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[5m]) / rate(apache_accesses_total[1h] 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: cisco
  72     rules:
  73       - alert: cisco fan alarm
  74         expr: rlPhdUnitEnvParamFan1Status{rlPhdUnitEnvParamFan1Status!="normal"} > 0 or rlPhdUnitEnvParamFan2Status{rlPhdUnitEnvParamFan2Status!="normal"} > 0
  75         for: 5m
  76         labels:
  77           alertgroup: "{{ $labels.site }}"
  78       - alert: cisco temperature alarm
  79         expr: rlPhdUnitEnvParamTempSensorStatus{rlPhdUnitEnvParamTempSensorStatus!="ok"} > 0
  80         for: 5m
  81         labels:
  82           alertgroup: "{{ $labels.site }}"
  83       - alert: cisco main power alarm
  84         expr: rlPhdUnitEnvParamMainPSStatus{rlPhdUnitEnvParamMainPSStatus!="normal"} > 0
  85         for: 5m
  86         labels:
  87           alertgroup: "{{ $labels.site }}"
  88       - alert: cisco redundant power alarm
  89         expr: rlPhdUnitEnvParamRedundantPSStatus{rlPhdUnitEnvParamRedundantPSStatus!="normal"} > 0
  90         for: 5m
  91         labels:
  92           alertgroup: "{{ $labels.site }}"
  93   - name: cpu
  94     rules:
  95       - alert: cpu pressure
  96         expr: rate(node_pressure_cpu_waiting_seconds_total[5m]) > 0.6
  97         for: 15m
  98         labels:
  99           alertgroup: "{{ $labels.instance }}"
 100         annotations:
 101           pressure: "{{ $value | humanizePercentage }}"
 102   - name: database
 103     rules:
 104       - alert: postgres replication delay
 105         expr: pg_replication_lag_seconds > 5
 106         for: 5m
 107         labels:
 108           alertgroup: database
 109         annotations:
 110           delay: "{{ $value | humanizeDuration }}"
 111   - name: fastly
 112     rules:
 113       - alert: error rate
 114         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
 115         for: 15m
 116         labels:
 117           alertgroup: fastly
 118         annotations:
 119           error_rate: "{{ $value | humanizePercentage }}"
 120   - name: filesystem
 121     rules:
 122       - alert: readonly filesystem
 123         expr: node_filesystem_readonly == 1
 124         for: 0m
 125         labels:
 126           alertgroup: "{{ $labels.instance }}"
 127       - alert: filesystem low on space
 128         expr: node_filesystem_avail_bytes / node_filesystem_size_bytes < 0.05
 129         for: 5m
 130         labels:
 131           alertgroup: "{{ $labels.instance }}"
 132         annotations:
 133           percentage_free: "{{ $value | humanizePercentage }}"
 134           free_bytes: "{{ with printf \"node_filesystem_avail_bytes{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value | humanize1024 }}bytes{{end}}"
 135           total_total: "{{ with printf \"node_filesystem_size_bytes{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value | humanize1024 }}bytes{{end}}"
 136       - alert: filesystem low on inodes
 137         expr: node_filesystem_files_free / node_filesystem_files < 0.1
 138         for: 5m
 139         labels:
 140           alertgroup: "{{ $labels.instance }}"
 141         annotations:
 142           percentage_free: "{{ $value | humanizePercentage }}"
 143           free_inodes: "{{ with printf \"node_filesystem_files_free{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value }}{{end}}"
 144           total_inodes: "{{ with printf \"node_filesystem_files{instance='%s',mountpoint='%s'}\" $labels.instance $labels.mountpoint | query }}{{ . | first | value }}{{end}}"
 145   - name: hwmon
 146     rules:
 147       - alert: hwmon fan alarm
 148         expr: node_hwmon_fan_alarm == 1
 149         for: 5m
 150         labels:
 151           alertgroup: "{{ $labels.instance }}"
 152         annotations:
 153           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}}"
 154           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}}"
 155       - alert: hwmon temperature alarm
 156         expr: node_hwmon_temp_alarm == 1
 157         for: 5m
 158         labels:
 159           alertgroup: "{{ $labels.instance }}"
 160         annotations:
 161           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}}"
 162           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}}"
 163           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}}"
 164       - alert: hwmon voltage alarm
 165         expr: node_hwmon_in_alarm == 1
 166         for: 5m
 167         labels:
 168           alertgroup: "{{ $labels.instance }}"
 169         annotations:
 170           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}}"
 171           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}}"
