+ - name: memory
+ rules:
+ - alert: low memory
+ expr: node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes < 0.1
+ for: 15m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ memory_free: "{{ $value | humanizePercentage }}"
+ - alert: memory pressure
+ expr: rate(node_pressure_memory_waiting_seconds_total[5m]) > 0.6
+ for: 60m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ pressure: "{{ $value | humanizePercentage }}"
+ - alert: oom kill detected
+ expr: increase(node_vmstat_oom_kill[1m]) > 0
+ for: 0m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ new_oom_kills: "{{ $value }}"
+ - name: mysql
+ rules:
+ - alert: mysql down
+ expr: mysql_up == 0
+ for: 1m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: mysql connection limit
+ expr: mysql_global_status_max_used_connections / mysql_global_variables_max_connections > 0.8
+ for: 1m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ connections_used: "{{ $value | humanizePercentage }}"
+ - alert: mysql connection errors
+ expr: increase(mysql_global_status_connection_errors_total[1m]) > 0
+ for: 0m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ error_count: "{{ $value }}"
+ - name: network
+ rules:
+ - alert: interface redundancy lost
+ expr: node_bonding_active < 2 and on (instance, master) label_replace(chef_network_interface{bond_mode="802.3ad"}, "master", "$1", "name", "(.*)")
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ link_count: "{{ $value }}"
+ - alert: interface transmit rate
+ expr: rate(node_network_transmit_bytes_total[1m]) / node_network_speed_bytes > 0.99
+ 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 > 0.99
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ bandwidth_used: "{{ $value | humanizePercentage }}"
+ - alert: interface transmit errors
+ expr: rate(node_network_transmit_errs_total{device!~"wg.*"}[1m]) / rate(node_network_transmit_packets_total{device!~"wg.*"}[1m]) > 0.01
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ error_rate: "{{ $value | humanizePercentage }}"
+ - alert: wireguard interface transmit errors
+ expr: rate(node_network_transmit_errs_total{device=~"wg.*"}[1m]) / rate(node_network_transmit_packets_total{device=~"wg.*"}[1m]) > 0.05
+ for: 1h
+ 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: nominatim
+ rules:
+ - alert: nominatim replication delay
+ expr: nominatim_replication_delay > 10800
+ for: 1h
+ labels:
+ alertgroup: nominatim
+ annotations:
+ delay: "{{ $value | humanizeDuration }}"
+ - alert: nominatim connections
+ expr: sum(nginx_connections_writing and on (instance) chef_role{name="nominatim"}) > 2500
+ for: 15m
+ labels:
+ alertgroup: nominatim
+ - name: overpass
+ rules:
+ - alert: overpass osm database age
+ expr: overpass_database_age_seconds{database="osm"} > 3600
+ for: 1h
+ labels:
+ alertgroup: overpass
+ annotations:
+ age: "{{ $value | humanizeDuration }}"
+ - alert: overpass area database age
+ expr: overpass_database_age_seconds{database="area"} > 86400
+ for: 1h
+ labels:
+ alertgroup: overpass
+ annotations:
+ age: "{{ $value | humanizeDuration }}"
+ - name: passenger
+ rules:
+ - alert: passenger down
+ expr: passenger_up == 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: passenger queuing
+ expr: passenger_top_level_request_queue > 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: passenger application queuing
+ expr: passenger_app_request_queue > 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - name: planet
+ rules:
+ - alert: planet dump overdue
+ expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/(pbf|planet)/.*"} > 7 * 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
+ for: 24h
+ labels:
+ alertgroup: planet
+ annotations:
+ overdue_by: "{{ $value | humanizeDuration }}"
+ - alert: notes dump overdue
+ expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/notes/.*"} > 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
+ for: 6h
+ labels:
+ alertgroup: planet
+ annotations:
+ overdue_by: "{{ $value | humanizeDuration }}"
+ - alert: daily replication feed delayed
+ expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/day/.*"} > 86400 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
+ for: 3h
+ labels:
+ alertgroup: planet
+ annotations:
+ delayed_by: "{{ $value | humanizeDuration }}"
+ - alert: hourly replication feed delayed
+ expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/hour/.*"} > 3600 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
+ for: 30m
+ labels:
+ alertgroup: planet
+ annotations:
+ delayed_by: "{{ $value | humanizeDuration }}"
+ - alert: minutely replication feed delayed
+ expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/minute/.*"} > 60 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
+ for: 5m
+ labels:
+ alertgroup: planet
+ annotations:
+ delayed_by: "{{ $value | humanizeDuration }}"
+ - alert: changeset replication feed delayed
+ expr: time() - file_stat_modif_time_seconds{path=~"/store/planet/replication/changesets/.*"} > 60 and ignoring (job, name, path) chef_role{name="planetdump"} == 1
+ for: 5m
+ labels:
+ alertgroup: planet
+ annotations:
+ delayed_by: "{{ $value | humanizeDuration }}"
+ - 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 > 30
+ for: 15m
+ 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{datname!="nominatim"}[1m]) > 5
+ for: 0m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ new_deadlocks: "{{ $value }}"
+ - alert: postgresql idle transactions
