Pending Workloads in Grafana

Monitoring Pending Workloads using the VisibilityOnDemand feature in Grafana.

This guide explains how to monitor pending workloads in Grafana using the VisibilityOnDemand feature.

The intended audience for this page are batch administrators for ClusterQueue visibility, and batch users for LocalQueue visibility.

Before you begin

Make sure the following conditions are met:

Setting Up Grafana for Pending Workloads

Step 1: Configure Cluster Permissions

To enable visibility, create a ClusterRole and ClusterRoleBinding for either ClusterQueue or LocalQueue:

  • For ClusterQueue visibility:
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: grafana-cluster-queue-reader
rules:
  - apiGroups: ["visibility.kueue.x-k8s.io"]
    resources: ["clusterqueues/pendingworkloads"]
    verbs: ["get", "list", "watch"]
  - apiGroups: ["kueue.x-k8s.io"]
    resources: ["clusterqueues"]
    verbs: ["get", "list", "watch"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: grafana-cluster-queue-reader-binding
subjects:
  - kind: ServiceAccount
    name: default
    namespace: default
roleRef:
  kind: ClusterRole
  name: grafana-cluster-queue-reader
  apiGroup: rbac.authorization.k8s.io
  • For LocalQueue visibility:
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: grafana-local-queue-reader
rules:
  - apiGroups: ["visibility.kueue.x-k8s.io"]
    resources: ["localqueues/pendingworkloads"]
    verbs: ["get", "list", "watch"]
  - apiGroups: ["kueue.x-k8s.io"]
    resources: ["localqueues"]
    verbs: ["get", "list", "watch"]
  - apiGroups: [""]
    resources: ["namespaces"]
    verbs: ["get", "list", "watch"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: grafana-local-queue-reader-binding
subjects:
  - kind: ServiceAccount
    name: default
    namespace: default
roleRef:
  kind: ClusterRole
  name: grafana-local-queue-reader
  apiGroup: rbac.authorization.k8s.io

Apply the appropriate configuration:

kubectl apply -f <filename>.yaml

Step 2: Generate a Service Account Token

Create a token for Grafana authentication:

TOKEN=$(kubectl create token default -n default)
echo $TOKEN

Save the token for use in Step 5.

Step 3: Set up port forwarding for Grafana

Access Grafana locally:

kubectl port-forward -n monitoring service/grafana 3000:3000

Grafana is now available at http://localhost:3000.

Step 4: Install the Infinity Plugin

  1. Open Grafana at http://localhost:3000.
  2. Log in (default credentials: admin/admin).
  3. Go to Connections > Add new connection.
  4. Search for Infinity and click Install.

Step 5: Configure the Infinity Data Source

  1. Go to Connections > Data sources and click + Add new data source.
  2. Select Infinity.
  3. Configure the data source:
    • Authentication: Set the Bearer Token to the token generated in Step 2.
    • Network: Enable Skip TLS Verify.
    • Security: Add https://kubernetes.default.svc to allowed hosts and set Query security to Allowed.
  4. Click Save & test to verify the configuration.

Step 6: Import the Pending Workloads Dashboard

  1. Download the appropriate dashboard JSON:
  2. In Grafana, go to Dashboards > New > Import.
  3. Select Upload dashboard JSON file and choose the downloaded file.
  4. Select the Infinity data source configured in Step 5.
  5. Click Import.

Step 7: Set Up ClusterQueue

To configure a basic ClusterQueue, apply the following:

apiVersion: kueue.x-k8s.io/v1beta1
kind: ResourceFlavor
metadata:
  name: "default-flavor"
---
apiVersion: kueue.x-k8s.io/v1beta1
kind: ClusterQueue
metadata:
  name: "cluster-queue"
spec:
  namespaceSelector: {} # match all.
  resourceGroups:
  - coveredResources: ["cpu", "memory"]
    flavors:
    - name: "default-flavor"
      resources:
      - name: "cpu"
        nominalQuota: 9
      - name: "memory"
        nominalQuota: 36Gi
---
apiVersion: kueue.x-k8s.io/v1beta1
kind: LocalQueue
metadata:
  namespace: "default"
  name: "user-queue"
spec:
  clusterQueue: "cluster-queue"

Apply the configuration:

kubectl apply -f https://kueue.sigs.k8s.io/examples/admin/single-clusterqueue-setup.yaml

Step 8: Create Sample Workloads

To populate the dashboard with data, create sample jobs:

apiVersion: batch/v1
kind: Job
metadata:
  generateName: sample-job-
  namespace: default
  labels:
    kueue.x-k8s.io/queue-name: user-queue
spec:
  parallelism: 3
  completions: 3
  suspend: true
  template:
    spec:
      containers:
      - name: dummy-job
        image: registry.k8s.io/e2e-test-images/agnhost:2.53
        command: [ "/bin/sh" ]
        args: [ "-c", "sleep 60" ]
        resources:
          requests:
            cpu: "1"
            memory: "200Mi"
      restartPolicy: Never

Apply the job multiple times:

for i in {1..6}; do kubectl create -f https://kueue.sigs.k8s.io/examples/jobs/sample-job.yaml; done

Step 9: View the Dashboard

  1. In Grafana, navigate to Dashboards.
  2. Select the imported dashboard (e.g., “Pending Workloads for ClusterQueue visibility”).
  3. Verify that pending workloads are displayed.

ClusterQueue Visibility Dashboard

LocalQueue Visibility Dashboard

Troubleshooting

No data in dashboard

Ensure jobs are created and the Infinity data source is correctly configured.

Permission errors

Verify the ClusterRole and ClusterRoleBinding are applied correctly.

Grafana inaccessible

Check port forwarding and ensure the Grafana service is running in the monitoring namespace.


最后修改 July 2, 2025: Cherry-pick website commits from upstream/main (#5825) (cd1b0ecf)