Mercurial > code > home > repos > victoriametrics
annotate alert_rules.py @ 36:2bc188c4117a
rules updates, incl front door group
| author | drewp@bigasterisk.com |
|---|---|
| date | Sun, 15 Oct 2023 18:34:29 -0700 |
| parents | 3b91d52b007d |
| children | 6e27d280b598 |
| rev | line source |
|---|---|
| 23 | 1 """ |
| 2 pdm run invoke push-config | |
| 3 | |
| 4 docs: https://prometheus.io/docs/prometheus/latest/configuration/alerting_rules/ | |
| 5 "Whenever the alert expression results in one or more vector | |
| 6 elements at a given point in time, the alert counts as active for | |
| 7 these elements' label sets." | |
| 8 also https://www.metricfire.com/blog/top-5-prometheus-alertmanager-gotchas/#Missing-metrics | |
| 9 | |
| 10 """ | |
| 11 | |
| 12 import json | |
| 13 | |
| 14 | |
| 15 def k8sRules(): | |
| 16 # from https://awesome-prometheus-alerts.grep.to/rules.html | |
| 17 return [ | |
| 18 { | |
| 34 | 19 "alert": "metricsTargetMissing", |
| 20 "expr": 'up{job!~"cm-acme-.*"} == 0', | |
| 36 | 21 'for': '10m', |
| 31 | 22 "labels": { |
| 23 "severity": "critical" | |
| 24 }, | |
| 23 | 25 "annotations": { |
| 34 | 26 "summary": "metrics target missing (instance {{ $labels.instance }})", |
| 27 "description": "A metrics target has disappeared. An exporter might be crashed.\n VALUE = {{ $value }}", | |
| 23 | 28 }, |
| 29 }, | |
| 30 { | |
| 31 "alert": "KubernetesMemoryPressure", | |
| 32 "expr": 'kube_node_status_condition{condition="MemoryPressure",status="true"} == 1', | |
| 33 "for": "2m", | |
| 31 | 34 "labels": { |
| 35 "severity": "critical" | |
| 36 }, | |
| 23 | 37 "annotations": { |
| 38 "summary": "Kubernetes memory pressure (instance {{ $labels.instance }})", | |
| 39 "description": "{{ $labels.node }} has MemoryPressure condition\n VALUE = {{ $value }}", | |
| 40 }, | |
| 41 }, | |
| 42 { | |
| 43 "alert": "KubernetesDiskPressure", | |
| 44 "expr": 'kube_node_status_condition{condition="DiskPressure",status="true"} == 1', | |
| 45 "for": "2m", | |
| 31 | 46 "labels": { |
| 47 "severity": "critical" | |
| 48 }, | |
| 23 | 49 "annotations": { |
| 50 "summary": "Kubernetes disk pressure (instance {{ $labels.instance }})", | |
| 51 "description": "{{ $labels.node }} has DiskPressure condition\n VALUE = {{ $value }}", | |
| 52 }, | |
| 53 }, | |
| 54 { | |
| 55 "alert": "KubernetesOutOfDisk", | |
| 56 "expr": 'kube_node_status_condition{condition="OutOfDisk",status="true"} == 1', | |
| 57 "for": "2m", | |
| 31 | 58 "labels": { |
| 59 "severity": "critical" | |
| 60 }, | |
| 23 | 61 "annotations": { |
| 62 "summary": "Kubernetes out of disk (instance {{ $labels.instance }})", | |
| 63 "description": "{{ $labels.node }} has OutOfDisk condition\n VALUE = {{ $value }}", | |
| 64 }, | |
| 65 }, | |
| 66 { | |
| 67 "alert": "KubernetesJobFailed", | |
| 68 "expr": "kube_job_status_failed > 0", | |
| 31 | 69 "labels": { |
| 70 "severity": "warning" | |
| 71 }, | |
| 23 | 72 "annotations": { |
| 73 "summary": "Kubernetes Job failed (instance {{ $labels.instance }})", | |
| 74 "description": "Job {{$labels.namespace}}/{{$labels.exported_job}} failed to complete\n VALUE = {{ $value }}", | |
| 75 }, | |
| 76 }, | |
| 77 { | |
| 78 "alert": "KubernetesPodCrashLooping", | |
| 79 "expr": "increase(kube_pod_container_status_restarts_total[1m]) > 3", | |
| 80 "for": "2m", | |
| 31 | 81 "labels": { |
| 82 "severity": "warning" | |
| 83 }, | |
| 23 | 84 "annotations": { |
| 85 "summary": "Kubernetes pod crash looping (instance {{ $labels.instance }})", | |
