cookbooks/prometheus/templates/default/alert_rules.yml.erb

   1 # DO NOT EDIT - This file is being maintained by Chef
   2
   3 groups:
   4   - name: alertmanager
   5     rules:
   6       - alert: prometheus target missing
   7         expr: up == 0
   8         for: 5m
   9         labels:
  10           alertgroup: "prometheus"
  11   - name: amsterdam
  12     rules:
  13       - alert: pdu current draw
  14         expr: rPDU2PhaseStatusCurrent{site="amsterdam",rPDU2PhaseStatusIndex="1"} / 10 > 10
  15         for: 5m
  16         labels:
  17           alertgroup: "amsterdam"
  18         annotations:
  19           current: "{{ $value | humanize }}A"
  20       - alert: site current draw
  21         expr: sum(rPDU2PhaseStatusCurrent{site="amsterdam",rPDU2PhaseStatusIndex="1"} / 10) > 13
  22         for: 5m
  23         labels:
  24           alertgroup: "amsterdam"
  25         annotations:
  26           current: "{{ $value | humanize }}A"
  27       - alert: site temperature
  28         expr: min(rPDU2SensorTempHumidityStatusTempC{site="amsterdam"}) / 10 < 18 or min(rPDU2SensorTempHumidityStatusTempC{site="amsterdam"}) / 10 > 25
  29         for: 5m
  30         labels:
  31           alertgroup: "amsterdam"
  32         annotations:
  33           temperature: "{{ $value | humanize }}C"
  34       - alert: site humidity
  35         expr: max(rPDU2SensorTempHumidityStatusRelativeHumidity{site="amsterdam"}) / 100 < 0.25 or max(rPDU2SensorTempHumidityStatusRelativeHumidity{site="amsterdam"}) / 100 > 0.65
  36         for: 5m
  37         labels:
  38           alertgroup: "amsterdam"
  39         annotations:
  40           humidity: "{{ $value | humanizePercentage }}"
  41   - name: apache
  42     rules:
  43       - alert: apache down
  44         expr: apache_up == 0
  45         for: 5m
  46         labels:
  47           alertgroup: "{{ $labels.instance }}"
  48       - alert: apache workers busy
  49         expr: sum(apache_workers{state="busy"}) by (instance) / sum(apache_scoreboard) by (instance) > 0.8
  50         for: 5m
  51         labels:
  52           alertgroup: "{{ $labels.instance }}"
  53         annotations:
  54           busy_workers: "{{ $value | humanizePercentage }}"
  55       - alert: apache low request rate
  56         expr: rate(apache_accesses_total[$__rate_interval]) / rate(apache_accesses_total[$__rate_interval] offset 1w) < 0.25 and rate(apache_accesses_total[1h] offset 1w) > 2
  57         for: 15m
  58         labels:
  59           alertgroup: "{{ $labels.instance }}"
  60         annotations:
  61           request_rate: "{{ $value | humanizePercentage }}"
  62   - name: chef
  63     rules:
  64       - alert: chef client not running
  65         expr: time() - node_systemd_timer_last_trigger_seconds{name="chef-client.timer"} > 3600
  66         for: 12h
  67         labels:
  68           alertgroup: "{{ $labels.instance }}"
  69         annotations:
  70           down_time: "{{ $value | humanizeDuration }}"
  71   - name: cpu
  72     rules:
  73       - alert: cpu pressure
  74         expr: rate(node_pressure_cpu_waiting_seconds_total[5m]) > 0.6
  75         for: 15m
  76         labels:
  77           alertgroup: "{{ $labels.instance }}"
  78         annotations:
  79           pressure: "{{ $value | humanizePercentage }}"
  80   - name: database
  81     rules:
  82       - alert: postgres replication delay
  83         expr: pg_replication_lag_seconds > 5
  84         for: 5m
  85         labels:
  86           alertgroup: database
  87         annotations:
  88           delay: "{{ $value | humanizeDuration }}"
  89   - name: fastly
  90     rules:
  91       - alert: error rate
  92         expr: sum(rate(fastly_rt_status_group_total{status_group="5xx"}[5m])) by (service_name, datacenter) / sum(rate(fastly_rt_status_group_total[5m])) by (service_name, datacenter) > 0.005
  93         for: 15m
  94         labels:
  95           alertgroup: fastly
  96         annotations:
  97           error_rate: "{{ $value | humanizePercentage }}"
  98   - name: filesystem
  99     rules:
 100       - alert: readonly filesystem
 101         expr: node_filesystem_readonly == 1
 102         for: 0m
 103         labels:
