Workshop Exercise 1.2 - Adding Deployment Templates

Table of Contents

• Objective
• Step 1 - Create the templates folder
• Step 2 - Create a Storage Clain
• Step 3 - Add the deployment

Objective

• Understand definitions for deployments
• Create a template for deploying a pod

Now that we have a chart, we probably want to put something in it. Let’s start by uploading the basic deployment for our PLC service.

Step 1 - Create the templates folder

Navigate to the codesys folder in your repo, either using the web UI for gitea or your local check-out directory, and create a “templates” folder alongside the Chart.yaml file, if not already done. This folder will contain all the yaml manifest files that will be deployed as part of the HELM chart for our vPLC application. If using the UI, you can just add ‘/templates/’ before the next yaml file.gitea

Step 2 - Create a Storage Claim

In order to persist state as well as internal Application details, the PLC deployment is going to need an object called a persistentVolumeClaim. This maps directly to a folder within the PLC pod and allows a new container firing up, whether it’s a new container or a restart of one that crashed, to access the same data.

Kubernetes allows us to move these pods between different hosts and storage mechanisms, like Openshift Data Foundation used in this environment, allows access to the data on the host where our vPLC pod will be fired up. This is especially useful when redundancy is needed in the case of a node failure.

Since we will be using our gitea repo to contain all configurations and manifests that will be needed by our application, let’s start off by creating the manifest for our data storage claim.

Let’s head to the templates folder and add a definition for our data storage.

Notice that rather than creating the persistentVolumes directly, we will let Openshift Data Foundations create the persistent volume for us by using the “ocs-storagecluster-cephfs” storage class.

Underneath this will provide us with a filesystem that can be mounted within our container. More details on Openshift storage mechanisms can be found in our documentation

{{- range $.Values.plcs }}
---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: {{ .name }}-data
  labels:
    app.kubernetes.io/part-of: plc-application-{{ .name }}
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 1Gi
  storageClassName: ocs-storagecluster-cephfs
  volumeMode: Filesystem
{{- end }}

Let’s do the same for a config storage, create conf-storage.yaml in the templates directory.

{{- range $.Values.plcs }}
---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: {{ .name }}-config
  labels:
    app.kubernetes.io/part-of: {{ .partOf }}
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 1Gi
  storageClassName: ocs-storagecluster-cephfs
  volumeMode: Filesystem
{{- end }}

!Remember to commit and push this file to the gitea repo.

Step 3 - Add the deployment

For the Codesys application to run, we need somethig to actually do the processing (work). In order deploy a pod with the vPLC image, we will create an object in kubernetes called a Deployment

Some advantages of using a deployment vs just using a pod is that the deployment will try to ensure that we always have the specified number of instances of a pod running. For this exercise, we will only need a single replica, but can investigate the effects of the pod going down and automatically being rescheduled.

Add the deployment.yaml file to your group git repository’s templates folder:

{{- range $.Values.plcs }}
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: {{ .name }}
  labels:
    app.kubernetes.io/part-of: {{ .partOf }}
spec:
  selector:
    matchLabels:
      app: {{ .name }}
  replicas: 1
  template:
    metadata:
      labels:
        app: {{ .name }}
    spec:
      hostname: {{ .name }}
      volumes:
        - name: data-storage
          persistentVolumeClaim: 
            claimName: {{ .name }}-data
        - name: config
          persistentVolumeCLaim:
            claimName: {{ .name }}-config
      containers:
        - name: codesys-plc
          image: quay.io/rh-ee-hvanniek/codesyscontrol:4.14.0.0
          ports:
            - name: opc-ua
              containerPort: 4840
              protocol: TCP
          volumeMounts:
            - mountPath: "/conf/codesyscontrol/"
              name: config
            - mountPath: "/data/codesyscontrol"
              name: data-storage

{{- end }}

!Remember to commit and push this file to the gitea repo.

With these templates created we are ready to let ArgoCD start doing its magic.

---

Navigation

Previous Exercise

Next Exercise

Click here to return to the Workshop Homepage