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: apache
  12     rules:
  13       - alert: apache down
  14         expr: apache_up == 0
  15         for: 5m
  16         labels:
  17           alertgroup: "{{ $labels.instance }}"
  18       - alert: apache workers busy
  19         expr: sum(apache_workers{state="busy"}) by (instance) / sum(apache_scoreboard) by (instance) > 0.8
  20         for: 5m
  21         labels:
  22           alertgroup: "{{ $labels.instance }}"
  23         annotations:
  24           busy_workers: "{{ $value | humanizePercentage }}"
  25   - name: cpu
  26     rules:
  27       - alert: cpu pressure
  28         expr: rate(node_pressure_cpu_waiting_seconds_total[5m]) > 0.3
  29         for: 15m
  30         labels:
  31           alertgroup: "{{ $labels.instance }}"
  32         annotations:
  33           pressure: "{{ $value | humanizePercentage }}"
  34   - name: database
  35     rules:
  36       - alert: postgres replication delay
  37         expr: pg_replication_lag_seconds > 5
  38         for: 5m
  39         labels:
  40           alertgroup: database
  41         annotations:
  42           delay: "{{ $value | humanizeDuration }}"
  43   - name: fastly
  44     rules:
  45       - alert: error rate
  46         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
  47         for: 15m
  48         labels:
  49           alertgroup: fastly
  50         annotations:
  51           error_rate: "{{ $value | humanizePercentage }}"
  52   - name: filesystem
  53     rules:
  54       - alert: readonly filesystem
  55         expr: node_filesystem_readonly == 1
  56         for: 0m
  57         labels:
  58           alertgroup: "{{ $labels.instance }}"
  59       - alert: filesystem low on space
  60         expr: node_filesystem_avail_bytes / node_filesystem_size_bytes < 0.05
  61         for: 5m
  62         labels:
  63           alertgroup: "{{ $labels.instance }}"
  64         annotations:
  65           percentage_free: "{{ $value | humanizePercentage }}"
  66           free_bytes: "{{ with printf \"node_filesystem_avail_bytes{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value | humanize1024 }}bytes{{end}}"
  67           total_total: "{{ with printf \"node_filesystem_size_bytes{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value | humanize1024 }}bytes{{end}}"
  68       - alert: filesystem low on inodes
  69         expr: node_filesystem_files_free / node_filesystem_files < 0.1
  70         for: 5m
  71         labels:
  72           alertgroup: "{{ $labels.instance }}"
  73         annotations:
  74           percentage_free: "{{ $value | humanizePercentage }}"
  75           free_inodes: "{{ with printf \"node_filesystem_files_free{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value }}{{end}}"
  76           total_inodes: "{{ with printf \"node_filesystem_files{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value }}{{end}}"
  77   - name: hwmon
  78     rules:
  79       - alert: hwmon fan alarm
  80         expr: node_hwmon_fan_alarm == 1
  81         for: 5m
  82         labels:
  83           alertgroup: "{{ $labels.instance }}"
  84         annotations:
  85           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}}"
  86           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}}"
  87       - alert: hwmon temperature alarm
  88         expr: node_hwmon_temp_alarm == 1
  89         for: 5m
  90         labels:
  91           alertgroup: "{{ $labels.instance }}"
  92         annotations:
  93           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}}"
  94           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}}"
  95           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}}"
  96       - alert: hwmon voltage alarm
  97         expr: node_hwmon_in_alarm == 1
  98         for: 5m
  99         labels:
 100           alertgroup: "{{ $labels.instance }}"
 101         annotations:
 102           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}}"
 103           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}}"
 104           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}}"
 105   - name: io
 106     rules:
 107       - alert: io pressure
 108         expr: rate(node_pressure_io_waiting_seconds_total[5m]) > 0.6
 109         for: 60m
 110         labels:
 111           alertgroup: "{{ $labels.instance }}"
 112         annotations:
 113           pressure: "{{ $value | humanizePercentage }}"
 114   - name: ipmi
 115     rules:
 116       - alert: ipmi fan alarm
 117         expr: ipmi_fan_speed_state > 0
 118         for: 5m
 119         labels:
 120           alertgroup: "{{ $labels.instance }}"
 121         annotations:
 122           fan_speed_rpm: "{{ with printf \"ipmi_fan_speed_rpm{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}rpm{{end}}"
 123       - alert: ipmi temperature alarm
 124         expr: ipmi_temperature_state > 0
 125         for: 5m
 126         labels:
 127           alertgroup: "{{ $labels.instance }}"
 128         annotations:
 129           temperature_celsius: "{{ with printf \"ipmi_temperature_celsius{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}C{{end}}"
 130       - alert: ipmi voltage alarm
 131         expr: ipmi_voltage_state > 0
 132         for: 5m
 133         labels:
 134           alertgroup: "{{ $labels.instance }}"
 135         annotations:
 136           voltage_volts: "{{ with printf \"ipmi_voltage_volts{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}V{{end}}"
 137       - alert: ipmi power alarm
 138         expr: ipmi_power_state > 0 or ipmi_sensor_state{type=~"Power .*"} > 0
 139         for: 5m
 140         labels:
 141           alertgroup: "{{ $labels.instance }}"