| 86 "description": "Pod {{ $labels.pod }} is crash looping\n VALUE = {{ $value }}", | |
| 87 }, | |
| 88 }, | |
| 89 { | |
| 31 | 90 "alert": |
| 91 "KubernetesClientCertificateExpiresNextWeek", | |
| 92 "expr": | |
| 93 'apiserver_client_certificate_expiration_seconds_count{job="apiserver"} > 0 and histogram_quantile(0.01, sum by (job, le) (rate(apiserver_client_certificate_expiration_seconds_bucket{job="apiserver"}[5m]))) < 7*24*60*60', | |
| 94 "labels": { | |
| 95 "severity": "warning" | |
| 96 }, | |
| 23 | 97 "annotations": { |
| 98 "summary": "Kubernetes client certificate expires next week (instance {{ $labels.instance }})", | |
| 99 "description": "A client certificate used to authenticate to the apiserver is expiring next week.\n VALUE = {{ $value }}", | |
| 100 }, | |
| 101 }, | |
| 102 { | |
| 103 "alert": "container_waiting", | |
| 34 | 104 "expr": "sum by (namespace, pod, container)(kube_pod_container_status_waiting!=0)", |
| 105 "annotations": { | |
| 106 "description": '', | |
| 107 "dashboard": "https://bigasterisk.com/k/clusters/local/namespaces/{{ $labels.namespace }}/pods/{{ $labels.pod }}", | |
| 108 }, | |
| 23 | 109 "for": "2m", |
| 110 }, | |
| 111 ] | |
| 112 | |
| 113 | |
|
32
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
114 def allRules(ctx): |
| 23 | 115 return { |
| 116 "groups": [ | |
| 117 { | |
| 118 "name": "k8s", | |
|
28
e114edff93dc
more explicit intervals. try to get a single day of notification out of a disk err increase
drewp@bigasterisk.com
parents:
27
diff
changeset
|
119 "interval": "1m", |
| 23 | 120 "rules": k8sRules(), |
| 121 }, | |
|
32
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
122 expectedK8sNodesGroup(ctx), |
| 23 | 123 # |
| 124 # any presence of starlette_request_duration_seconds_created{app_name="starlette",method="GET",path="/",status_code="200"} 1.6460176156784086e+09 means someone forgot to set app name | |
| 125 { | |
| 31 | 126 "name": |
| 127 "Outages", | |
| 128 "interval": | |
| 129 "1m", | |
| 23 | 130 "rules": [ |
| 131 { | |
| 132 "alert": "powereagleStalled", | |
| 133 "expr": "rate(house_power_w[100m]) == 0", | |
| 134 "for": "0m", | |
| 31 | 135 "labels": { |
| 136 "severity": "losingData" | |
| 137 }, | |
| 23 | 138 "annotations": { |
| 139 "summary": "power eagle data stalled", | |
| 140 "description": "logs at https://bigasterisk.com/k/clusters/local/namespaces/default/deployments/power-eagle/logs", | |
| 141 }, | |
| 142 }, | |
| 143 { | |
| 144 "alert": "powereagleAbsent", | |
| 145 "expr": "absent_over_time(house_power_w[5m])", | |
| 146 "for": "2m", | |
| 31 | 147 "labels": { |
| 148 "severity": "losingData" | |
| 149 }, | |
| 23 | 150 "annotations": { |
| 151 "summary": "power eagle data missing", | |
| 152 "description": "logs at https://bigasterisk.com/k/clusters/local/namespaces/default/deployments/power-eagle/logs", | |
| 153 }, | |
| 154 }, | |
| 155 { | |
| 156 "alert": "absent_zigbee", | |
| 157 "expr": 'absent(container_last_seen{container="zigbee2mqtt"})', | |
| 158 }, | |
| 159 { | |
| 160 "alert": "net_routes_sync", | |
| 161 "expr": 'rate(starlette_request_duration_seconds_count{app_name="net_routes",path="/routes"}[5m]) < 1/70', | |
| 162 "for": "10m", | |
| 31 | 163 "labels": { |
| 164 "severity": "houseUsersAffected" | |
| 165 }, | |
| 23 | 166 "annotations": { |
| 167 "summary": "net_routes is not getting regular updates" | |
| 168 }, | |
| 169 }, | |
| 170 ], | |
| 171 }, | |
| 172 { | |
| 36 | 173 "name": |
| 174 "disk_errs", | |
| 175 "interval": | |
| 176 "2d", | |
| 31 | 177 "rules": [{ |
| 36 | 178 "alert": "zpool_device_error_increase", |
