How to pass multiple docker images through values.yml to template.yml in Helm

Question

I am trying to run a application that has around 40 microservices. How to pass 40 different docker images from values.yml file to template.yml file.

template file

      name:{{ .values.name }}
spec:
   containers:
      - image: {{ .values.container.image }}

values file

name:A
 container:
     image:A
name :B
  container :
       image:B

i have 40 more docker images like that, how to pass all those images to template. And will passing like that creates 40 different pods?, because we would need 40 different pods. Any guidance is highly appreciated.

Answer 1

Focusing only on images and templates you can create a helm template that will spawn an X amount of pods by:

• Creating a Chart.yaml file
• Creating a values.yaml file with the variable that store all image names
• Creating a template with a {{ range }} directive
• Testing

Below is the structure of files and directories:

❯ tree helm-dir
helm-dir
├── Chart.yaml
├── templates
│   └── pod.yaml
└── values.yaml

1 directory, 3 files

---

Create Chart.yaml file

Below is the Chart.yaml file:

apiVersion: v2
name: helm-templates
description: A Helm chart for spawning pods from images
version: 0.1.0

---

Create a values.yaml file with the variable that store all image names

Below is the simple values.yaml file with different images name that will be used with a template:

different_images: 
  - ubuntu 
  - nginx

---

Create a template that with a {{ range }} directive

This template is stored in templates directory with a name pod.yaml

Below YAML definition will be a template for all the pods:

{{- range .Values.different_images }}
apiVersion: v1
kind: Pod
metadata:
  name: {{ . }} 
  labels:
    app: {{ . }} 
spec:
  restartPolicy: Never
  containers:
  - name: {{ . }} 
    image: {{ . }} 
    imagePullPolicy: Always
    command: 
    - sleep 
    - infinity
---
{{- end }} 

{{- range .Values.different_images }} will iterate over different_images variable and replace the {{ . }} with an image name.

---

Test

Run below command from helm-dir directory to check if helm YAML definitions of pods are correctly created:

$ helm install NAME . --dry-run --debug

You should get an output with multiple pods definition that look similar to one below:

# Source: helm-templates/templates/pod.yaml
apiVersion: v1
kind: Pod
metadata:
  name: ubuntu
  labels:
    app: ubuntu
spec:
  restartPolicy: Never
  containers:
  - name: ubuntu
    ports:
      - containerPort: 3000
    image: ubuntu
    imagePullPolicy: Always
    command: 
    - sleep 
    - infinity
    resources:
      requests:
        memory: 500Mi
        cpu: 500m 

You can now run: $ helm install NAME .

and check if pods spawned correctly with $ kubectl get pods:

NAME     READY   STATUS    RESTARTS   AGE
nginx    1/1     Running   0          8s
ubuntu   1/1     Running   0          8s

Please take a look on additional resources:

• Helm.sh: Chart best practices
• Helm.sh: Chart template guide: Control structures
• Stackoverflow.com: How can I iteratively create pods from list using helm