Terraform | Coding With Taz

I’ve been using a container for running Terraform for a while but just for local development. More recently though the need to share modules has become more prevalent.

One solution for this is to use a container to not only share modules for development but for deployment as well. This also allows the containers to be versioned, limiting breaking changes affecting multiple pipelines at once.

In this post I am going to cover:

Building a container with shared terraform modules
Pushing the built container to Azure Container Registry
Configuring the dev environment to use the built container
Deploy infrastructure using the built container

NOTE: All of the code used here can be found on my GitHub including the shared modules.

Prerequisites

For this post I will running on Windows and using the following programs:

Building the Container

The container needs to not only have what is needed for development but what is needed to run as a container job in Azure Pipelines e.g. Node. The Microsoft Docs provide more detail about this.

The container is an Alpine Linux base with Node, PowerShell Core, Azure CLI and Terraform installed.

Dockerfile

ARG IMAGE_REPO=alpine
ARG IMAGE_VERSION=3
ARG TERRAFORM_VERSION
ARG POWERSHELL_VERSION
ARG NODE_VERSION=lts-alpine3.14

FROM node:${NODE_VERSION} AS node_base
RUN echo "NODE Version:" && node --version
RUN echo "NPM Version:" && npm --version

FROM ${IMAGE_REPO}:${IMAGE_VERSION} AS installer-env
ARG TERRAFORM_VERSION
ARG POWERSHELL_VERSION
ARG POWERSHELL_PACKAGE=powershell-${POWERSHELL_VERSION}-linux-alpine-x64.tar.gz
ARG POWERSHELL_DOWNLOAD_PACKAGE=powershell.tar.gz
ARG POWERSHELL_URL=https://github.com/PowerShell/PowerShell/releases/download/v${POWERSHELL_VERSION}/${POWERSHELL_PACKAGE}
RUN apk upgrade --update && \
    apk add --no-cache bash wget curl

# Terraform
RUN wget --quiet https://releases.hashicorp.com/terraform/${TERRAFORM_VERSION}/terraform_${TERRAFORM_VERSION}_linux_amd64.zip && \
    unzip terraform_${TERRAFORM_VERSION}_linux_amd64.zip && \
    mv terraform /usr/bin
    
# PowerShell Core
RUN curl -s -L ${POWERSHELL_URL} -o /tmp/${POWERSHELL_DOWNLOAD_PACKAGE}&& \
    mkdir -p /opt/microsoft/powershell/7 && \
    tar zxf /tmp/${POWERSHELL_DOWNLOAD_PACKAGE} -C /opt/microsoft/powershell/7 && \
    chmod +x /opt/microsoft/powershell/7/pwsh 

FROM ${IMAGE_REPO}:${IMAGE_VERSION} 
ENV NODE_HOME /usr/local/bin/node
# Copy only the files we need from the previous stages
COPY --from=installer-env ["/usr/bin/terraform", "/usr/bin/terraform"]
COPY --from=installer-env ["/opt/microsoft/powershell/7", "/opt/microsoft/powershell/7"]
RUN ln -s /opt/microsoft/powershell/7/pwsh /usr/bin/pwsh
COPY --from=node_base ["${NODE_HOME}", "${NODE_HOME}"]

# Copy over Modules
RUN mkdir modules
COPY modules modules

LABEL maintainer="Coding With Taz"
LABEL "com.azure.dev.pipelines.agent.handler.node.path"="${NODE_HOME}"

ENV APK_DEV "gcc libffi-dev musl-dev openssl-dev python3-dev make"
ENV APK_ADD "bash sudo shadow curl py3-pip graphviz git"
ENV APK_POWERSHELL="ca-certificates less ncurses-terminfo-base krb5-libs libgcc libintl libssl1.1 libstdc++ tzdata userspace-rcu zlib icu-libs"
# Install additional packages
RUN apk upgrade --update && \
    apk add --no-cache --virtual .pipeline-deps readline linux-pam && \
    apk add --no-cache --virtual .build ${APK_DEV} && \
    apk add --no-cache ${APK_ADD} ${APK_POWERSHELL} && \
    # Install Azure CLI
    pip --no-cache-dir install --upgrade pip && \
    pip --no-cache-dir install wheel && \
    pip --no-cache-dir install azure-cli && \
    apk del .build && \
    apk del .pipeline-deps 