 172           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}}"
 173   - name: io
 174     rules:
 175       - alert: io pressure
 176         expr: rate(node_pressure_io_waiting_seconds_total[5m]) > 0.6
 177         for: 60m
 178         labels:
 179           alertgroup: "{{ $labels.instance }}"
 180         annotations:
 181           pressure: "{{ $value | humanizePercentage }}"
 182   - name: ipmi
 183     rules:
 184       - alert: ipmi fan alarm
 185         expr: ipmi_fan_speed_state > 0
 186         for: 5m
 187         labels:
 188           alertgroup: "{{ $labels.instance }}"
 189         annotations:
 190           fan_speed_rpm: "{{ with printf \"ipmi_fan_speed_rpm{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}rpm{{end}}"
 191       - alert: ipmi temperature alarm
 192         expr: ipmi_temperature_state > 0
 193         for: 5m
 194         labels:
 195           alertgroup: "{{ $labels.instance }}"
 196         annotations:
 197           temperature_celsius: "{{ with printf \"ipmi_temperature_celsius{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}C{{end}}"
 198       - alert: ipmi voltage alarm
 199         expr: ipmi_voltage_state > 0
 200         for: 5m
 201         labels:
 202           alertgroup: "{{ $labels.instance }}"
 203         annotations:
 204           voltage_volts: "{{ with printf \"ipmi_voltage_volts{instance='%s',id='%s'}\" $labels.instance $labels.id | query }}{{ . | first | value | humanize }}V{{end}}"
 205       - alert: ipmi power alarm
 206         expr: ipmi_power_state > 0 or ipmi_sensor_state{type=~"Power .*"} > 0
 207         for: 5m
 208         labels:
 209           alertgroup: "{{ $labels.instance }}"
 210   - name: mail
 211     rules:
 212       - alert: exim queue length
 213         expr: exim_queue > exim_queue_limit
 214         for: 60m
 215         labels:
 216           alertgroup: mail
 217         annotations:
 218           queue_length: "{{ $value }}"
 219       - alert: mailman queue length
 220         expr: mailman_queue_length > 200
 221         for: 60m
 222         labels:
 223           alertgroup: mail
 224         annotations:
 225           queue_length: "{{ $value }}"
 226   - name: mdadm
 227     rules:
 228       - alert: mdadm array inactive
 229         expr: node_md_state{state="inactive"} > 0
 230         for: 0m
 231         labels:
 232           alertgroup: "{{ $labels.instance }}"
 233         annotations:
 234           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 235           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 236           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 237           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 238       - alert: mdadm array degraded
 239         expr: sum (node_md_disks{state="active"}) without (state) < node_md_disks_required
 240         for: 0m
 241         labels:
 242           alertgroup: "{{ $labels.instance }}"
 243         annotations:
 244           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 245           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 246           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 247           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 248       - alert: mdadm disk failed
 249         expr: node_md_disks{state="failed"} > 0
 250         for: 0m
 251         labels:
 252           alertgroup: "{{ $labels.instance }}"
 253         annotations:
 254           required: "{{ with printf \"node_md_disks_required{instance='%s',device='%s'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 255           active: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='active'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 256           failed: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='failed'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 257           spare: "{{ with printf \"node_md_disks{instance='%s',device='%s',state='spare'}\" $labels.instance $labels.device | query }}{{ . | first | value | humanize }} disks{{end}}"
 258   - name: memory
 259     rules:
 260       - alert: low memory
 261         expr: node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes < 0.1
 262         for: 15m
 263         labels:
 264           alertgroup: "{{ $labels.instance }}"
 265         annotations:
 266           memory_free: "{{ $value | humanizePercentage }}"
 267       - alert: memory pressure
 268         expr: rate(node_pressure_memory_waiting_seconds_total[5m]) > 0.6
 269         for: 60m
 270         labels:
 271           alertgroup: "{{ $labels.instance }}"
 272         annotations:
