🍆Azure Kubernetes Service

4-Dimensional Way

1. Kubernetes core concepts

2. Azure services integration with AKS clusters.

3. Azure DevOps

4. Terraform

Kubernetes core concepts

Two ways of Deployment

1. Imperative way using the kubectl - implement these pods,replica sets, deployments and then services

2. Declarative way using YAML - Using Live template writing deployment.

Azure services integrated with AKS

1. Azure Disks

2. Azure MySQL

3. Azure Files

4. Storage Account

5. Azure Public IP

6. Standard Load Balancer

7. Virtual Network

8. Subnets

9. Azure AD

10. AD Groups

11. AD users

12. Azure DNS Zones

13. Azure Container Registry

14. Azure DevOps

15. Azure Build & Release Pipelines

16. Azure Container Instances Virtual Nodes

17. Azure Kubernetes Service

18. AKS Linux & Windows Node Pools

19. Azure VM Scale sets

20. Azure Managed Service Identity

Azure DevOps

1. Build and push dock images to ACR

2. Deploy to Azure Kubernetes clusters

3. starter pipelines write from scratch

4. Release pipelines to deploy across Environments(Dev,QA)

Terraform

1. Terraform command basics

2. Terraform language basics

3. provisioning the production grade AKS cluster using Terraform from our local desktop.

4. provisioning AKS cluster using Terraform and then Azure DevOps

GitHub repositories

1. Azure AKS Kubernetes masterclass Main Repo - https://github.com/stacksimplify/azure-aks-kubernetes-masterclass

2. Azure DeOps for running Kubernetes workloads on Azure AKS Cluster - https://github.com/stacksimplify/azure-devops-github-acr-aks-app1

3. Provision Azure AKS cluster using Azure DevOps and Terraform - https://github.com/stacksimplify/azure-devops-aks-kubernetes-terraform-pipeline

4. Docker Fundamentals - https://github.com/stacksimplify/docker-fundamentals

5. Course Presentation with 250 slides outlining various architectures and designs - https://github.com/stacksimplify/azure-aks-kubernetes-masterclass/tree/master/ppt-presentation

Docker Fundamentals

Course-Pre-requisites

1. Azure Cloud Subscription

2. Github Repositories

Create Azure AKS Cluster

AKS - Introduction

• AKS is known as Azure Kubernetes Service

• AKS is highly available, secure and fully managed Kubernetes service

• It's available in 36 regions and compare than other CSP's aks is available in highest regions.

• we can run any type of workloads

  • Windows - .Net Apps

  • Linux - Java

  • IOT - Device deployment

  • Machine Learning Model Training with AKS

Kubernetes Architecture

• Any Kubernetes it will have a cluster control plan and worker nods.

• In Azure perspective will have an Azure AKS cluster control plan and Azure AKS node pools.

• Usually in regular Kubernetes, you call the node says worker nodes.

• AWS you call it as manager node groups.

• In AKS, means like in Azure, you're going to call them as node pools for the work nodes.

• cluster control plane is called as master component.

• In master component what all components are available

  • container runtime docker -

  • etcd - etcd is nothing but consistently and highly available. Key value store used as a Kubernetes backing store for all cluster data.

    • It stores all the master, and worker node information.

  • kube scheduler - It is responsible for distributing containers across multiple nodes, which is nothing but multiple worker nodes.

  • kube API server - one more component available. It acts as a front end for the Kubernetes control plane. It exposes the Kubernetes APA.

    • Command line tools, like kube CTL, or users, and even the master components like scheduler, controller manager, and etcd and worker node components like kubelet.

      Everything can talk to this kube APA server to perform the operations on the master,

  • kube control manager - controllers are responsible for noticing, and then responding when nodes, containers, or endpoints go down.

    • We'll have different controllers available.

      • Node controllers who are responsible for noticing and responding when worker nodes are down.

      • Replication controllers are responsible for maintaining the correct number of pods for every application.

      • Endpoint controllers, and then service account, and then token controllers.

  • worker nodes

    • The common thing here is container runtime

    • Container runtime is the underlying software where we run all the Kubernetes components.

    • We are using docker as container runtime.

  • kubelet

    • Kubelet is a heart of worker nodes.

    • kubelet is the agent that runs on every node in the cluster.

    • Agent is responsible for making sure that containers are running in a pod on a node and these will be always in constant communication with the kube scheduler.

    • kube scheduler, from master node, talks to worker nodes to the kubelet.

    • kube proxy - It is a network proxy that runs on each node in your cluster.It maintains the network rules on the nodes.

Create the AKs cluster