for: 5m
labels:
alertgroup: "prometheus"
+ - name: apache
+ rules:
+ - alert: apache down
+ expr: apache_up == 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: apache workers busy
+ expr: sum(apache_workers{state="busy"}) by (instance) / sum(apache_scoreboard) by (instance) > 0.8
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ busy_workers: "{{ $value | humanizePercentage }}"
- name: database
rules:
- alert: postgres replication delay
alertgroup: database
annotations:
delay: "{{ $value | humanizeDuration }}"
+ - name: filesystem
+ rules:
+ - alert: filesystem low on space
+ expr: node_filesystem_avail_bytes / node_filesystem_size_bytes < 0.05
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ percentage_free: "{{ $value | humanizePercentage }}"
+ free_bytes: "{{ with printf \"node_filesystem_avail_bytes{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value | humanize1024 }}bytes{{end}}"
+ total_total: "{{ with printf \"node_filesystem_size_bytes{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value | humanize1024 }}bytes{{end}}"
+ - alert: filesystem low on inodes
+ expr: node_filesystem_files_free / node_filesystem_files < 0.1
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ percentage_free: "{{ $value | humanizePercentage }}"
+ free_inodes: "{{ with printf \"node_filesystem_files_free{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value }}{{end}}"
+ total_inodes: "{{ with printf \"node_filesystem_files{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value }}{{end}}"
- name: hwmon
rules:
- alert: hwmon fan alarm
alertgroup: "{{ $labels.instance }}"
annotations:
voltage_volts: "{{ with printf \"ipmi_voltage_volts{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}V{{end}}"
+ - alert: ipmi power alarm
+ expr: ipmi_power_state > 0 or ipmi_sensor_state{type=~"Power .*"} > 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
- name: mdadm
rules:
- alert: mdadm array inactive
active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
+ - alert: mdadm array degraded
+ expr: sum (node_md_disks{state="active"}) without (state) < node_md_disks_required
+ for: 0m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
+ active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
+ failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
+ spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
- alert: mdadm disk failed
expr: node_md_disks{state="failed"} > 0
for: 0m
alertgroup: "{{ $labels.instance }}"
annotations:
new_oom_kills: "{{ $value }}"
+ - name: network
+ rules:
+ - alert: interface transmit rate
+ expr: rate(node_network_transmit_bytes_total[1m]) / node_network_speed_bytes > 0.98
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ bandwidth_used: "{{ $value | humanizePercentage }}"
+ - alert: interface receive rate
+ expr: rate(node_network_receive_bytes_total[1m]) / node_network_speed_bytes > 0.98
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ bandwidth_used: "{{ $value | humanizePercentage }}"
+ - alert: interface transmit errors
+ expr: rate(node_network_transmit_errs_total[1m]) / rate(node_network_transmit_packets_total[1m]) > 0.01
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ error_rate: "{{ $value | humanizePercentage }}"
+ - alert: interface receive errors
+ expr: rate(node_network_receive_errs_total[1m]) / rate(node_network_receive_packets_total[1m]) > 0.01
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ error_rate: "{{ $value | humanizePercentage }}"
+ - alert: conntrack entries
+ expr: node_nf_conntrack_entries / node_nf_conntrack_entries_limit > 0.8
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ entries_used: "{{ $value | humanizePercentage }}"
+ - name: postgresql
+ rules:
+ - alert: postgresql down
+ expr: pg_up == 0
+ for: 1m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: postgresql replication delay
+ expr: pg_replication_lag_seconds > 5
+ for: 1m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ delay: "{{ $value | humanizeDuration }}"
+ - alert: postgresql connection limit
+ expr: sum (pg_stat_activity_count) by (instance, server) / sum (pg_settings_max_connections) by (instance, server) > 0.8
+ for: 1m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ connections_used: "{{ $value | humanizePercentage }}"
+ - alert: postgresql deadlocks
+ expr: increase(pg_stat_database_deadlocks[1m]) > 5
+ for: 0m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ new_deadlocks: "{{ $value }}"
+ - alert: postgresql slow queries
+ expr: pg_slow_queries > 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ queries: "{{ $value }}"
+ - name: smart
+ rules:
+ - alert: smart failure
+ expr: smart_health_status == 0
+ for: 60m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: smart ssd wearout approaching
+ expr: smart_percentage_used >= 90
+ for: 60m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ percentage_used: "{{ $value | humanizePercentage }}"
+ - name: ssl
+ rules:
+ - alert: ssl certificate probe failed
+ expr: ssl_probe_success == 0
+ for: 60m
+ labels:
+ alertgroup: ssl
+ - alert: ssl certificate expiry
+ expr: ssl_verified_cert_not_after{chain_no="0"} - time() < 86400 * 14
+ for: 0m
+ labels:
+ alertgroup: ssl
+ annotations:
+ expires_in: "{{ $value | humanizeDuration }}"
+ - alert: ssl certificate revoked
+ expr: ssl_ocsp_response_status == 1
+ for: 0m
+ labels:
+ alertgroup: ssl
+ - alert: ocsp status unknown
+ expr: ssl_ocsp_response_status == 1
+ for: 0m
+ labels:
+ alertgroup: ssl
+ - name: systemd
+ rules:
+ - alert: systemd failed service
+ expr: node_systemd_unit_state{state="failed"} == 1
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
- name: tile
rules:
- alert: renderd replication delay
alertgroup: "{{ $labels.instance }}"
annotations:
skew: "{{ with printf \"node_timex_offset_seconds{instance='%s'}\" $labels.instance | query }} {{ . | humanizeDuration }}{{ end }}"
+ - name: web
+ rules:
+ - alert: web error rate
+ expr: sum(rate(api_call_count_total{status=~"5.*"}[5m])) by (instance) / sum(rate(api_call_count_total[5m])) by (instance) > 0.002
+ for: 5m
+ labels:
+ alertgroup: web
+ annotations:
+ error_rate: "{{ $value | humanizePercentage }}"
projects / chef.git / blobdiff