Grafana
inari.grafana
A Docker Compose file for Grafana alone, generated from the catalog and checked by the safety inspector. No account needed.
Dashboard and visualisation server that queries time-series, log and SQL data sources directly and renders panels, with alert rules, notification routing and dashboards defined as JSON.
Grafana queries data sources where they live rather than ingesting anything itself, so a single dashboard can place a metrics panel beside a log panel and a SQL table over the same time range. Panels compose into dashboards with template variables, letting one dashboard serve many hosts, services or tenants instead of being duplicated per target. Unified alerting evaluates rules against those same queries and routes notifications through contact points with grouping and silencing. Dashboards and data sources can be provisioned from files, which is what keeps a deployment reproducible rather than hand-assembled.
You know it worked when
- The interface loads and an administrator signs in with changed credentials.
- A configured data source passes its connection test.
- A panel querying that source renders data.
- A dashboard survives a container restart.
- An alert rule transitions to firing and delivers a notification through its contact point.
Known sharp edges
- The default administrator credentials are widely known and remain accepted until changed.
- Dashboards, users and alert rules live in the configured database, and with the embedded default on a non-persistent volume all of it is lost on recreation.
- Anonymous access and open sign-up are configuration flags that are easy to enable for convenience and easy to forget afterwards.
- Data source credentials are held by the server, so anyone able to edit a panel can query whatever those credentials permit.