Step 1: Label Your Nodes

In Kubernetes, if you have two nodes and want to schedule specific pods onto a particular node based on labels or certain properties, you can achieve that using node labels combined with either nodeSelector or nodeAffinity. Both are native scheduling mechanisms that let you control where your workload runs, but they differ in flexibility.

---

Step 1: Label Your Nodes

First, list your nodes to identify their names:

kubectl get nodes --show-labels

Then, apply a label to the node you want to target (for example, label one node as disktype=ssd):

kubectl label nodes node1 disktype=ssd
kubectl label nodes node2 disktype=hdd

You can verify the labels applied with:

kubectl get nodes --show-labels

This assigns descriptive metadata to each node that can be referenced during scheduling.

---

Step 2: Use nodeSelector (Simple Approach)

The easiest way to schedule pods to a specific node is via nodeSelector. It matches key-value pairs of node labels.

Example Pod YAML:

apiVersion: v1
kind: Pod
metadata:
  name: nginx-ssd
spec:
  containers:
    - name: nginx
      image: nginx
  nodeSelector:
    disktype: ssd

This configuration ensures that the pod runs only on nodes labeled disktype=ssd.

However, note that nodeSelector performs a strict match — if no node has that label, the pod will remain unscheduled.

---

Step 3: Use nodeAffinity (Recommended Approach)

Node affinity provides more control and flexibility than nodeSelector. It supports both required (hard rule) and preferred (soft rule) conditions.

Example YAML using node affinity:

apiVersion: v1
kind: Pod
metadata:
  name: nginx-app
spec:
  containers:
    - name: nginx
      image: nginx:latest
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: disktype
            operator: In
            values:
            - ssd

This pod will only be scheduled on nodes that have the label disktype=ssd. The operator: In ensures that any node with that label value qualifies.

---

Step 4: Combining with Pod Labels (Optional)

If your pod has specific labels (like app=tier1), you can combine label selectors with node affinity to target both the pod type and the node property together. For example, only app=tier1 pods go to disktype=ssd nodes.

In a deployment YAML:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: tier1-deployment
spec:
  replicas: 2
  selector:
    matchLabels:
      app: tier1
  template:
    metadata:
      labels:
        app: tier1
    spec:
      affinity:
        nodeAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
            - matchExpressions:
              - key: disktype
                operator: In
                values:
                - ssd
      containers:
      - name: web
        image: nginx

Now any tier1 pod from this deployment will be placed only on nodes with the disktype=ssd label.

---

Summary Table

• nodeSelector

  • Description: Direct label match
  • Flexibility: Simple but rigid
  • Example Use: Perfect for small clusters with fixed rules

• nodeAffinity

  • Description: Advanced matching with operators
  • Flexibility: Highly flexible
  • Example Use: Ideal for dynamic environments with complex scheduling needs

---