 273           pressure: "{{ $value | humanizePercentage }}"
 274       - alert: oom kill detected
 275         expr: increase(node_vmstat_oom_kill[1m]) > 0
 276         for: 0m
 277         labels:
 278           alertgroup: "{{ $labels.instance }}"
 279         annotations:
 280           new_oom_kills: "{{ $value }}"
 281   - name: network
 282     rules:
 283       - alert: interface transmit rate
 284         expr: rate(node_network_transmit_bytes_total[1m]) / node_network_speed_bytes > 0.98
 285         for: 5m
 286         labels:
 287           alertgroup: "{{ $labels.instance }}"
 288         annotations:
 289           bandwidth_used: "{{ $value | humanizePercentage }}"
 290       - alert: interface receive rate
 291         expr: rate(node_network_receive_bytes_total[1m]) / node_network_speed_bytes > 0.98
 292         for: 5m
 293         labels:
 294           alertgroup: "{{ $labels.instance }}"
 295         annotations:
 296           bandwidth_used: "{{ $value | humanizePercentage }}"
 297       - alert: interface transmit errors
 298         expr: rate(node_network_transmit_errs_total[1m]) / rate(node_network_transmit_packets_total[1m]) > 0.01
 299         for: 5m
 300         labels:
 301           alertgroup: "{{ $labels.instance }}"
 302         annotations:
 303           error_rate: "{{ $value | humanizePercentage }}"
 304       - alert: interface receive errors
 305         expr: rate(node_network_receive_errs_total[1m]) / rate(node_network_receive_packets_total[1m]) > 0.01
 306         for: 5m
 307         labels:
 308           alertgroup: "{{ $labels.instance }}"
 309         annotations:
 310           error_rate: "{{ $value | humanizePercentage }}"
 311       - alert: conntrack entries
 312         expr: node_nf_conntrack_entries / node_nf_conntrack_entries_limit > 0.8
 313         for: 5m
 314         labels:
 315           alertgroup: "{{ $labels.instance }}"
 316         annotations:
 317           entries_used: "{{ $value | humanizePercentage }}"
 318   - name: planet
 319     rules:
 320       - alert: planet dump overdue
 321         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/(pbf|planet)/.*"} > 7 * 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 322         for: 24h
 323         labels:
 324           alertgroup: planet
 325         annotations:
 326           overdue_by: "{{ $value | humanizeDuration }}"
 327       - alert: notes dump overdue
 328         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/notes/.*"} > 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 329         for: 6h
 330         labels:
 331           alertgroup: planet
 332         annotations:
 333           overdue_by: "{{ $value | humanizeDuration }}"
 334       - alert: daily replication feed delayed
 335         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/day/.*"} > 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 336         for: 3h
 337         labels:
 338           alertgroup: planet
 339         annotations:
 340           delayed_by: "{{ $value | humanizeDuration }}"
 341       - alert: hourly replication feed delayed
 342         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/hour/.*"} > 3600 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 343         for: 30m
 344         labels:
 345           alertgroup: planet
 346         annotations:
 347           delayed_by: "{{ $value | humanizeDuration }}"
 348       - alert: minutely replication feed delayed
 349         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/minute/.*"} > 60 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 350         for: 5m
 351         labels:
 352           alertgroup: planet
 353         annotations:
 354           delayed_by: "{{ $value | humanizeDuration }}"
 355       - alert: changeset replication feed delayed
 356         expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/changesets/.*"} > 60 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
 357         for: 5m
 358         labels:
 359           alertgroup: planet
 360         annotations:
 361           delayed_by: "{{ $value | humanizeDuration }}"
 362   - name: postgresql
 363     rules:
 364       - alert: postgresql down
 365         expr: pg_up == 0
 366         for: 1m
 367         labels:
 368           alertgroup: "{{ $labels.instance }}"
 369       - alert: postgresql replication delay
 370         expr: pg_replication_lag_seconds > 5
 371         for: 1m
 372         labels:
 373           alertgroup: "{{ $labels.instance }}"
 374         annotations:
 375           delay: "{{ $value | humanizeDuration }}"
 376       - alert: postgresql connection limit