 142   - name: mail
 143     rules:
 144       - alert: exim queue length
 145         expr: exim_queue > exim_queue_limit
 146         for: 60m
 147         labels:
 148           alertgroup: mail
 149         annotations:
 150           queue_length: "{{ $value }}"
 151       - alert: mailman queue length
 152         expr: mailman_queue_length > 200
 153         for: 60m
 154         labels:
 155           alertgroup: mail
 156         annotations:
 157           queue_length: "{{ $value }}"
 158   - name: mdadm
 159     rules:
 160       - alert: mdadm array inactive
 161         expr: node_md_state{state="inactive"} > 0
 162         for: 0m
 163         labels:
 164           alertgroup: "{{ $labels.instance }}"
 165         annotations:
 166           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 167           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 168           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 169           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 170       - alert: mdadm array degraded
 171         expr: sum (node_md_disks{state="active"}) without (state) < node_md_disks_required
 172         for: 0m
 173         labels:
 174           alertgroup: "{{ $labels.instance }}"
 175         annotations:
 176           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 177           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 178           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 179           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 180       - alert: mdadm disk failed
 181         expr: node_md_disks{state="failed"} > 0
 182         for: 0m
 183         labels:
 184           alertgroup: "{{ $labels.instance }}"
 185         annotations:
 186           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 187           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 188           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 189           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 190   - name: memory
 191     rules:
 192       - alert: low memory
 193         expr: node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes < 0.1
 194         for: 15m
 195         labels:
 196           alertgroup: "{{ $labels.instance }}"
 197         annotations:
 198           memory_free: "{{ $value | humanizePercentage }}"
 199       - alert: memory pressure
 200         expr: rate(node_pressure_memory_waiting_seconds_total[5m]) > 0.6
 201         for: 60m
 202         labels:
 203           alertgroup: "{{ $labels.instance }}"
 204         annotations:
 205           pressure: "{{ $value | humanizePercentage }}"
 206       - alert: oom kill detected
 207         expr: increase(node_vmstat_oom_kill[1m]) > 0
 208         for: 0m
 209         labels:
 210           alertgroup: "{{ $labels.instance }}"
 211         annotations:
 212           new_oom_kills: "{{ $value }}"
 213   - name: network
 214     rules:
 215       - alert: interface transmit rate
 216         expr: rate(node_network_transmit_bytes_total[1m]) / node_network_speed_bytes > 0.98
 217         for: 5m
 218         labels:
 219           alertgroup: "{{ $labels.instance }}"
 220         annotations:
 221           bandwidth_used: "{{ $value | humanizePercentage }}"
 222       - alert: interface receive rate
 223         expr: rate(node_network_receive_bytes_total[1m]) / node_network_speed_bytes > 0.98
 224         for: 5m
 225         labels:
 226           alertgroup: "{{ $labels.instance }}"
 227         annotations:
 228           bandwidth_used: "{{ $value | humanizePercentage }}"
 229       - alert: interface transmit errors
 230         expr: rate(node_network_transmit_errs_total[1m]) / rate(node_network_transmit_packets_total[1m]) > 0.01
 231         for: 5m
 232         labels:
 233           alertgroup: "{{ $labels.instance }}"
 234         annotations:
 235           error_rate: "{{ $value | humanizePercentage }}"
 236       - alert: interface receive errors
 237         expr: rate(node_network_receive_errs_total[1m]) / rate(node_network_receive_packets_total[1m]) > 0.01
 238         for: 5m
 239         labels:
 240           alertgroup: "{{ $labels.instance }}"
 241         annotations:
 242           error_rate: "{{ $value | humanizePercentage }}"
 243       - alert: conntrack entries
 244         expr: node_nf_conntrack_entries / node_nf_conntrack_entries_limit > 0.8
 245         for: 5m
 246         labels:
 247           alertgroup: "{{ $labels.instance }}"
 248         annotations:
 249           entries_used: "{{ $value | humanizePercentage }}"
 250   - name: planet
 251     rules:
 252       - alert: planet dump overdue
 253         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/(pbf|planet)/.*"} > 7 * 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 254         for: 24h
 255         labels:
 256           alertgroup: planet
 257         annotations:
 258           overdue_by: "{{ $value | humanizeDuration }}"
 259       - alert: notes dump overdue
 260         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/notes/.*"} > 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 261         for: 6h
 262         labels:
 263           alertgroup: planet
 264         annotations:
 265           overdue_by: "{{ $value | humanizeDuration }}"
 266       - alert: daily replication feed delayed
 267         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/day/.*"} > 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 268         for: 3h
 269         labels:
 270           alertgroup: planet
 271         annotations:
 272           delayed_by: "{{ $value | humanizeDuration }}"
 273       - alert: hourly replication feed delayed
