Skip to content

DaemonSets

A DaemonSet ensures that all (or some) nodes run a copy of a pod. As nodes are added to the cluster, pods are added to them. As nodes are removed from the cluster, those pods are garbage collected.

DaemonSets are typically used for cluster-wide services that need to run on every node.

Common Use Cases

Use Case Examples
Logging Collectors Fluentd, Filebeat, Logstash
Monitoring Agents Prometheus Node Exporter, Datadog Agent
Storage Daemons Ceph, GlusterFS
Network Plugins Calico, Cilium, Weave Net
Security Agents Falco, Twistlock

How DaemonSets Work

  • When you create a DaemonSet, it automatically creates a pod on each node
  • When a new node joins the cluster, a pod is scheduled on it
  • When a node is removed, the pod is garbage collected
  • Deleting a DaemonSet cleans up all the pods it created

Resources

DaemonSets

DaemonSets

Perform Rolling Update on DaemonSet

References

Basic DaemonSet for logging agent

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: fluentd
  namespace: kube-system
  labels:
    app: fluentd
spec:
  selector:
    matchLabels:
      app: fluentd
  template:
    metadata:
      labels:
        app: fluentd
    spec:
      containers:
        - name: fluentd
          image: fluentd:v1.16
          resources:
            limits:
              memory: 200Mi
            requests:
              cpu: 100m
              memory: 200Mi
          volumeMounts:
            - name: varlog
              mountPath: /var/log
            - name: containers
              mountPath: /var/lib/docker/containers
              readOnly: true
      terminationGracePeriodSeconds: 30
      volumes:
        - name: varlog
          hostPath:
            path: /var/log
        - name: containers
          hostPath:
            path: /var/lib/docker/containers

DaemonSet running only on specific nodes using nodeSelector

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: gpu-monitor
spec:
  selector:
    matchLabels:
      app: gpu-monitor
  template:
    metadata:
      labels:
        app: gpu-monitor
    spec:
      nodeSelector:
        gpu: "true"
      containers:
        - name: gpu-monitor
          image: nvidia/dcgm-exporter:latest

DaemonSet with tolerations to run on control plane nodes

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: node-exporter
spec:
  selector:
    matchLabels:
      app: node-exporter
  template:
    metadata:
      labels:
        app: node-exporter
    spec:
      tolerations:
        - key: node-role.kubernetes.io/control-plane
          operator: Exists
          effect: NoSchedule
      containers:
        - name: node-exporter
          image: prom/node-exporter:latest
          ports:
            - containerPort: 9100
              hostPort: 9100
Create DaemonSet
oc apply -f daemonset.yaml
Get DaemonSets
oc get daemonsets -A
Describe DaemonSet
oc describe daemonset fluentd -n kube-system
Check DaemonSet Pods
oc get pods -l app=fluentd -o wide
Delete DaemonSet
oc delete daemonset fluentd
Create DaemonSet
kubectl apply -f daemonset.yaml
Get DaemonSets
kubectl get daemonsets -A
Describe DaemonSet
kubectl describe daemonset fluentd -n kube-system
Check DaemonSet Pods
kubectl get pods -l app=fluentd -o wide
Delete DaemonSet
kubectl delete daemonset fluentd

DaemonSet vs Deployment

Feature Deployment DaemonSet
Replicas User-defined count One per node (automatic)
Scaling Manual or HPA Automatic with cluster
Node Coverage Scheduled by kube-scheduler Guaranteed per node
Use Case Application workloads Node-level services