alerts.yaml

# used for generate prometheusrule and alertmanagerconfig
- name: mongodb-down
  alertType: monitor
  expr: mongodb_up{service="__fullname__",namespace="__namespace__"} == 0
  message: MongoDB [namespace {{ $labels.namespace }}] [service {{ $labels.service }}] [instance {{ $labels.instance }}] down
  for: 0m
  alertLevels:
    - severity: critical
  receivers:
    - alertChannelID: 1
      interval: 10m
- name: mongodb-too-many-connections
  alertType: monitor
  expr: avg by(instance) (rate(mongodb_connections{state="current",service="__fullname__",namespace="__namespace__"}[1m])) / avg by(instance) (sum (mongodb_connections{service="__fullname__",namespace="__namespace__"})) * 100 > 80
  message: MongoDB too many connections [namespace {{ $labels.namespace }}] [service {{ $labels.service }}] [instance {{ $labels.instance }}]
  for: 5m
  alertLevels:
    - severity: error
  receivers:
    - alertChannelID: 1
      interval: 10m
- name: mongodb-low-avalible-connections
  alertType: monitor
  expr: mongodb_connections{state="available",service="__fullname__",namespace="__namespace__"} < 100
  message: MongoDB few available connections [namespace {{ $labels.namespace }}] [service {{ $labels.service }}] [instance {{ $labels.instance }}]
  for: 5m
  alertLevels:
    - severity: critical
  receivers:
    - alertChannelID: 1
      interval: 10m
- name: mongodb-too-many-memories
  alertType: monitor
  expr: mongodb_mongod_wiredtiger_cache_bytes{type="total",service="__fullname__",namespace="__namespace__"} / on(pod,namespace,instance,job,service) mongodb_mongod_wiredtiger_cache_max_bytes{service="__fullname__",namespace="__namespace__"} * 100 > 80
  message: MongoDB too many memories [namespace {{ $labels.namespace }}] [service {{ $labels.service }}] [instance {{ $labels.instance }}]
  for: 5m
  alertLevels:
    - severity: error
  inhibitLabels: [instance]
  receivers:
    - alertChannelID: 1
      interval: 10m