grafana
The leading platform for analytics and monitoring
Component Information
| Property | Value |
|---|---|
| Chart Type | application |
| Upstream Project | grafana |
| Maintainers | Platform Engineering Team (repo) |
Why Grafana?
Grafana provides a unified interface for querying metrics and logs. The kube-prometheus-stack chart bundles it with Prometheus and includes dashboards for Kubernetes components.
Grafana supports multiple data sources. In this platform:
- Prometheus: For metrics (resource usage, application performance, SLO burn rates)
- Loki: For logs (structured log queries, correlation with metrics)
The ability to correlate metrics and logs in the same interface helps with debugging. You can see a spike in errors in a Prometheus graph, then query Loki for the corresponding error logs.
Grafana dashboards can be stored as ConfigMaps, making observability infrastructure reproducible and versionable in Git.
Architecture Role
Grafana sits at Layer 3 of the platform, the Developer-Facing Applications layer. It’s a user interface that consumes data from services below.
Key integration points:
- Prometheus: Configured as a data source for metrics
- Loki: Configured as a data source for logs
- Vault → External Secrets: Admin credentials synced
- Gateway API: Exposed via HTTPRoute for browser access
- ConfigMaps: Dashboards stored as code
The configuration uses Unified Alerting (Grafana’s built-in alerting) as the primary alerting interface.
See Observability Model for the complete observability architecture.
Configuration Values
Grafana is bundled inside the kube-prometheus-stack deployment. Its values are managed in that chart; see the Prometheus component page for the generated values table.