| 179 "labels": { | |
| 180 "severity": "warning" | |
| 181 }, | |
| 182 "expr": 'increase(zpool_device_error_count[3d]) > 0', | |
| 183 }, { | |
| 31 | 184 "alert": "zpool_device_error_count", |
| 185 "labels": { | |
| 186 "severity": "warning" | |
|
28
e114edff93dc
more explicit intervals. try to get a single day of notification out of a disk err increase
drewp@bigasterisk.com
parents:
27
diff
changeset
|
187 }, |
| 36 | 188 "expr": 'zpool_device_error_count > 0', |
| 31 | 189 }], |
|
28
e114edff93dc
more explicit intervals. try to get a single day of notification out of a disk err increase
drewp@bigasterisk.com
parents:
27
diff
changeset
|
190 }, |
|
e114edff93dc
more explicit intervals. try to get a single day of notification out of a disk err increase
drewp@bigasterisk.com
parents:
27
diff
changeset
|
191 { |
| 31 | 192 "name": |
| 34 | 193 "front_door", |
| 194 "interval": | |
| 195 "5m", | |
| 196 "rules": [ | |
| 197 { | |
| 36 | 198 "alert": "front_door_reader_esp32_no_mqtt", |
| 199 'expr': 'hw_connected{job="fingerprint"} < 1', | |
| 200 "annotations": { | |
| 201 "summary": "see https://bigasterisk.com/front-door-lock/" | |
| 202 }, | |
| 203 }, | |
| 204 { | |
| 205 "alert": "fronr_door_reader_svc_down", | |
| 206 'expr': 'up{job="fingerprint"} < 1', | |
| 207 "annotations": { | |
| 208 "summary": "see https://bigasterisk.com/front-door-lock/" | |
| 209 }, | |
| 210 }, | |
| 211 { | |
| 212 "alert": "fronr_door_reader_svc_reader_no_mqtt", | |
| 213 'expr': 'mqtt_connected{job="fingerprint"} < 1', | |
| 214 "annotations": { | |
| 215 "summary": "see https://bigasterisk.com/front-door-lock/" | |
| 216 }, | |
| 34 | 217 }, |
| 218 { | |
| 36 | 219 "alert": "front_door_lock_svc_down", |
| 220 'expr': 'up{job="front-door-lock"} < 1', | |
| 221 "annotations": { | |
| 222 "summary": "see https://bigasterisk.com/front-door-lock/" | |
| 223 }, | |
| 34 | 224 }, |
| 36 | 225 { |
| 226 "alert": "front_door_lock_svc_no_mqtt", | |
| 227 'expr': 'mqtt_connected{job="front-door-lock"} < 1', | |
| 228 "annotations": { | |
| 229 "summary": "see https://bigasterisk.com/front-door-lock/" | |
| 230 }, | |
| 231 }, | |
| 232 { | |
| 233 "alert": "front_door_lock_esp32_no_mqtt", | |
| 234 'expr': 'hw_connected{job="front-door-lock"} < 1', | |
| 235 "annotations": { | |
| 236 "summary": "see https://bigasterisk.com/front-door-lock/" | |
| 237 }, | |
| 238 }, | |
| 239 ], | |
| 34 | 240 }, |
| 241 { | |
| 242 "name": | |
| 31 | 243 "alerts", |
| 23 | 244 "rules": [ |
| 245 { | |
| 246 "alert": "kube_node_status_bad_condition", | |
| 247 "for": "2h", | |
| 31 | 248 "labels": { |
| 249 "severity": "warning" | |
| 250 }, | |
| 23 | 251 "expr": 'kube_node_status_condition{condition=~".*Pressure",status="true"} > 0', |
| 252 }, | |
| 253 { | |
| 254 "alert": "housePower", | |
|
28
e114edff93dc
more explicit intervals. try to get a single day of notification out of a disk err increase
drewp@bigasterisk.com
parents:
27
diff
changeset
|
255 "for": "1h", |
| 31 | 256 "labels": { |
| 257 "severity": "waste" | |
| 258 }, | |
| 23 | 259 "expr": "house_power_w > 4000", |
| 31 | 260 "annotations": { |
| 261 "summary": "house power usage over 4KW" | |
| 262 }, | |
| 23 | 263 }, |
| 264 { | |
| 265 "alert": "host_root_fs_space_low", | |
| 266 "for": "20m", | |
| 31 | 267 "labels": { |
| 268 "severity": "warning" | |
| 269 }, | |
| 34 | 270 "expr": 'disk_free{host!="garage",path="/"} < 20G', |
| 23 | 271 }, |
| 272 { | |
| 273 "alert": "zpool_space_low", | |
| 274 "for": "20m", | |
| 31 | 275 "labels": { |
