The VMAlertmanagerConfig provides way to configure VMAlertmanager configuration with CRD. It allows to define different configuration parts, which will be merged by operator into config.

It behaves like other config parts - VMServiceScrape and etc.

Read Usage and Special case before using.

Specification #

You can see the full actual specification of the VMAlertmanagerConfig resource in the API docs -> VMAlertmanagerConfig.

Also, you can check out the examples section.

Validation #

The operator performs validation of VMAlertmanagerConfig. In case of any misconfiguration it adds corresponding validation fail message into status.lastSyncError field and updates status.status field to failed status.

For example VMAlertmanagerConfig without required fields:

apiVersion: operator.victoriametrics.com/v1beta1
kind: VMAlertmanagerConfig
metadata:
  name: invalid-config
  namespace: default
spec:
  receivers:
  - name: blackhole
  - name: pagerduty
    pagerduty_configs:
    - url: http://example
  route:
    group_by:
    - alertname
    receiver: blackhole
    routes:
    - matchers:
      - alertname="pd"
      receiver: pagerduty
status:
  lastErrorParentAlertmanagerName: default/example-alertmanager
  lastSyncError: 'receiver at idx=2 is invalid: at idx=0 pagerduty_configs one of
    ''routing_key'' or ''service_key'' must be configured'
  lastSyncErrorTimestamp: 1722950290
  status: failed

Usage #

VMAlertmanagerConfig allows delegating notification configuration to the kubernetes cluster users. The application owner may configure notifications by defining it at VMAlertmanagerConfig.

With the combination of VMRule and VMServiceScrape it allows delegating configuration observability to application owners, and uses popular GitOps practice.

Operator combines VMAlertmanagerConfigs into a single configuration file for VMAlertmanager.

apiVersion: operator.victoriametrics.com/v1beta1
kind: VMAlertmanagerConfig
metadata:
  name: example-email-web
  namespace: production
spec:
  route:
    receiver: email
    group_interval: 1m
    routes:
      - receiver: email
        matchers:
          - {severity =~ "warning|critical", app_name = "blog"}
  receivers:
    - name: email
      email_configs:
        - to: some-email@example.com
          from: alerting@example.com
          smarthost: example.com:25
          text: ALARM

Special Case #

VMAlertmanagerConfig has enforced namespace matcher. Alerts must have a proper namespace label, with the same value as name of namespace for VMAlertmanagerConfig.

It can be disabled, by setting the following value to the VMAlertmanager: spec.disableNamespaceMatcher: true.

Examples #

apiVersion: operator.victoriametrics.com/v1beta1
kind: VMAlertmanagerConfig
metadata:
  name: example
  namespace: default
spec:
  inhibit_rules:
    - equals: []
      target_matchers: []
      source_matchers: []
  route:
    routes:
      - receiver: webhook
        continue: true
    receiver: email
    group_by: []
    continue: false
    matchers:
      - job = "alertmanager"
    group_wait: 30s
    group_interval: 45s
    repeat_interval: 1h
  mute_time_intervals:
    - name: base
      time_intervals:
        - times:
            - start_time: ""
              end_time: ""
          weekdays: []
          days_of_month: []
          months: []
          years: []
  receivers:
      email_configs: []
      webhook_configs:
        - url: http://some-other-wh
      pagerduty_configs: []
      pushover_configs: []
      slack_configs: []
      opsgenie_configs: []
      victorops_configs: []
      wechat_configs: []