Monitoring
Use Fleet Observability for fleet-wide workload metrics. It provides authenticated ingestion, trusted identity labels, scoped queries, and GPU observability through vClustervClusterAn open-source software product that creates and manages tenant clusters within Kubernetes infrastructure. vCluster provides tenant isolation capabilities while reducing infrastructure costs. Platform.
You can also scrape the Platform's own Prometheus metrics. The older OpenTelemetry pages in this section remain as reference architectures for deployments that can't use Fleet Observability. They aren't Platform-managed monitoring features.
Workload metrics​
| If you want to | Use |
|---|---|
| Monitor one connected cluster with Prometheus and Grafana | Single-cluster monitoring |
| Monitor workloads across a fleet | Fleet Observability |
| Query metrics with trusted project and tenant identity | Fleet Observability labels |
| Monitor Platform components | Platform health metrics |
| Understand older collector architectures | Fleet metrics architectures |
Single-cluster monitoring​
For a single connected cluster, install a current kube-prometheus-stack Helm chart from
the Platform UI. The chart deploys Prometheus, Grafana, the Prometheus Operator, and cluster
exporters. Check the
upstream releases
for the latest chart version and confirm it supports your Kubernetes version before
installing.
-
Go to Infrastructure > Control PlaneControl PlaneThe container orchestration layer that exposes the API and interfaces to define, deploy, and manage the lifecycle of containers. In vCluster, each tenant cluster has its own control plane components. Clusters and select a cluster.
-
Navigate to the Apps tab.
-
Click Install App and configure the Helm chart with the following settings. Set Chart Version to the version you verified in the previous step.
Setting Value Chart Repository URL https://prometheus-community.github.io/helm-chartsChart Name kube-prometheus-stackChart Version <compatible-version>Namespace kube-prometheus-stackRelease Name kube-prometheus-stack -
Review the upstream chart configuration for storage, retention, resource requests, and ingress.
-
Click Install.
After installation, you have a complete monitoring setup with Prometheus scraping cluster metrics and Grafana for visualization. If your tenant clusters run in Shared Nodes mode, see Prometheus node metrics on shared nodes to configure kubelet scraping.
Don't use kube-prometheus-stack if you want to aggregate metrics across multiple tenant clusters. Use Fleet Observability instead.
Reference architectures​
The following guides predate Fleet Observability. Use them only when you need to own the collector and backend configuration directly:
- Fleet monitoring with OpenTelemetry covers Shared Nodes and Private Nodes collector examples, PromQL queries, and Grafana dashboards.
- Aggregating metrics with OpenTelemetry shows a smaller collector-to-Prometheus example.
- Prometheus node metrics on shared nodes explains kubelet discovery and scraping behavior in the Shared Nodes tenancy model.
These architectures use self-reported labels and don't provide Platform-enforced access scoping. See Fleet metrics architectures before using them.
Platform health metrics​
vCluster Platform exposes Prometheus-conformant metrics from its internal components, including the API gateway, integrated Kubernetes API serverAPI ServerThe core component of Kubernetes that exposes the Kubernetes API. It is the front-end for the Kubernetes control plane and handles all REST operations, validating and configuring data for API objects., controller manager, and Go runtime. These metrics cover request counts, latency, and error rates for all platform operations.
Use a Prometheus ServiceMonitor to scrape these metrics automatically, or access the /metrics endpoint directly. For platform pod log level and output format, see Platform Process Logging.
For workload log collection, platform audit logs, and platform process logs, see Logging.