| 276 "severity": "warning" | |
| 277 }, | |
| 23 | 278 "expr": 'last_over_time(zfs_pool_free_bytes{pool="stor7"}[1h]) < 100G', |
| 279 }, | |
| 280 { | |
| 281 "alert": "disk_week_incr", | |
| 282 "for": "20m", | |
| 31 | 283 "labels": { |
| 284 "severity": "warning" | |
| 285 }, | |
| 23 | 286 "expr": 'round(increase(disk_used{path=~"/my/.*"}[1d])/1M) > 5000', |
| 31 | 287 "annotations": { |
| 288 "summary": "high mb/week on zfs dir" | |
| 289 }, | |
| 23 | 290 }, |
| 291 { | |
| 292 "alert": "high_logging", | |
| 31 | 293 "for": "3h", |
| 294 "labels": { | |
| 295 "severity": "waste" | |
| 296 }, | |
| 23 | 297 "expr": "sum by (container) (rate(kubelet_container_log_filesystem_used_bytes[3h])) > 4k", |
| 31 | 298 "annotations": { |
| 299 "summary": "high log output rate" | |
| 300 }, | |
| 23 | 301 }, |
| 302 { | |
| 303 "alert": "stale_process", | |
| 304 "for": "1d", | |
| 31 | 305 "labels": { |
| 306 "severity": "dataRisk" | |
| 307 }, | |
| 23 | 308 "expr": "round((time() - filestat_modification_time/1e9) / 86400) > 14", |
| 31 | 309 "annotations": { |
| 310 "summary": "process time is old" | |
| 311 }, | |
| 23 | 312 }, |
| 313 { | |
| 314 "alert": "starlette", | |
| 315 "for": "1m", | |
| 31 | 316 "labels": { |
| 317 "severity": "fix" | |
| 318 }, | |
| 23 | 319 "expr": 'starlette_request_duration_seconds_created{app_name="starlette"}', |
| 31 | 320 "annotations": { |
| 321 "summary": "set starlette app name" | |
| 322 }, | |
| 23 | 323 }, |
| 324 { | |
| 325 "alert": "ssl_certs_expiring_soon", | |
| 326 "expr": "min((min_over_time(probe_ssl_earliest_cert_expiry[1d])-time())/86400) < 10", | |
| 31 | 327 "labels": { |
| 328 "severity": "warning" | |
| 329 }, | |
| 23 | 330 "annotations": { |
| 331 "summary": "cert expiring soon. See https://bigasterisk.com/grafana/d/z1YtDa3Gz/certs?orgId=1\nVALUE = {{ $value }}" | |
| 332 }, | |
| 333 }, | |
| 334 ], | |
| 335 }, | |
|
32
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
336 ] + hostsExpectedOnline(ctx)['groups'] |
| 23 | 337 } |
| 338 | |
| 339 | |
| 340 def _runJson(ctx, cmd): | |
| 341 return json.loads(ctx.run(cmd, hide="stdout").stdout) | |
| 342 | |
| 343 | |
| 344 def hostsExpectedOnline(ctx): | |
| 345 return _runJson(ctx, "cd /my/serv/lanscape; pdm run python hosts_expected_online.py") | |
| 346 | |
| 347 | |
|
32
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
348 def expectedK8sNodesGroup(ctx): |
| 23 | 349 getNode = _runJson(ctx, "kubectl get node -o json") |
| 350 hosts = [item["metadata"]["name"] for item in getNode["items"]] | |
| 351 optionalHosts = {'slash'} | |
| 352 return { | |
|
32
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
353 "name": |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
354 "k8s_expected_nodes", |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
355 "rules": [{ |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
356 "alert": "kube_node_log_size_report_" + h, |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
357 "expr": 'absent(kubelet_container_log_filesystem_used_bytes{instance="%s"})' % h, |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
358 "for": "1h", |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
359 "annotations": { |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
360 "summary": f"no recent k8s log size report from host {h}" |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
361 }, |
|
eb1de82c93aa
refactor the merging of all the groups
drewp@bigasterisk.com
parents:
31
diff
changeset
|
362 } for h in hosts if h not in optionalHosts], |
| 23 | 363 } |