RUN echo "PS1='\n\[\033[01;35m\][\[\033[0m\]Terraform\[\033[01;35m\]]\[\033[0m\]\n\[\033[01;35m\][\[\033[0m\]\[\033[01;32m\]\w\[\033[0m\]\[\033[01;35m\]]\[\033[0m\]\n \[\033[01;33m\]->\[\033[0m\] '" >> ~/.bashrc 
CMD tail -f /dev/null

The container can be built locally using the docker build command and providing the PowerShell and Terraform versions e.g.

docker build --build-arg TERRAFORM_VERSION="1.0.10" --build-arg POWERSHELL_VERSION="7.1.5" -t my-terraform .

Pushing the container to Azure Container Registry

Next thing to do is to build the container and push it to the Azure Container Registry (if you need to know how to set that up in Azure DevOps see my previous post on Configuring ACR). In this pipeline I have also added a Snyk scan to check for vulnerabilities in my container (happy to report there wasn’t any at the time of writing). If you are not familiar with Snyk I recommend you check out their website.

For the build number I have used the version of Terraform and then the date and revision but you can use whatever makes sense for example you could use Semver.

I also setup some pipeline variables for the container registry connection and the container registry name e.g. <your registry>.azurecr.io

trigger: 
    - main

pr: none

name: $(terraformVersion)_$(Date:yyyyMMdd)$(Rev:.r)

variables:
 dockerFilePath: dockerfile
 imageRepository: iac/terraform
 terraformVersion: 1.0.10
 powershellVersion: 7.1.5

pool:
  vmImage: "ubuntu-latest"

steps:
  - task: Docker@2
    displayName: "Build Terraform Image"
    inputs:
      containerRegistry: '$(containerRegistryConnection)'
      repository: '$(imageRepository)'
      command: 'build'
      Dockerfile: '$(dockerfilePath)'
      arguments: '--build-arg TERRAFORM_VERSION="$(terraformVersion)" --build-arg POWERSHELL_VERSION="$(powershellVersion)"'
      tags: | 
        $(Build.BuildNumber)
  - task: SnykSecurityScan@1
    inputs:
      serviceConnectionEndpoint: 'Snyk'
      testType: 'container'
      dockerImageName: '$(containerRegistry)/$(imageRepository):$(Build.BuildNumber)'
      dockerfilePath: '$(dockerfilePath)'
      monitorWhen: 'always'
      severityThreshold: 'high'
      failOnIssues: true
  - task: Docker@2
    displayName: "Build and Push Terraform Image"
    inputs:
      containerRegistry: '$(containerRegistryConnection)'
      repository: '$(imageRepository)'
      command: 'Push'
      Dockerfile: '$(dockerfilePath)'
      tags: | 
        $(Build.BuildNumber)

Once the container is built it can be viewed in the Azure Portal inside your Azure Container Registry.

Configuring the Dev Environment

Now the container has been created and pushed to the Azure Container Registry the next job is to configure Visual Studio Code.

To start with we need to make sure the extension Remote Containers is installed in Visual Studio Code

In the project where you want to use the container, create a folder called .devcontainer and then a file inside the folder called devcontainer.json and add the following (updating the container registry and container details e.g. name, version, etc.)

// For format details, see https://aka.ms/devcontainer.json. For config options, see the README at:
// https://github.com/microsoft/vscode-dev-containers/tree/v0.205.1/containers/docker-existing-dockerfile
{
	"name": "Terraform Dev",

	// Sets the run context to one level up instead of the .devcontainer folder.
	"context": "..",

	// Update the 'dockerFile' property if you aren't using the standard 'Dockerfile' filename.
	"image": "<your container registry>.azurecr.io/iac/terraform:1.0.10_20211108.1",

	// Set *default* container specific settings.json values on container create.
	"settings": {},
	
	// Add the IDs of extensions you want installed when the container is created.
	"extensions": [
		"ms-vscode.azure-account",
		"ms-azuretools.vscode-azureterraform",
		"hashicorp.terraform",
		"ms-azure-devops.azure-pipelines"
	]
}

NOTE: You may notice that there is a number of extensions in the above config. I use these extensions in Visual Studio Code for Terraform, Azure Pipelines, etc. and therefore they would also need installing in order to make use of them in the container environment.