 274         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/hour/.*"} > 3600 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 275         for: 30m
 276         labels:
 277           alertgroup: planet
 278         annotations:
 279           delayed_by: "{{ $value | humanizeDuration }}"
 280       - alert: minutely replication feed delayed
 281         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/minute/.*"} > 60 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 282         for: 5m
 283         labels:
 284           alertgroup: planet
 285         annotations:
 286           delayed_by: "{{ $value | humanizeDuration }}"
 287       - alert: changeset replication feed delayed
 288         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/changesets/.*"} > 60 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 289         for: 5m
 290         labels:
 291           alertgroup: planet
 292         annotations:
 293           delayed_by: "{{ $value | humanizeDuration }}"
 294   - name: postgresql
 295     rules:
 296       - alert: postgresql down
 297         expr: pg_up == 0
 298         for: 1m
 299         labels:
 300           alertgroup: "{{ $labels.instance }}"
 301       - alert: postgresql replication delay
 302         expr: pg_replication_lag_seconds > 5
 303         for: 1m
 304         labels:
 305           alertgroup: "{{ $labels.instance }}"
 306         annotations:
 307           delay: "{{ $value | humanizeDuration }}"
 308       - alert: postgresql connection limit
 309         expr: sum (pg_stat_activity_count) by (instance, server) / sum (pg_settings_max_connections) by (instance, server) > 0.8
 310         for: 1m
 311         labels:
 312           alertgroup: "{{ $labels.instance }}"
 313         annotations:
 314           connections_used: "{{ $value | humanizePercentage }}"
 315       - alert: postgresql deadlocks
 316         expr: increase(pg_stat_database_deadlocks[1m]) > 5
 317         for: 0m
 318         labels:
 319           alertgroup: "{{ $labels.instance }}"
 320         annotations:
 321           new_deadlocks: "{{ $value }}"
 322       - alert: postgresql slow queries
 323         expr: pg_slow_queries > 0
 324         for: 5m
 325         labels:
 326           alertgroup: "{{ $labels.instance }}"
 327         annotations:
 328           queries: "{{ $value }}"
 329   - name: smart
 330     rules:
 331       - alert: smart failure
 332         expr: smart_health_status == 0
 333         for: 60m
 334         labels:
 335           alertgroup: "{{ $labels.instance }}"
 336       - alert: smart ssd wearout approaching
 337         expr: smart_percentage_used >= 90
 338         for: 60m
 339         labels:
 340           alertgroup: "{{ $labels.instance }}"
 341         annotations:
 342           percentage_used: "{{ $value | humanizePercentage }}"
 343   - name: ssl
 344     rules:
 345       - alert: ssl certificate probe failed
 346         expr: ssl_probe_success == 0
 347         for: 60m
 348         labels:
 349           alertgroup: ssl
 350       - alert: ssl certificate expiry
 351         expr: ssl_verified_cert_not_after{chain_no="0"} - time() < 86400 * 14
 352         for: 0m
 353         labels:
 354           alertgroup: ssl
 355         annotations:
 356           expires_in: "{{ $value | humanizeDuration }}"
 357       - alert: ssl certificate revoked
 358         expr: ssl_ocsp_response_status == 1
 359         for: 0m
 360         labels:
 361           alertgroup: ssl
 362       - alert: ocsp status unknown
 363         expr: ssl_ocsp_response_status == 1
 364         for: 0m
 365         labels:
 366           alertgroup: ssl
 367   - name: systemd
 368     rules:
 369       - alert: systemd failed service
 370         expr: node_systemd_unit_state{state="failed"} == 1
 371         for: 5m
 372         labels:
 373           alertgroup: "{{ $labels.instance }}"
 374   - name: tile
 375     rules:
 376       - alert: renderd replication delay
 377         expr: renderd_replication_delay > 120
 378         for: 5m
 379         labels:
 380           alertgroup: tile
 381         annotations:
 382           delay: "{{ $value | humanizeDuration }}"
 383       - alert: missed tile rate
 384         expr: sum(rate(modtile_http_response_total{code="404"}[5m])) by (instance) / sum(rate(modtile_http_response_total[5m])) by (instance) > 0.05
 385         for: 5m
 386         labels:
 387           alertgroup: tile
 388         annotations:
 389           miss_rate: "{{ $value | humanizePercentage }}"
 390   - name: time
 391     rules:
 392       - alert: clock not synchronising
 393         expr: min_over_time(node_timex_sync_status[1m]) == 0 and node_timex_maxerror_seconds >= 16
 394         for: 5m
 395         labels:
 396           alertgroup: "{{ $labels.instance }}"
 397       - alert: clock skew detected
 398         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)
 399         for: 5m
 400         labels:
 401           alertgroup: "{{ $labels.instance }}"
 402         annotations:
 403           skew: "{{ with printf \"node_timex_offset_seconds{instance='%s'}\" $labels.instance | query }} {{ . | humanizeDuration }}{{ end }}"
 404   - name: web
 405     rules:
 406       - alert: web error rate
 407         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
 408         for: 5m
 409         labels:
 410           alertgroup: web
 411         annotations:
 412           error_rate: "{{ $value | humanizePercentage }}"
 413       - alert: job processing rate
 414         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
 415         for: 5m
 416         labels:
 417           alertgroup: web
 418         annotations:
 419           job_processing_rate: "{{ $value | humanizePercentage }}"