View alert status by node or service

When alerts fire, you need to know which nodes or services are affected and what state each alert is in. Use the Alert Status page to get that picture at a glance, triage by state, and silence alerts without switching pages.

Check alerts for a specific node

To see all alerts affecting a single node:

1. Go to Alerts > Status.
2. Toggle Group by node in the toolbar.
3. Find your node and expand it to see its individual alerts and their states.

Filter alerts by state

Use the State dropdown in the toolbar to show only alerts in a specific state: Normal, Pending, Firing, Recovering, No Data, Error or Silenced.

Check alerts for a specific service

To filter the alert list by service:

1. Go to Alerts > Status.
2. Click Show/Hide filters in the toolbar to reveal the column filters.
3. In the Service column filter, enter the service name.

The table updates to show only alerts associated with that service.

Get details on an alert

Click any alert row to open the details pane. From here you can:

• See what triggered the alert: the Details tab shows the summary, description, state and duration, node, service, severity, triggered at timestamp, and the MetricsQL expression. Check Rule configuration to see evaluation settings, template name, folder, and rule health.
• Debug custom templates or verify label values: switch to Raw data to inspect the full label set and JSON payload.
• Move between alerts: use the arrow buttons in the pane header to go to the next or previous alert without closing the pane.

Silence an alert

You need Editor role or higher to silence alerts.

1. Click the actions menu on the alert row.
2. Click Silence. PMM opens the Silences page with the alert labels pre-filled.
3. Set the duration and confirm.

Silenced alerts stay visible in the table with a Silenced badge. To unsilence, open the actions menu and click Unsilence.

Make custom alerts appear in this view

Built-in PMM alert templates automatically include the labels that this page uses to group alerts by node and service.

The following templates show data even when no alert is firing, so you can see how close a node or service is to triggering an alert before it does:

• Node high CPU load
• MongoDB DBPath disk space utilization
• PostgreSQL high rate of transaction rollbacks
• PostgreSQL connections in use

Support for more templates to show data when no alert is firing will be added in future releases.

If your custom templates do not show up correctly, add these labels to the template definition:

• node_name: identifies the monitored node.
• service_name: identifies the monitored service.

For instructions, see Alert rules and alert templates.