Pod affinity and anti-affinity in Kubernetes - Mini Project: Build & Apply

Pod Affinity and Anti-Affinity in Kubernetes

📖 Scenario: You are managing a Kubernetes cluster for a small online store. You want to control how pods are placed on nodes to improve reliability and performance.Specifically, you want some pods to be placed together (affinity) and some pods to be placed apart (anti-affinity).

🎯 Goal: Learn how to write pod specifications with affinity rules using podAffinity and podAntiAffinity in Kubernetes pod manifests.You will create a pod manifest with labels, then add affinity and anti-affinity rules, and finally check the pod spec output.

📋 What You'll Learn

Create a pod manifest with specific labels

Add pod affinity rules to schedule pods together

Add pod anti-affinity rules to avoid scheduling pods together

Print the final pod manifest YAML

💡 Why This Matters

🌍 Real World

Pod affinity and anti-affinity help control pod placement in Kubernetes clusters to improve availability and performance by grouping or separating pods.

💼 Career

Understanding pod affinity rules is important for Kubernetes administrators and DevOps engineers to optimize workload distribution and fault tolerance.

1

Create a basic pod manifest with labels

Create a YAML manifest for a pod named web-server in namespace default. The pod should have a container named nginx using image nginx:latest. Add labels app: web and tier: frontend under metadata.labels.

Kubernetes

# Write the pod manifest YAML with name, namespace, container, and labels

Need a hint?

Use metadata.labels to add labels. Containers go under spec.containers.

2

Add pod affinity to schedule pods together

Add a podAffinity rule under spec.affinity to require scheduling this pod on nodes where pods with label app: web are already running. Use requiredDuringSchedulingIgnoredDuringExecution with topologyKey: kubernetes.io/hostname.

Kubernetes

apiVersion: v1
kind: Pod
metadata:
  name: web-server
  namespace: default
  labels:
    app: web
    tier: frontend
spec:
  containers:
  - name: nginx
    image: nginx:latest
  # Add podAffinity rule below

Need a hint?

Put podAffinity under spec.affinity. Use requiredDuringSchedulingIgnoredDuringExecution with a list of rules.

3

Add pod anti-affinity to avoid scheduling pods together

Add a podAntiAffinity rule under spec.affinity to avoid scheduling this pod on nodes where pods with label tier: frontend are running. Use preferredDuringSchedulingIgnoredDuringExecution with weight 100 and topologyKey: kubernetes.io/hostname.

Kubernetes

apiVersion: v1
kind: Pod
metadata:
  name: web-server
  namespace: default
  labels:
    app: web
    tier: frontend
spec:
  affinity:
    podAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
      - labelSelector:
          matchLabels:
            app: web
        topologyKey: kubernetes.io/hostname
  # Add podAntiAffinity rule below
  containers:
  - name: nginx
    image: nginx:latest

Need a hint?

Put podAntiAffinity under spec.affinity. Use preferredDuringSchedulingIgnoredDuringExecution with weight and podAffinityTerm.

4

Print the final pod manifest YAML

Print the complete pod manifest YAML with the affinity and anti-affinity rules included.

Kubernetes

apiVersion: v1
kind: Pod
metadata:
  name: web-server
  namespace: default
  labels:
    app: web
    tier: frontend
spec:
  affinity:
    podAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
      - labelSelector:
          matchLabels:
            app: web
        topologyKey: kubernetes.io/hostname
    podAntiAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
      - weight: 100
        podAffinityTerm:
          labelSelector:
            matchLabels:
              tier: frontend
          topologyKey: kubernetes.io/hostname
  containers:
  - name: nginx
    image: nginx:latest

# Print the pod manifest YAML below

Need a hint?

Use a print statement to confirm the manifest is ready.