Creating a simple deployment in Azure Kubernetes Service using YAML (Less than 5 min)

Declarative approach...

November 2, 2021

Chendrayan Venkatesan

3-Minute Read

Introduction

In my previous blog posts, we have seen imperative approaches for creating pods, deployments, replica sets and exposing them to load balancer service. The links below are for your reference.

YAML Basics

YAML is a human-readable data serialization language used for configurations and many other purposes. YAML stands for Yet Another Markup Language. If you are familiar with Python, you can correlate the indentation. Yes, YAML follows the same indentation. Below is the different data formats used in the YAML

YAML is case sensitive.

Simple Key-Value pairs

name: Chen
department: IT
city: Bangalore

Dictionary

person:
  name: Chen
  department: IT
  city: Bangalore

Array

skills:
  - Azure
  - IT Automation
  - PowerShell
  - Python

More than one list

employees:
  - name: Chen
    city: Bangalore
  - name: Shashi
    city: Bangalore

Project Folder Structure

📦Collabrains.Cloud
┣ 📂database
┃ ┗ 📜employee.json
┣ 📂manifest
┃ ┣ 📜create-deployment.yaml
┃ ┗ 📜create-service.yaml
┣ 📜Dockerfile
┗ 📜employee-rest-api.ps1

Deploy a REST API in AKS! To begin with, let us follow a simple project structure. The employee.json file underneath the database folder is the source of employee information.

{
   "employee": [
       {
           "Id": "1",
           "FirstName": "Chendrayan",
           "SurName": "Venkatesan",
           "Country": "India",
           "City": "Bangalore",
           "DateOfJoining": "June 16, 2008"
       },
       {
           "Id": "2",
           "FirstName": "Shahsi",
           "SurName": "Shetty",
           "Country": "India",
           "City": "Bangalore",
           "DateOfJoining": "January 01, 2010"
       },
       {
           "Id": "3",
           "FirstName": "Matt",
           "SurName": "Hans",
           "Country": "Nederlands",
           "City": "Amsterdam",
           "DateOfJoining": "March 03, 2005"
       }
   ]
}

The main file is employee-rest-api.ps1, and this code hosts a REST API to serve other applications.

Start-PodeServer {
    Add-PodeEndpoint -Address * -Port 3000 -Protocol Http
    
    Add-PodeRoute -Method Get -Path '/employee' -ScriptBlock {             
        $Employee = Get-Content .\database\employee.json -Raw
        Write-PodeJsonResponse -Value $($Employee)
    }

    Add-PodeRoute -Method Get -Path '/employee/:id' -ScriptBlock {             
        $Employee = ((Get-Content .\database\employee.json -Raw  | ConvertFrom-Json).employee).Where(
            {
                ($_.Id -eq $WebEvent.Parameters['id'])
            }
        )
        Write-PodeJsonResponse -Value $($Employee)
    }
}

Dockerize the application (Dockerfile)

FROM mcr.microsoft.com/powershell:latest

WORKDIR /usr/src/app/

COPY . .    

RUN pwsh -c 'Install-Module Pode -force'

CMD [ "pwsh", "-c", "cd /usr/src/app; ./employee-rest-api.ps1" ]

Build, tag & Push the docker image

PS C:\AKS-Learning\Collabrains.Cloud> docker build -t chenv/employee-rest-api:v1.0.0 .

PS C:\AKS-Learning\Collabrains.Cloud> docker push chenv/employee-rest-api:v1.0.0

Deployment Definition (create-deployment.yaml)

Required Fields : apiVersion, kind, metadata, spec

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp
spec:
  selector:
    matchLabels:
      app: myapp
  template:
    metadata:
      labels:
        app: myapp
    spec:
      containers:
      - name: myapp
        image: chenv/ps-web-app:v1.0.0
        resources:
          limits:
            memory: "128Mi"
            cpu: "500m"
        ports:
            - containerPort: 80

Service Definition (create-service.yaml)

Required Fields : apiVersion, kind, metadata, spec

apiVersion: v1
kind: Service
metadata:
  name: deployment-loadbalancer-service
spec:
  type: LoadBalancer
  selector:
    app: webfrontend
  ports:
    - name: http
      port: 80
      targetPort: 3000

Deployment & Expose

PS C:\> kubectl apply -f .\manifest\create-deployment.yaml

PS C:\> kubectl apply -f .\manifest\create-service.yaml

References

Credits

dfinke - PowerShell Microservice is very helpful.

Summary

Awesome, now that we know how to do deployments using YAML (declaratively). There are lot more in AKS..,. Please feel free to subscribe to my YouTube channel - iAutomate and follow me on twitter ChendrayanV

Say Something

Comments

Nothing yet.