 104           alertgroup: "{{ $labels.instance }}"
 105       - alert: filesystem low on space
 106         expr: node_filesystem_avail_bytes / node_filesystem_size_bytes < 0.05
 107         for: 5m
 108         labels:
 109           alertgroup: "{{ $labels.instance }}"
 110         annotations:
 111           percentage_free: "{{ $value | humanizePercentage }}"
 112           free_bytes: "{{ with printf \"node_filesystem_avail_bytes{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value | humanize1024 }}bytes{{end}}"
 113           total_total: "{{ with printf \"node_filesystem_size_bytes{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value | humanize1024 }}bytes{{end}}"
 114       - alert: filesystem low on inodes
 115         expr: node_filesystem_files_free / node_filesystem_files < 0.1
 116         for: 5m
 117         labels:
 118           alertgroup: "{{ $labels.instance }}"
 119         annotations:
 120           percentage_free: "{{ $value | humanizePercentage }}"
 121           free_inodes: "{{ with printf \"node_filesystem_files_free{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value }}{{end}}"
 122           total_inodes: "{{ with printf \"node_filesystem_files{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value }}{{end}}"
 123   - name: hwmon
 124     rules:
 125       - alert: hwmon fan alarm
 126         expr: node_hwmon_fan_alarm == 1
 127         for: 5m
 128         labels:
 129           alertgroup: "{{ $labels.instance }}"
 130         annotations:
 131           fan_rpm: "{{ with printf \"node_hwmon_fan_rpm{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}rpm{{end}}"
 132           fan_min_rpm: "{{ with printf \"node_hwmon_fan_min_rpm{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}rpm{{end}}"
 133       - alert: hwmon temperature alarm
 134         expr: node_hwmon_temp_alarm == 1
 135         for: 5m
 136         labels:
 137           alertgroup: "{{ $labels.instance }}"
 138         annotations:
 139           temp_celsius: "{{ with printf \"node_hwmon_temp_celsius{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}C{{end}}"
 140           temp_max_celsius: "{{ with printf \"node_hwmon_temp_max_celsius{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}C{{end}}"
 141           temp_crit_celsius: "{{ with printf \"node_hwmon_temp_crit_celsius{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}C{{end}}"
 142       - alert: hwmon voltage alarm
 143         expr: node_hwmon_in_alarm == 1
 144         for: 5m
 145         labels:
 146           alertgroup: "{{ $labels.instance }}"
 147         annotations:
 148           in_volts: "{{ with printf \"node_hwmon_in_volts{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}V{{end}}"
 149           in_min_volts: "{{ with printf \"node_hwmon_in_min_volts{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}V{{end}}"
 150           in_max_volts: "{{ with printf \"node_hwmon_in_max_volts{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}V{{end}}"
 151   - name: io
 152     rules:
 153       - alert: io pressure
 154         expr: rate(node_pressure_io_waiting_seconds_total[5m]) > 0.6
 155         for: 60m
 156         labels:
 157           alertgroup: "{{ $labels.instance }}"
 158         annotations:
 159           pressure: "{{ $value | humanizePercentage }}"
 160   - name: ipmi
 161     rules:
 162       - alert: ipmi fan alarm
 163         expr: ipmi_fan_speed_state > 0
 164         for: 5m
 165         labels:
 166           alertgroup: "{{ $labels.instance }}"
 167         annotations:
 168           fan_speed_rpm: "{{ with printf \"ipmi_fan_speed_rpm{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}rpm{{end}}"
 169       - alert: ipmi temperature alarm
 170         expr: ipmi_temperature_state > 0
 171         for: 5m
 172         labels:
 173           alertgroup: "{{ $labels.instance }}"
 174         annotations:
 175           temperature_celsius: "{{ with printf \"ipmi_temperature_celsius{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}C{{end}}"
 176       - alert: ipmi voltage alarm
 177         expr: ipmi_voltage_state > 0
 178         for: 5m
 179         labels:
 180           alertgroup: "{{ $labels.instance }}"
 181         annotations:
