Cloud-Native Microservices With Kubernetes - 2nd Edition

Creating a Pod Disruption Budget

To better understand how to create a PDB for your application, let's consider the following example:

We have a critical application deployed as a Deployment named hello-world with 5 replicas.

kubectl apply -f - <
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: hello-world
spec:
  replicas: 5
  selector:
    matchLabels:
      app: hello-world
  template:
    metadata:
      labels:
        app: hello-world
    spec:
      containers:
      - name: hello-world
        image: gcr.io/google-samples/hello-app:1.0
        ports:
        - containerPort: 8080
EOF

After many tests and considerations, you have decided that during disruptions, 3 replicas should always be available. To enforce this rule, the PDB can be created using the following command:

kubectl create pdb hello-world --selector app=hello-world --min-available 3

A PDB, like ClusterIP, uses labels to select the Pods to which it applies. In this case, we used the label app=hello-world to select all Pods of the Deployment hello-world. It also accepts two parameters to define the availability rules: --min-available and --max-unavailable. In this case, we used --min-available 3, which means that at least 3 Pods must be available at all times.

You can verify that it was created successfully by running the following command: