Working with notification policies - Amazon Managed Grafana

Working with notification policies

This documentation topic is designed for Grafana workspaces that support Grafana version 9.x.

For Grafana workspaces that support Grafana version 10.x, see Working in Grafana version 10.

For Grafana workspaces that support Grafana version 8.x, see Working in Grafana version 8.

Notification policies determine how alerts are routed to contact points. Policies have a tree structure, where each policy can have one or more child policies. Each policy, except for the root policy, can also match specific alert labels. Each alert is evaluated by the root policy and then by each child policy. If you enable the Continue matching subsequent sibling nodes option for a specific policy, then evaluation continues even after one or more matches. A parent policy’s configuration settings and contact point information govern the behavior of an alert that does not match any of the child policies. A root policy governs any alert that does not match a specific policy.

Note

You can create and edit notification policies for Grafana managed alerts. Notification policies for Alertmanager alerts are read-only.

Grouping notifications

Grouping categorizes alert notifications of similar nature into a single funnel. This allows you to control alert notifications during larger outages when many parts of a system fail at once causing a high number of alerts to initiate simultaneously.

Grouping example

Suppose you have 100 services connected to a database in different environments. These services are differentiated by the label env=environmentname. An alert rule is in place to monitor whether your services can reach the database. The alert rule creates alerts named alertname=DatabaseUnreachable.

If a network partition occurs, where half of your services can no longer reach the database, 50 different alerts are initiated. For this situation, you want to receive a single-page notification (as opposed to 50) with a list of the environments that are affected.

You can configure grouping to be group_by: [alertname] (not using the env label, which is different for each service). With this configuration in place, Grafana sends a single compact notification that has all the affected environments for this alert rule.

Special Groups

Grafana has two special groups. The default group, group_by: null groups all alerts together into a single group. You can also use a special label named ... to group alerts by all labels, effectively disabling grouping, and sending each alert into its own group.

Working with notifications

The following procedures show you how to create and manage notification policies.

To edit the root notification policy
  1. From your Grafana console, in the Grafana menu, choose the Alerting (bell) icon to open the Alerting page.

  2. Choose Notification policies.

  3. From the Alertmanager dropdown, select the Alertmanager you want to edit.

  4. In the Root policy section, choose the Edit icon (pen).

  5. In Default contact point, update the contact point where notifications should be sent for rules when alert rules do not match any specific policy.

  6. In Group by, choose the labels (or special groups) to group alerts by.

  7. In Timing options, select from the following options.

    • Group wait – Time to wait to buffer alerts of the same group before sending an initial notification. The default is 30 seconds.

    • Group interval – Minimum time interval between two notifications for a group. The default is 5 minutes.

    • Repeat interval – Minimum time interval before resending a notification if no new alerts were added to the group. The default is 4 hours.

  8. Choose Save to save your changes.

To add a new, top-level specific policy
  1. From your Grafana console, in the Grafana menu, choose the Alerting (bell) icon to open the Alerting page.

  2. Choose Notification policies.

  3. From the Alertmanager dropdown, select the Alertmanager you want to edit.

  4. In the Specific routing section, choose New specific policy.

  5. In the Matching labels section, add one or more matching alert labels. More information about label matching is later in this topic.

  6. In Contact point, add the contact point to send notifications to if the alert matches this specific policy. Nested policies override this contact point.

  7. Optionally, enable Continue matching subsequent sibling nodes to continue matching sibling policies even after the alert matched the current policy. When this policy is enabled, you can get more than one notification for the same alert.

  8. Optionally select Override grouping to specify a grouping different from the root policy.

  9. Optionally select Override general timings to override the timing options in the group notification policy.

  10. Choose Save policy to save your changes.

To add a nested policy
  1. Expand the specific policy you want to create a nested policy under.

  2. Choose Add nested policy, then add the details (as when adding a top-level specific policy).

  3. Choose Save policy to save your changes.

To edit a specific policy
  1. From the Alerting page, choose Notification policies to open the page that listing existing policies.

  2. Select the policy that you want to edit, then choose the Edit icon (pen).

  3. Make any changes (as when adding a top-level specific policy).

  4. Choose Save policy.

Searching for policies

You can search within the tree of policies by Label matchers or contact points.

  • To search by contact point, enter a partial or full name of a contact point in the Search by contact point field.

  • To search by label, enter a valid label matcher in the Search by label field. Multiple matchers can be entered, separated by a comma. For example, a valid matcher input could be severity=high, region=~EMEA|NA.

    Note

    When searching by label, all matched policies will be exact matches. Partial matches and regex-style matches are not supported.

How label matching works

A policy matches an alert if the alert's labels match all the Matching Labels specified on the policy.

  • Label – The name of the label to match. It must exactly match the label name of the alert.

  • Operator – The operator used to compare the label value with the matching label value. The available operators are:

    • = Select labels whose value exactly matches the provided string.

    • != Select labels whose value does not match the provided string.

    • =~ Select labels whose value match the regex interpreted value of the provided string (the provided string is interpreted as a regular expression.

    • != Select labels that do not match the provided regular expression.

  • Value – The value to match the label value to. It can match as a string or as a regular expression, depending on the operator chosen.