 182           voltage_volts: "{{ with printf \"ipmi_voltage_volts{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}V{{end}}"
 183       - alert: ipmi power alarm
 184         expr: ipmi_power_state > 0 or ipmi_sensor_state{type=~"Power .*"} > 0
 185         for: 5m
 186         labels:
 187           alertgroup: "{{ $labels.instance }}"
 188   - name: mail
 189     rules:
 190       - alert: exim queue length
 191         expr: exim_queue > exim_queue_limit
 192         for: 60m
 193         labels:
 194           alertgroup: mail
 195         annotations:
 196           queue_length: "{{ $value }}"
 197       - alert: mailman queue length
 198         expr: mailman_queue_length > 200
 199         for: 60m
 200         labels:
 201           alertgroup: mail
 202         annotations:
 203           queue_length: "{{ $value }}"
 204   - name: mdadm
 205     rules:
 206       - alert: mdadm array inactive
 207         expr: node_md_state{state="inactive"} > 0
 208         for: 0m
 209         labels:
 210           alertgroup: "{{ $labels.instance }}"
 211         annotations:
 212           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 213           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 214           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 215           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 216       - alert: mdadm array degraded
 217         expr: sum (node_md_disks{state="active"}) without (state) < node_md_disks_required
 218         for: 0m
 219         labels:
 220           alertgroup: "{{ $labels.instance }}"
 221         annotations:
 222           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 223           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 224           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 225           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 226       - alert: mdadm disk failed
 227         expr: node_md_disks{state="failed"} > 0
 228         for: 0m
 229         labels:
 230           alertgroup: "{{ $labels.instance }}"
 231         annotations:
 232           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 233           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 234           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 235           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 236   - name: memory
 237     rules:
 238       - alert: low memory
 239         expr: node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes < 0.1
 240         for: 15m
 241         labels:
 242           alertgroup: "{{ $labels.instance }}"
 243         annotations:
 244           memory_free: "{{ $value | humanizePercentage }}"
 245       - alert: memory pressure
 246         expr: rate(node_pressure_memory_waiting_seconds_total[5m]) > 0.6
 247         for: 60m
 248         labels:
 249           alertgroup: "{{ $labels.instance }}"
 250         annotations:
 251           pressure: "{{ $value | humanizePercentage }}"
 252       - alert: oom kill detected
 253         expr: increase(node_vmstat_oom_kill[1m]) > 0
 254         for: 0m
 255         labels:
 256           alertgroup: "{{ $labels.instance }}"
 257         annotations:
 258           new_oom_kills: "{{ $value }}"
 259   - name: network
 260     rules:
 261       - alert: interface transmit rate
 262         expr: rate(node_network_transmit_bytes_total[1m]) / node_network_speed_bytes > 0.98
 263         for: 5m
 264         labels:
 265           alertgroup: "{{ $labels.instance }}"
 266         annotations:
 267           bandwidth_used: "{{ $value | humanizePercentage }}"
 268       - alert: interface receive rate
 269         expr: rate(node_network_receive_bytes_total[1m]) / node_network_speed_bytes > 0.98
 270         for: 5m
 271         labels:
 272           alertgroup: "{{ $labels.instance }}"
 273         annotations:
 274           bandwidth_used: "{{ $value | humanizePercentage }}"
 275       - alert: interface transmit errors
 276         expr: rate(node_network_transmit_errs_total[1m]) / rate(node_network_transmit_packets_total[1m]) > 0.01
 277         for: 5m
 278         labels:
 279           alertgroup: "{{ $labels.instance }}"
 280         annotations:
 281           error_rate: "{{ $value | humanizePercentage }}"
