DaemonSets for per-node workloads in Kubernetes - Step-by-Step Execution

Process Flow - DaemonSets for per-node workloads

Create DaemonSet YAML

↓

kubectl apply DaemonSet

↓

Kubernetes Scheduler

↓

DaemonSet Controller creates Pod on each Node

↓

Pods run on all current Nodes

↓

New Node joins Cluster?

Yes↓

DaemonSet Controller adds Pod to new Node

↓

Pods run on all Nodes including new one

↓

Node removed?

Yes↓

Pod on removed Node deleted

↓

DaemonSet maintains one Pod per Node

This flow shows how a DaemonSet ensures one Pod runs on each node, including new nodes joining the cluster.

Execution Sample

Kubernetes

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: node-monitor
spec:
  selector:
    matchLabels:
      name: node-monitor
  template:
    metadata:
      labels:
        name: node-monitor
    spec:
      containers:
      - name: monitor
        image: busybox
        command: ["sh", "-c", "while true; do echo monitoring; sleep 30; done"]

This YAML creates a DaemonSet named node-monitor that runs a simple monitoring container on every node.

Process Table

Step	Action	Result	Pods per Node
1	Apply DaemonSet YAML with kubectl	DaemonSet resource created	0
2	DaemonSet Controller detects DaemonSet	Starts creating Pods	0
3	Create Pod on Node1	Pod running on Node1	Node1:1
4	Create Pod on Node2	Pod running on Node2	Node1:1, Node2:1
5	New Node3 joins cluster	DaemonSet Controller creates Pod on Node3	Node1:1, Node2:1, Node3:1
6	Node2 removed from cluster	Pod on Node2 deleted	Node1:1, Node3:1
7	DaemonSet maintains Pods on all nodes	Pods running on all current nodes	Node1:1, Node3:1

💡 DaemonSet ensures exactly one Pod per node; stops when no more nodes exist or DaemonSet deleted.

Status Tracker

Variable	Start	After Step 3	After Step 4	After Step 5	After Step 6	Final
Pods per Node	{}	{"Node1":1}	{"Node1":1, "Node2":1}	{"Node1":1, "Node2":1, "Node3":1}	{"Node1":1, "Node3":1}	{"Node1":1, "Node3":1}

Key Moments - 3 Insights

Why does the DaemonSet create a new Pod when a new node joins?

Because the DaemonSet controller watches the cluster nodes and ensures one Pod per node. When a new node appears (see Step 5), it creates a Pod there automatically.

What happens to Pods on nodes that are removed from the cluster?

Pods on removed nodes are deleted (Step 6) because the node no longer exists. The DaemonSet controller updates the Pod list accordingly.

Can a DaemonSet create more than one Pod per node?

No. The DaemonSet ensures exactly one Pod per node, as shown in the Pods per Node column in the execution table.

Visual Quiz - 3 Questions

Look at the execution table, how many Pods are running after Step 4?

A 1 Pod on Node1 and 1 Pod on Node2

B 1 Pod on Node1

C 1 Pod on Node1, Node2, and Node3

D No Pods running yet

💡 Hint

Check the 'Pods per Node' column at Step 4 in the execution table.

At which step does the DaemonSet add a Pod to a new node?

A Step 4

B Step 5

C Step 3

D Step 6

💡 Hint

Look for when Node3 first appears with a Pod in the execution table.

If Node2 is removed, what happens to the Pod on Node2?

A It remains running

B It moves to another node

C It is deleted

D It duplicates on Node1

💡 Hint

See Step 6 in the execution table where Node2 is removed.

Concept Snapshot

DaemonSet runs one Pod on each node in a Kubernetes cluster.
Apply a DaemonSet YAML with kubectl to create it.
Pods are automatically added to new nodes.
Pods on removed nodes are deleted.
Ensures per-node workload consistency.

Full Transcript

A DaemonSet in Kubernetes ensures that a specific Pod runs on every node in the cluster. When you create a DaemonSet using kubectl apply, the DaemonSet controller watches the cluster nodes and creates one Pod per node. If a new node joins, the controller adds a Pod there automatically. If a node is removed, the Pod on that node is deleted. This way, the DaemonSet keeps exactly one Pod running on each node, useful for per-node tasks like monitoring or logging.