yml
• Add a rule_files entry pointing to alerts/*.yml
• Reload Prometheus — rules load immediately and appear in the Rules tab
• Each rule includes severity labels and a human-readable summary annotation for alert managers

What You Get

• alerts/cpu.yml — 7 CPU alerting rules
• alerts/memory.yml — 6 memory alerting rules
• alerts/disk.yml — 8 disk and inode alerting rules
• alerts/network.yml — 6 network alerting rules
• alerts/containers.yml — 10 container alerting rules
• alerts/uptime.yml — 5 uptime and scrape alerting rules
• README.md — integration steps and threshold tuning guide

Requirements

Prometheus 2.x or later. Node Exporter required for system metric rules. cAdvisor required for container metric rules. Both are optional — unused rules remain inactive without errors.