 282       - alert: interface receive errors
 283         expr: rate(node_network_receive_errs_total[1m]) / rate(node_network_receive_packets_total[1m]) > 0.01
 284         for: 5m
 285         labels:
 286           alertgroup: "{{ $labels.instance }}"
 287         annotations:
 288           error_rate: "{{ $value | humanizePercentage }}"
 289       - alert: conntrack entries
 290         expr: node_nf_conntrack_entries / node_nf_conntrack_entries_limit > 0.8
 291         for: 5m
 292         labels:
 293           alertgroup: "{{ $labels.instance }}"
 294         annotations:
 295           entries_used: "{{ $value | humanizePercentage }}"
 296   - name: planet
 297     rules:
 298       - alert: planet dump overdue
 299         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/(pbf|planet)/.*"} > 7 * 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 300         for: 24h
 301         labels:
 302           alertgroup: planet
 303         annotations:
 304           overdue_by: "{{ $value | humanizeDuration }}"
 305       - alert: notes dump overdue
 306         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/notes/.*"} > 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 307         for: 6h
 308         labels:
 309           alertgroup: planet
 310         annotations:
 311           overdue_by: "{{ $value | humanizeDuration }}"
 312       - alert: daily replication feed delayed
 313         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/day/.*"} > 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 314         for: 3h
 315         labels:
 316           alertgroup: planet
 317         annotations:
 318           delayed_by: "{{ $value | humanizeDuration }}"
 319       - alert: hourly replication feed delayed
 320         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/hour/.*"} > 3600 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 321         for: 30m
 322         labels:
 323           alertgroup: planet
 324         annotations:
 325           delayed_by: "{{ $value | humanizeDuration }}"
 326       - alert: minutely replication feed delayed
 327         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/minute/.*"} > 60 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 328         for: 5m
 329         labels:
 330           alertgroup: planet
 331         annotations:
 332           delayed_by: "{{ $value | humanizeDuration }}"
 333       - alert: changeset replication feed delayed
 334         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/changesets/.*"} > 60 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 335         for: 5m
 336         labels:
 337           alertgroup: planet
 338         annotations:
 339           delayed_by: "{{ $value | humanizeDuration }}"
 340   - name: postgresql
 341     rules:
 342       - alert: postgresql down
 343         expr: pg_up == 0
 344         for: 1m
 345         labels:
 346           alertgroup: "{{ $labels.instance }}"
 347       - alert: postgresql replication delay
 348         expr: pg_replication_lag_seconds > 5
 349         for: 1m
 350         labels:
 351           alertgroup: "{{ $labels.instance }}"
 352         annotations:
 353           delay: "{{ $value | humanizeDuration }}"
 354       - alert: postgresql connection limit
 355         expr: sum (pg_stat_activity_count) by (instance, server) / sum (pg_settings_max_connections) by (instance, server) > 0.8
 356         for: 1m
 357         labels:
 358           alertgroup: "{{ $labels.instance }}"
 359         annotations:
 360           connections_used: "{{ $value | humanizePercentage }}"
 361       - alert: postgresql deadlocks
 362         expr: increase(pg_stat_database_deadlocks[1m]) > 5
 363         for: 0m
 364         labels:
 365           alertgroup: "{{ $labels.instance }}"
 366         annotations:
 367           new_deadlocks: "{{ $value }}"
 368       - alert: postgresql slow queries
 369         expr: pg_slow_queries > 0
 370         for: 5m
 371         labels:
 372           alertgroup: "{{ $labels.instance }}"
 373         annotations:
 374           queries: "{{ $value }}"
 375   - name: smart
 376     rules:
 377       - alert: smart failure
 378         expr: smart_health_status == 0
