Deployments

Background

A Kubernetes Deployments is a resource object that tells Kubernetes how to create or update instances of pods. Several use case of Deployments :

• Efficiently scale the number of pods using ReplicaSet
• Roll out the updated code in a controlled manner
• Roll back to an earlier deployment version if needed
• Ensure the desired state is maintained by starting new pods if pods fail or need to be updated

Details

Below is a Deployments sample manifest with key concept to note :

• we have 2 instances (pods) running, indicated by the .spec.replicas field
• the application will be configured under spec.template.spec.containers field

apiVersion: apps/v1
kind: Deployment
metadata:
  name: order-service
spec:
  replicas: 2
  selector:
    matchLabels:
      app: order-service
  template:
    metadata:
      labels:
        app: order-service
      annotations:
        prometheus.io/scrape: "true"
        prometheus.io/port: "9873"
        prometheus.io/path: "/actuator/prometheus"
    spec:
      containers:
      - name: order-service
        image: mariesto/order-service:latest
        ports:
        - containerPort: 9873
        env:
        - name: SPRING_DATASOURCE_URL
          value: jdbc:postgresql://host.docker.internal:5432/order-db
        - name: SPRING_DATASOURCE_USERNAME
          value: postgres
        - name: SPRING_DATASOURCE_PASSWORD
          value: pwd123
        resources:
          requests:
            memory: "512Mi"
            cpu: "500m"
          limits:
            memory: "1024Mi"
            cpu: "1000m"

Command

• Create a new deployment

kubectl apply -f [deploymentFileName].yaml

• Get deployment status

kubectl get deployments

• Get deployment details

kubectl describe deployments