Grafana Dashboard 6417 Kubernetes Pods

💡 Quick Answer: Grafana dashboard 6417 (“Kubernetes Pods”) displays real-time CPU, memory, network I/O, and filesystem usage per pod. Import it via Grafana UI → Dashboards → Import → ID 6417 → select your Prometheus data source. Requires kube-state-metrics and node-exporter running in your cluster.

The Problem

You need a comprehensive pod-level monitoring dashboard that shows:

• CPU usage vs requests vs limits per pod
• Memory consumption and OOM risk
• Network receive/transmit bytes per pod
• Filesystem usage inside containers
• Pod restart counts and ready status

Dashboard 6417 is one of the most popular Kubernetes dashboards on Grafana.com with 10M+ downloads.

The Solution

Prerequisites

# Verify Prometheus is scraping Kubernetes metrics
kubectl get pods -n monitoring -l app.kubernetes.io/name=prometheus
kubectl get pods -n monitoring -l app.kubernetes.io/name=kube-state-metrics

# Verify metrics are available
kubectl port-forward -n monitoring svc/prometheus 9090:9090 &
curl -s "http://localhost:9090/api/v1/query?query=container_cpu_usage_seconds_total" | jq '.data.result | length'

Import Dashboard 6417

Method 1: Grafana UI

1. Open Grafana → Dashboards → Import
2. Enter dashboard ID: 6417
3. Click Load
4. Select your Prometheus data source
5. Click Import

Method 2: Grafana API

# Download dashboard JSON
curl -s https://grafana.com/api/dashboards/6417/revisions/1/download \
  > dashboard-6417.json

# Import via Grafana API
curl -X POST http://admin:admin@grafana:3000/api/dashboards/import \
  -H "Content-Type: application/json" \
  -d "{
    \"dashboard\": $(cat dashboard-6417.json),
    \"overwrite\": true,
    \"inputs\": [{
      \"name\": \"DS_PROMETHEUS\",
      \"type\": \"datasource\",
      \"pluginId\": \"prometheus\",
      \"value\": \"Prometheus\"
    }]
  }"

Method 3: Helm values (kube-prometheus-stack)

# values.yaml for kube-prometheus-stack
grafana:
  dashboardProviders:
    dashboardproviders.yaml:
      apiVersion: 1
      providers:
      - name: default
        folder: Kubernetes
        type: file
        options:
          path: /var/lib/grafana/dashboards/default
  dashboards:
    default:
      kubernetes-pods:
        gnetId: 6417
        revision: 1
        datasource: Prometheus

helm upgrade --install kube-prometheus-stack prometheus-community/kube-prometheus-stack \
  -n monitoring -f values.yaml

Dashboard Panels Explained

Customize for Your Cluster

// Add namespace variable filter
{
  "templating": {
    "list": [
      {
        "name": "namespace",
        "query": "label_values(kube_pod_info, namespace)",
        "type": "query",
        "multi": true,
        "includeAll": true
      },
      {
        "name": "pod",
        "query": "label_values(kube_pod_info{namespace=~\"$namespace\"}, pod)",
        "type": "query",
        "multi": true,
        "includeAll": true
      }
    ]
  }
}

Alternative Dashboards

Common Issues

“No data” on all panels

Prometheus data source not configured or wrong name. Go to Grafana → Configuration → Data Sources → verify Prometheus URL (usually http://prometheus-server:9090 or http://prometheus-kube-prometheus-prometheus:9090).

Missing container_* metrics

cAdvisor metrics not being scraped. Check if Prometheus has a kubelet job: up{job="kubelet"}. For kube-prometheus-stack, this is configured automatically.

Dashboard shows only system pods

Namespace filter needs adjustment. Edit the dashboard variable to include your application namespaces.

Best Practices

• Pin dashboard revision in Helm values — prevents unexpected changes on upgrade
• Add namespace and pod filters — default dashboard shows everything, gets noisy on large clusters
• Set alert rules from dashboard queries — CPU > 90% of limit, memory > 80% of limit
• Use container_memory_working_set_bytes not container_memory_usage_bytes — working set is what triggers OOMKill

Key Takeaways

• Dashboard 6417 is the standard Kubernetes pod monitoring dashboard (10M+ downloads)
• Import via UI (ID 6417), API, or Helm values for GitOps
• Requires Prometheus + kube-state-metrics + cAdvisor (kubelet) metrics
• Customize with namespace/pod variable filters for large clusters
• Use working set memory (not RSS) for accurate OOM risk visualization