Override repeat_interval by child Notification policy

Grafana Alerting
1

I am trying to achieve sort out repeating of long-time running alerts by priority. By default policy is it repeated each 4h, which is sufficient for most alerts. Few alerts are more critical, others are with low priority (and repair takes several days)

  • What happened?

    • long running alert is repeated by 4h

  • What expected

    • repeat matched alert each 2h30m (or whatever else) for instances with higher priority (matched by tags)

Is there any other option to enable overriding?

apiVersion: 1
policies:
    - orgId: 2
      receiver: grafana-default-email
      group_by:
        - alertname
      routes:
        - receiver: Teams LKN Alerting - mould fridge
          object_matchers:
            - - area
              - =
              - MldResin
          repeat_interval: 2h30m
      group_wait: 30s
      group_interval: 5m
      repeat_interval: 4h

obrazek
obrazek746×231 8.76 KB

  • self-hosted Grafana 12.2
2

Yes, this is the correct way to override timing options without changing the parent policy.

Child notification policies inherit settings from their parent, but any timing options you set in a child policy (like group_wait, group_interval, or repeat_interval) will override the parent’s values for alerts that match that child policy.

This design lets you customize notification behavior for narrower scopes (such as severity or team), while keeping shared defaults at the top level.

More about notification policies here: Notification policies | Grafana documentation

3

Thank you for reply. Created issue #113692 on github.

Please if someone have note to this behavior, post additional comments here.