🐋 Kubernetes interpolation

Here’s a cool little interpolation trick I found on the internet while trying to figure out how to build a new environment variable from existing ones with Kubernetes:

Let’s say you have a ConfigMap:

configmap.yaml
yaml
apiVersion: v1
kind: ConfigMap
metadata:
  name: some-configmap
data:
  DB_NAME: "app-db"
  DB_PORT: "5432"
  DB_HOST: "app-db-svc"

And a secret because we want to keep things safe:

secret.yaml
yaml
apiVersion: v1
kind: Secret
metadata:
  name: some-secret
data:
  DB_PASSWORD: aHVudGVyMg== # hunter2
  DB_USER: YXBwLXVzZXI= # app-user
type: Opaque

Now you can render them

deployment.yaml
yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: sick-app
  name: sick-app
spec:
  replicas: 3
  selector:
    matchLabels:
      app: sick-app
  template:
    metadata:
      labels:
        app: sick-app
    spec:
      containers:
        - image: your-registry.io/your-repo/sick-app
          name: app
          ports:
            - containerPort: 80
          envFrom:
            - configMapRef:
                name: some-configmap
            - secretRef:
                name: some-secret
          env:
            # Let's say you're using Prisma
            # You want a DATABASE_URL like this:
            # postgresql://app-user:hunter2@app-db-svc:5432/app-db
            - name: DATABASE_URL
              value: postgresql://$(DB_USER):$(DB_PASSWORD)@$(DB_HOST):$(DB_PORT)/$(DB_NAME)
          resources: {}

Now you don’t have to define a new secret!