+ expr: sum(pg_process_idle_seconds_count{state="idle in transaction"}) by (instance, server) > sum(pg_process_idle_seconds_bucket{state="idle in transaction",le="300"}) by (instance, server)
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ queries: "{{ $value }}"
+ - name: prometheus
+ rules:
+ - alert: prometheus configuration error
+ expr: prometheus_config_last_reload_successful == 0
+ for: 10m
+ labels:
+ alertgroup: "prometheus"
+ - alert: prometheus target missing
+ expr: up == 0
+ for: 10m
+ labels:
+ alertgroup: "prometheus"
+ - alert: node exporter text file scrape error
+ expr: node_textfile_scrape_error > 0
+ for: 10m
+ labels:
+ alertgroup: "prometheus"
+ - name: raid
+ rules:
+ - alert: raid controller battery failed
+ expr: ohai_controller_info{battery_status="failed"} > 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: raid controller battery recharging
+ expr: ohai_controller_info{battery_status="recharging"} > 0
+ for: 4h
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: raid array degraded
+ expr: ohai_array_info{status="degraded"} > 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: raid disk failed
+ expr: ohai_disk_info{status="failed"} > 0
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - name: rasdaemon
+ rules:
+ - alert: memory controller errors
+ expr: increase(rasdaemon_mc_events_total[1m]) > 0
+ for: 0m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ new_errors: "{{ $value }}"
+ - alert: pcie aer errors
+ expr: increase(rasdaemon_aer_events_total[1m]) > 0
+ for: 0m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ new_errors: "{{ $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 / 100 >= 0.8
+ for: 60m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ annotations:
+ percentage_used: "{{ $value | humanizePercentage }}"
+ - name: smokeping
+ rules:
+ - alert: packet loss
+ expr: 1 - (rate(smokeping_response_duration_seconds_count[5m]) / rate(smokeping_requests_total[5m])) > 0.02
+ for: 10m
+ labels:
+ alertgroup: smokeping
+ annotations:
+ loss_rate: "{{ $value | humanizePercentage }}"
+ - name: snmp
+ rules:
+ - alert: snmp pdus missing
+ expr: max_over_time(snmp_scrape_pdus_returned[1d]) - snmp_scrape_pdus_returned > 0
+ for: 15m
+ labels:
+ alertgroup: snmp
+ annotations:
+ missing_pdus: "{{ $value }}"
+ - 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: statuscake
+ rules:
+ - alert: statuscake uptime check failing
+ expr: statuscake_paused == 0 and statuscake_up == 0
+ for: 10m
+ labels:
+ alertgroup: statuscake
+ - name: systemd
+ rules:
+ - alert: systemd failed service
+ expr: node_systemd_unit_state{state="failed",name!="chef-client.service"} == 1
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: systemd failed chef client service
+ expr: sum_over_time(node_systemd_unit_state{state="inactive",name="chef-client.service"}[6h]) == 0
+ for: 0m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - name: taginfo
+ rules:
+ - alert: taginfo planet age
+ expr: time() - taginfo_data_from_seconds > 129600 and on (instance) chef_role{name="taginfo"}
+ for: 0m
+ labels:
+ alertgroup: taginfo
+ annotations:
+ age: "{{ $value | humanizeDuration }}"
+ - alert: taginfo database age
+ expr: time() - taginfo_database_update_finish_seconds > 129600 and on (instance) chef_role{name="taginfo"}
+ for: 0m
+ labels:
+ alertgroup: taginfo
+ annotations:
+ age: "{{ $value | humanizeDuration }}"
+ - alert: taginfo database size
+ expr: abs(delta(taginfo_database_size_bytes[30m])) / taginfo_database_size_bytes > 0.1
+ for: 30m
+ labels:
+ alertgroup: taginfo
+ annotations:
+ size_change: "{{ $value | humanizePercentage }}"
+ - name: tile
+ rules:
+ - alert: renderd replication delay
+ expr: renderd_replication_delay > 120
+ for: 15m
+ labels:
+ alertgroup: tile
+ annotations:
+ delay: "{{ $value | humanizeDuration }}"
+ - alert: missed tile rate
+ expr: sum(rate(modtile_http_response_total{code="404"}[5m])) by (instance) / sum(rate(modtile_http_response_total[5m])) by (instance) > 0.05
+ for: 5m
+ labels:
+ alertgroup: tile
+ annotations:
+ miss_rate: "{{ $value | humanizePercentage }}"
+ - alert: tile render rate
+ expr: sum(rate(renderd_zoom_metatiles_total[5m])) by (instance) == 0
+ for: 15m
+ labels:
+ alertgroup: tile
+ annotations:
+ render_rate: "{{ $value }} tiles/s"
+ - name: time
+ rules:
+ - alert: clock not synchronising
+ expr: min_over_time(node_timex_sync_status[1m]) == 0 and node_timex_maxerror_seconds >= 16
+ for: 5m
+ labels:
+ alertgroup: "{{ $labels.instance }}"
+ - alert: clock skew detected
+ 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)
+ for: 5m
+ labels:
+ 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=~"50[0-8]|5[1-9][0-9]"}[5m])) by (instance) / sum(rate(api_call_count_total[5m])) by (instance) > 0.002 and sum(rate(api_call_count_total{status=~"50[0-8]|5[1-9][0-9]"}[5m])) by (instance) > 0.01
+ for: 5m
+ labels:
+ alertgroup: web
+ annotations:
+ error_rate: "{{ $value | humanizePercentage }}"
+ - alert: job processing rate
+ expr: rate(pg_stat_user_tables_n_tup_del{datname="openstreetmap",relname="delayed_jobs"}[1h]) / rate(pg_stat_user_tables_n_tup_ins{datname="openstreetmap",relname="delayed_jobs"}[1h]) < 0.9 and ignoring(job, name, datname, relname, schemaname, server) chef_role{name="db-master"} == 1
+ for: 1h
+ labels:
+ alertgroup: web
+ annotations:
+ job_processing_rate: "{{ $value | humanizePercentage }}"
projects / chef.git / blobdiff