 377         expr: sum (pg_stat_activity_count) by (instance, server) / sum (pg_settings_max_connections) by (instance, server) > 0.8
 378         for: 1m
 379         labels:
 380           alertgroup: "{{ $labels.instance }}"
 381         annotations:
 382           connections_used: "{{ $value | humanizePercentage }}"
 383       - alert: postgresql deadlocks
 384         expr: increase(pg_stat_database_deadlocks[1m]) > 5
 385         for: 0m
 386         labels:
 387           alertgroup: "{{ $labels.instance }}"
 388         annotations:
 389           new_deadlocks: "{{ $value }}"
 390       - alert: postgresql slow queries
 391         expr: pg_slow_queries > 0
 392         for: 5m
 393         labels:
 394           alertgroup: "{{ $labels.instance }}"
 395         annotations:
 396           queries: "{{ $value }}"
 397   - name: smart
 398     rules:
 399       - alert: smart failure
 400         expr: smart_health_status == 0
 401         for: 60m
 402         labels:
 403           alertgroup: "{{ $labels.instance }}"
 404       - alert: smart ssd wearout approaching
 405         expr: smart_percentage_used >= 90
 406         for: 60m
 407         labels:
 408           alertgroup: "{{ $labels.instance }}"
 409         annotations:
 410           percentage_used: "{{ $value | humanizePercentage }}"
 411   - name: ssl
 412     rules:
 413       - alert: ssl certificate probe failed
 414         expr: ssl_probe_success == 0
 415         for: 60m
 416         labels:
 417           alertgroup: ssl
 418       - alert: ssl certificate expiry
 419         expr: ssl_verified_cert_not_after{chain_no="0"} - time() < 86400 * 14
 420         for: 0m
 421         labels:
 422           alertgroup: ssl
 423         annotations:
 424           expires_in: "{{ $value | humanizeDuration }}"
 425       - alert: ssl certificate revoked
 426         expr: ssl_ocsp_response_status == 1
 427         for: 0m
 428         labels:
 429           alertgroup: ssl
 430       - alert: ocsp status unknown
 431         expr: ssl_ocsp_response_status == 1
 432         for: 0m
 433         labels:
 434           alertgroup: ssl
 435   - name: statuscake
 436     rules:
 437       - alert: statuscake uptime check failing
 438         expr: statuscake_uptime{status="down",paused="false"} > 0
 439         for: 0m
 440         labels:
 441           alertgroup: statuscake
 442   - name: systemd
 443     rules:
 444       - alert: systemd failed service
 445         expr: node_systemd_unit_state{state="failed",name!="chef-client.service"} == 1
 446         for: 5m
 447         labels:
 448           alertgroup: "{{ $labels.instance }}"
 449       - alert: systemd failed service
 450         expr: node_systemd_unit_state{state="failed",name="chef-client.service"} == 1
 451         for: 6h
 452         labels:
 453           alertgroup: "{{ $labels.instance }}"
 454   - name: tile
 455     rules:
 456       - alert: renderd replication delay
 457         expr: renderd_replication_delay > 120
 458         for: 15m
 459         labels:
 460           alertgroup: tile
 461         annotations:
 462           delay: "{{ $value | humanizeDuration }}"
 463       - alert: missed tile rate
 464         expr: sum(rate(modtile_http_response_total{code="404"}[5m])) by (instance) / sum(rate(modtile_http_response_total[5m])) by (instance) > 0.05
 465         for: 5m
 466         labels:
 467           alertgroup: tile
 468         annotations:
 469           miss_rate: "{{ $value | humanizePercentage }}"
 470   - name: time
 471     rules:
 472       - alert: clock not synchronising
 473         expr: min_over_time(node_timex_sync_status[1m]) == 0 and node_timex_maxerror_seconds >= 16
 474         for: 5m
 475         labels:
 476           alertgroup: "{{ $labels.instance }}"
 477       - alert: clock skew detected
 478         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)
 479         for: 5m
 480         labels:
 481           alertgroup: "{{ $labels.instance }}"
 482         annotations:
 483           skew: "{{ with printf \"node_timex_offset_seconds{instance='%s'}\" $labels.instance | query }} {{ . | humanizeDuration }}{{ end }}"
 484   - name: web
 485     rules:
 486       - alert: web error rate
 487         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
 488         for: 5m
 489         labels:
 490           alertgroup: web
 491         annotations:
 492           error_rate: "{{ $value | humanizePercentage }}"
 493       - alert: job processing rate
 494         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
 495         for: 15m
 496         labels:
 497           alertgroup: web
 498         annotations:
 499           job_processing_rate: "{{ $value | humanizePercentage }}"