Azure for AWS engineers · Lesson 7 of 9 · 3 min read
Infrastructure as code and pipelines
ARM templates and Bicep
Enlarge diagram
Swipe sideways to see all of it, or tap to open it full screen.
ARM templates are JSON (≈ CloudFormation). Bicep is a cleaner DSL that compiles to ARM — Microsoft's recommended native IaC:
param location string = resourceGroup().location
resource sa 'Microsoft.Storage/storageAccounts@2023-05-01' = {
name: 'stshopprodlogs'
location: location
sku: { name: 'Standard_ZRS' }
kind: 'StorageV2'
properties: {
minimumTlsVersion: 'TLS1_2'
allowBlobPublicAccess: false
}
}
az deployment group create -g rg-shop-prod --template-file main.bicep
az deployment group what-if -g rg-shop-prod --template-file main.bicep # ≈ change sets / plan
Terraform on Azure
Your Terraform skills transfer; only the provider and resources change:
terraform {
required_providers { azurerm = { source = "hashicorp/azurerm", version = "~> 4.0" } }
backend "azurerm" {
resource_group_name = "rg-tfstate"
storage_account_name = "sttfstateacme"
container_name = "tfstate"
key = "prod/aks.tfstate"
use_azuread_auth = true
}
}
provider "azurerm" {
features {}
subscription_id = var.subscription_id
}
resource "azurerm_resource_group" "aks" {
name = "rg-shop-prod"
location = "centralindia"
}
resource "azurerm_kubernetes_cluster" "shop" {
name = "aks-shop-prod"
location = azurerm_resource_group.aks.location
resource_group_name = azurerm_resource_group.aks.name
dns_prefix = "shop-prod"
sku_tier = "Standard"
oidc_issuer_enabled = true
workload_identity_enabled = true
default_node_pool {
name = "system"
vm_size = "Standard_D4s_v5"
node_count = 3
zones = ["1", "2", "3"]
}
identity { type = "SystemAssigned" }
network_profile {
network_plugin = "azure"
network_plugin_mode = "overlay"
}
}
- State in an Azure Storage blob container, locked with blob leases (≈ S3 + lockfile/DynamoDB).
- Authenticate CI with OIDC/workload identity federation — no client secrets.
- Secrets for Terraform from Key Vault via data sources or pipeline variable groups.
- For multi-cloud repos: provider-specific modules, shared conventions (naming, tagging), and separate state per cloud and environment.
Azure DevOps Pipelines
# azure-pipelines.yml
trigger: { branches: { include: [main] } }
pool: { vmImage: ubuntu-latest }
variables:
- group: shop-prod-secrets # variable group linked to Key Vault, values marked secret
stages:
- stage: Build
jobs:
- job: build
steps:
- task: Maven@4
inputs: { mavenPomFile: pom.xml, goals: package }
- task: Docker@2
inputs:
containerRegistry: acr-shop-prod # service connection
repository: orders
command: buildAndPush
tags: $(Build.SourceVersion)
- stage: Deploy
dependsOn: Build
jobs:
- deployment: prod
environment: shop-prod # approvals and checks configured on the environment
strategy:
runOnce:
deploy:
steps:
- task: AzureCLI@2
inputs:
azureSubscription: sc-shop-prod # service connection using workload identity federation
scriptType: bash
scriptLocation: inlineScript
inlineScript: az aks get-credentials -g rg-shop-prod -n aks-shop-prod && kubectl get ns
- Service connections authenticate to Azure — prefer workload identity federation over secrets.
- Variable groups linked to Key Vault supply secrets, marked secret (masked in logs).
- Environments with approvals and checks ≈ GitHub environment protection / Jenkins
input. - GitHub Actions to Azure:
azure/loginwith OIDC (federated credential on an app registration or managed identity).