nixfiles/hosts/monitoring-3/rules.yml

groups:
- name: alert.rules
  rules:
  - alert: InstanceDown
    expr: up{job="node-exporter"} == 0
    for: 5m
    labels:
      severity: critical
    annotations:
      summary: "Endpoint {{ $labels.instance }} down"
      description: "{{ $labels.instance }} is unreachable"
  - alert: InstanceConfigNotSync
    expr: nixos_current_system_is_sync{} == 0
    for: 2h
    labels:
      severity: warning
    annotations:
      summary: "Current system of {{ $labels.instance }} not in sync with config"
      description: "The current system hash of {{ $labels.instance }} does not match the one generated by hydra based on the current config"
  - alert: BackupStorageFull
    expr: ((avg_over_time(node_filesystem_avail_bytes{job="node-exporter"}[5m]) / avg_over_time(node_filesystem_size_bytes{job="node-exporter"}[5m])) * 100) < 5
    for: 30m
    labels:
      severity: critical
    annotations:
      summary: "Storage of {{ $labels.instance }} is full"
      description: "Storage of {{ $labels.instance }} for {{ $labels.mountpoint }} on {{ $labels.device }} is full"
  - alert: ClerieBackupJobLastSuccessfulRunBehind
    expr: time() - clerie_backup_last_successful_run_time{} >= 9000
    for: 5m
    labels:
      severity: warning
    annotations:
      summary: "Last successful backup on {{ $labels.instance }} older than 1h"
      description: "Backup job {{ $labels.backup_job }} to target {{ $labels.backup_instance }} finished successfully over an hour ago"
  - alert: InstanceJustBooted
    expr: time() - node_boot_time_seconds{job="node-exporter"} <= 300
    labels:
      severity: warning
    annotations:
      summary: "Uptime of {{ $labels.instance }} less than 5 min"
      description: "{{ $labels.instance }} just booted"
  - alert: EventInstanceDown
    expr: up{job="node-exporter", service_level="event"} == 0
    for: 2h
    labels:
      mute: instance
      severity: muted
    annotations:
      summary: "Event instance {{ $labels.instance }} down"
      description: "{{ $labels.instance }} has been down for more than 2 hours."
  - alert: KernelChanged
    expr: avg_over_time(nixos_current_system_kernel_is_booted_system_kernel{job="nixos-exporter"}[5m]) == 0
    for: 2h
    labels:
      severity: warning
    annotations:
      summary: "Kernel of {{ $labels.instance }} changed"
      description: "The Kernel {{ $labels.instance }} booted with, isn't the target Kernel. A reboot may be required."