 379         for: 60m
 380         labels:
 381           alertgroup: "{{ $labels.instance }}"
 382       - alert: smart ssd wearout approaching
 383         expr: smart_percentage_used >= 90
 384         for: 60m
 385         labels:
 386           alertgroup: "{{ $labels.instance }}"
 387         annotations:
 388           percentage_used: "{{ $value | humanizePercentage }}"
 389   - name: ssl
 390     rules:
 391       - alert: ssl certificate probe failed
 392         expr: ssl_probe_success == 0
 393         for: 60m
 394         labels:
 395           alertgroup: ssl
 396       - alert: ssl certificate expiry
 397         expr: ssl_verified_cert_not_after{chain_no="0"} - time() < 86400 * 14
 398         for: 0m
 399         labels:
 400           alertgroup: ssl
 401         annotations:
 402           expires_in: "{{ $value | humanizeDuration }}"
 403       - alert: ssl certificate revoked
 404         expr: ssl_ocsp_response_status == 1
 405         for: 0m
 406         labels:
 407           alertgroup: ssl
 408       - alert: ocsp status unknown
 409         expr: ssl_ocsp_response_status == 1
 410         for: 0m
 411         labels:
 412           alertgroup: ssl
 413   - name: statuscake
 414     rules:
 415       - alert: statuscake uptime check failing
 416         expr: statuscake_uptime{status="down",paused="false"} > 0
 417         for: 0m
 418         labels:
 419           alertgroup: statuscake
 420   - name: systemd
 421     rules:
 422       - alert: systemd failed service
 423         expr: node_systemd_unit_state{state="failed",name!="chef-client.service"} == 1
 424         for: 5m
 425         labels:
 426           alertgroup: "{{ $labels.instance }}"
 427       - alert: systemd failed service
 428         expr: node_systemd_unit_state{state="failed",name="chef-client.service"} == 1
 429         for: 6h
 430         labels:
 431           alertgroup: "{{ $labels.instance }}"
 432   - name: tile
 433     rules:
 434       - alert: renderd replication delay
 435         expr: renderd_replication_delay > 120
 436         for: 15m
 437         labels:
 438           alertgroup: tile
 439         annotations:
 440           delay: "{{ $value | humanizeDuration }}"
 441       - alert: missed tile rate
 442         expr: sum(rate(modtile_http_response_total{code="404"}[5m])) by (instance) / sum(rate(modtile_http_response_total[5m])) by (instance) > 0.05
 443         for: 5m
 444         labels:
 445           alertgroup: tile
 446         annotations:
 447           miss_rate: "{{ $value | humanizePercentage }}"
 448   - name: time
 449     rules:
 450       - alert: clock not synchronising
 451         expr: min_over_time(node_timex_sync_status[1m]) == 0 and node_timex_maxerror_seconds >= 16
 452         for: 5m
 453         labels:
 454           alertgroup: "{{ $labels.instance }}"
 455       - alert: clock skew detected
 456         expr: (node_timex_offset_seconds > 0.05 and deriv(node_timex_offset_seconds[5m]) >= 0) or (node_timex_offset_seconds < -0.05 and deriv(node_timex_offset_seconds[5m]) <= 0)
 457         for: 5m
 458         labels:
 459           alertgroup: "{{ $labels.instance }}"
 460         annotations:
 461           skew: "{{ with printf \"node_timex_offset_seconds{instance='%s'}\" $labels.instance | query }} {{ . | humanizeDuration }}{{ end }}"
 462   - name: web
 463     rules:
 464       - alert: web error rate
 465         expr: sum(rate(api_call_count_total{status=~"50[0-8]|5[1-9][0-9]"}[5m])) by (instance) / sum(rate(api_call_count_total[5m])) by (instance) > 0.002
 466         for: 5m
 467         labels:
 468           alertgroup: web
 469         annotations:
 470           error_rate: "{{ $value | humanizePercentage }}"
 471       - alert: job processing rate
 472         expr: rate(pg_stat_user_tables_n_tup_del{datname="openstreetmap",relname="delayed_jobs"}[5m]) / rate(pg_stat_user_tables_n_tup_ins{datname="openstreetmap",relname="delayed_jobs"}[5m]) < 0.9 and ignoring(job, name, datname, relname, schemaname, server) chef_role{name="db-master"} == 1
 473         for: 15m
 474         labels:
 475           alertgroup: web
 476         annotations:
 477           job_processing_rate: "{{ $value | humanizePercentage }}"