TIP: If you right-click on an extension in Visual Studio Code and select ‘Copy Extension ID’ you can easily get the extension information you need to add other extensions to the list.

Now, make sure to login to the Azure Container Registry (either in another window or the terminal in Visual Studio Code) with the Azure CLI for authentication e.g.

az acr login -n <your container registry name>

This needs to be done to be able to pull down the container. Once the login is successful, select the icon in the bottom left of Visual Studio Code to ‘Open a Remote Window’

Then select ‘Reopen in Container’ this will download the container from the Azure Container Registry and load up the project in the container (this can take a minute or so first time).

Once the project is loaded you can create Terraform files as normal and take advantage of the shared modules inside the container.

So lets create a small example. I work a lot in Azure so I am using a shared module to create an Azure Function and another module to format the naming convention for the resources.

terraform {
  required_providers {
    azurerm = {
      source  = "hashicorp/azurerm"
      version = "~> 2.83"
    }
  }
  backend "local" {}
  required_version = ">= 1.0.10"
}

provider "azurerm" {
  features {}
}


module "rgname" {
    source        = "/modules/naming"
    name          = "myapp"
    env           = "rg-${var.env}"
    resource_type = ""
    location      = var.location
    separator     = "-"
}

resource "azurerm_resource_group" "rg" {
  name     = module.rgname.result
  location = "uksouth"
}

module "funcApp" {
  source                    = "/modules/linux_azure_function"
  resource_group            = azurerm_resource_group.rg.name
  resource_group_location   = azurerm_resource_group.rg.location
  env                       = var.env
  appName                   = var.appName
  funcWorkerRuntime         = "dotnet-isolated"
  dotnetVersion             = "v5.0"
  additionalFuncAppSettings = {
    mysetting = "somevalue"
  }
  tags                      = var.tags
}

From the terminal window I can now authenticate to Azure by logging in via the CLI

az login

Then I can run the terraform commands

terraform init
terraform plan

This produces the terraform plan for the resources that would be created.

Deploy Infrastructure Using the Container

So now I have created a new terraform configuration its time to deploy the changes using the same container.

To do this I am using Azure Pipelines YAML. There are several parts to the pipeline, firstly, in order to store the state for the pipeline there needs to be an Azure Storage Account to store the state file. I like to add this to the pipeline using Azure CLI so that the account is created if it doesn’t exist but also updates it if there are changes.

 - task: AzureCLI@2
    displayName: 'Create/Update State File Storage'
    inputs:
        azureSubscription: '$(subscription)'
        scriptType: bash
        scriptLocation: inlineScript
        inlineScript: |
          az group create --location $(location) --name $(terraformGroup)
          az storage account create --name $(terraformStorageName) --resource-group $(terraformGroup) --location $(location) --sku $(terraformStorageSku) --min-tls-version TLS1_2 --https-only true --allow-blob-public-access false
          az storage container create --name $(terraformContainerName) --account-name $(terraformStorageName)
        addSpnToEnvironment: false

The terraform backend configuration is set to local for development and so I need a step in the pipeline to update it to use backend “azurerm”.

  - bash: |
      sed -i 's/backend "local" {}/backend "azurerm" {}/g' main.tf
    displayName: 'Update Backend in terraform file'

For the Terraform commands I tend to use the Microsoft Terraform Tasks with additional command options for the plan file

- task: TerraformTaskV2@2
    displayName: 'Terraform Init'
    inputs:
      backendServiceArm: '$(subscription)'
      backendAzureRmResourceGroupName: '$(terraformGroup)'
      backendAzureRmStorageAccountName: '$(terraformStorageName)'
      backendAzureRmContainerName: '$(terraformContainerName)'
      backendAzureRmKey: '$(terraformStateFilename)'
  - task: TerraformTaskV2@2
    displayName: 'Terraform Plan'
    inputs:
      command: plan
      commandOptions: '-out=tfplan'
      environmentServiceNameAzureRM: '$(subscription)'
  - task: TerraformTaskV2@2
    displayName: 'Terraform Apply'
    inputs:
      command: apply
      commandOptions: '-auto-approve tfplan'
      environmentServiceNameAzureRM: '$(subscription)'

