Fleet metrics architectures
Use Fleet Observability for new fleet-wide metrics deployments. It's the Platform-managed observability feature. The other pages in this section are operator-maintained reference architectures that predate Fleet Observability.
The architectures don't share a label contract. Dashboards, alerts, and recording rules built for one architecture don't automatically work with another.
Compare the architectures​
| Fleet Observability | Fleet monitoring with OpenTelemetry | Aggregating Metrics | |
|---|---|---|---|
| Status | Recommended Platform feature | Reference architecture | Reference architecture |
| Identity labels | vcluster_platform_*, authenticated | vcluster_*/cluster, self-reported | loft_*, self-reported |
| Read access control | Scoped per tenant or cluster, Platform-authorized | None, any reader sees all data | None, any reader sees all data |
| GPU observability | DCGM, NVSentinel | Not covered | Not covered |
| Requires Argo CD for the bundled deploy path | Yes (manual config available for Standalone) | No | No |
| Runs on embedded Kubernetes platform deployments | No | Yes | Yes |
| Included examples | Query examples | PromQL library and two Grafana dashboards | Prometheus and Grafana example |
| Operational ownership | Platform gateway and access policy, plus your metrics backend | You maintain and validate the collector Apps and backend | You maintain and validate the collector and backend |
When a reference architecture is necessary​
Fleet Observability's bundled Argo CD deployment doesn't cover vCluster Standalone. You can still configure a collector manually. Embedded Kubernetes deployments can't run the observability gateway.
If either constraint applies, the older guides can help you design a direct collector path:
- Use Fleet monitoring with OpenTelemetry as a detailed example for Shared Nodes and Private Nodes deployments.
- Use Aggregating Metrics as a smaller example built around a single OpenTelemetry Collector DaemonSet.
These guides aren't maintained as turnkey production configurations. Validate chart versions, security settings, resource sizing, and backend compatibility for your environment.
Next steps​
- Start a supported deployment with the Fleet Observability quickstart.
- See the labels reference before migrating dashboards or alerts from an older architecture.