Grafana SLIs

This page documents Service Level Indicators (SLIs) for monitoring the health of Grafana. To set up Service Level Objectives (SLOs), see Set up SLOs with Sloth.

These metrics are recommended as Service Level Indicators for Grafana.

The slo_group label

Grafana’s HTTP request metrics include a slo_group label that Grafana Labs uses internally to classify endpoints by their service level tier. When building SLIs, filter on this label to exclude endpoints that are intentionally not covered by any SLO:

slo_group value	Meaning
high-fast	Default for most handlers; expected to respond quickly
high-medium	Handlers that may take longer to respond
high-slow	Proxy and downstream handlers; high availability but no latency guarantees
low	Experimental or unstable handlers not covered by the standard SLO
none	Handlers where errors are expected (for example, WebSocket connections); exclude from SLIs

Use slo_group!="none" in your SLI queries to exclude the none group.

HTTP API

Metric	Type	Description
grafana_http_request_duration_seconds	Histogram	HTTP request latency by handler, method, status code, and slo_group
grafana_http_request_in_flight	Gauge	Number of HTTP requests currently being served
grafana_api_response_status_total	Counter	API responses by HTTP status code
grafana_page_response_status_total	Counter	Page responses by HTTP status code
grafana_proxy_response_status_total	Counter	Proxy responses by HTTP status code

Datasource health

Metric	Type	Description
grafana_datasource_request_duration_seconds	Histogram	Datasource request latency by datasource, type, method, and status code
grafana_datasource_request_total	Counter	Datasource requests by datasource, type, method, and status code
grafana_datasource_request_in_flight	Gauge	Number of datasource requests currently in flight

Database

Metric	Type	Description
grafana_database_conn_in_use	Gauge	Number of database connections currently in use
grafana_database_conn_idle	Gauge	Number of idle database connections
grafana_database_conn_wait_count_total	Counter	Total number of times a request waited for a database connection
grafana_database_conn_wait_duration_seconds	Gauge	Total time blocked waiting for a database connection

Alerting

Metric	Type	Description
grafana_alerting_active_alerts	Gauge	Number of active alerts
grafana_alerting_alerts	Gauge	Number of alerts by state in the scheduler
grafana_alerting_schedule_alert_rules	Gauge	Number of alert rules considered for evaluation at the next tick
grafana_alerting_execution_time_milliseconds	Summary	Alert rule evaluation execution time
grafana_alerting_notification_latency_seconds	Summary	Alert notification send latency

Plugins

Metric	Type	Description
grafana_plugin_request_duration_milliseconds	Histogram	Plugin request duration by endpoint, plugin ID, and status
grafana_plugin_request_total	Counter	Plugin requests by endpoint, plugin ID, and status

Rendering

Metric	Type	Description
grafana_rendering_queue_size	Gauge	Number of rendering requests currently queued