So, putting it all together the whole pipeline looks likes this:

trigger:
   - main
 
pr: none
parameters:
  - name: env
    displayName: 'Environment'
    type: string
    default: 'dev'
    values:
      - dev
      - test
      - prod
  - name: location
    displayName: 'Resource Location'
    type: string
    default: 'uksouth'
  - name: appName
    displayName: 'Application Name'
    type: string
    default: 'myapp'
  - name: tags 
    displayName: 'Tags'
    type: object 
    default: 
     Environment: "dev"
     Project: "Demo"
variables:
  isMain: $[eq(variables['Build.SourceBranch'], 'refs/heads/main')]
  location: 'uksouth'
  terraformGroup: 'rg-dev-terraform-uksouth'
  terraformStorageName: 'devterraformuksouth2329'
  terraformStorageSku: 'Standard_LRS'
  terraformContainerName: 'infrastructure'
  terraformStateFilename: 'deploy.tfstate'
  
jobs:
- job: infrastructure
  displayName: 'Build Infrastructure'
  pool:
    vmImage: ubuntu-latest
  container:
    image: $(containerRegistry)/iac/terraform:1.0.10_20211108.1
    endpoint: 'ACR Connection'
  steps:
  - task: AzureCLI@2
    displayName: 'Create/Update State File Storage'
    inputs:
        azureSubscription: '$(subscription)'
        scriptType: bash
        scriptLocation: inlineScript
        inlineScript: |
          az group create --location $(location) --name $(terraformGroup)
          az storage account create --name $(terraformStorageName) --resource-group $(terraformGroup) --location $(location) --sku $(terraformStorageSku) --min-tls-version TLS1_2 --https-only true --allow-blob-public-access false
          az storage container create --name $(terraformContainerName) --account-name $(terraformStorageName)
        addSpnToEnvironment: false
  - bash: |
      sed -i 's/backend "local" {}/backend "azurerm" {}/g' main.tf
    displayName: 'Update Backend in terraform file'
  - template: 'autovars.yml'
    parameters:
      env: ${{ parameters.env }}
      location: ${{ parameters.location }}
      appName: ${{ parameters.appName }}
      tags: ${{ parameters.tags }}
  - task: TerraformTaskV2@2
    displayName: 'Terraform Init'
    inputs:
      backendServiceArm: '$(subscription)'
      backendAzureRmResourceGroupName: '$(terraformGroup)'
      backendAzureRmStorageAccountName: '$(terraformStorageName)'
      backendAzureRmContainerName: '$(terraformContainerName)'
      backendAzureRmKey: '$(terraformStateFilename)'
  - task: TerraformTaskV2@2
    displayName: 'Terraform Plan'
    inputs:
      command: plan
      commandOptions: '-out=tfplan'
      environmentServiceNameAzureRM: '$(subscription)'
  - task: TerraformTaskV2@2
    displayName: 'Terraform Apply'
    inputs:
      command: apply
      commandOptions: '-auto-approve tfplan'
      environmentServiceNameAzureRM: '$(subscription)'

As with the container build pipeline I used some pipeline variables here for the subscription connection and the container registry e.g. <your registry>.azurecr.io

After the pipeline ran, a quick check in the Azure Portal shows the resources were created as expected

Final Thoughts

I really like using containers for local development and with the remote containers extension for Visual Studio Code its great to be able to run from within a container and share code in this way. I am sure that other things could be shared using this method too.

Being able to version the containers and isolate breaking changes across multiple pipelines is also a bonus. I expect this process could be better, maybe even include pinning of provider versions in Terraform, etc. but its a good start.

I’ve been using Terraform for a while with Azure Pipelines and have always passed the pipeline parameters or variables to Terraform using the -var command line parameter. This has worked really well until I encountered a need to pass more complex objects into Terraform which supports objects, maps and lists.

The Problem

When attempting to pass complex objects into -var Azure Pipelines outputs errors like ‘object is not a string’. After trying a number of work arounds that failed I ended up changing my Terraform to take strings and then perform actions on them e.g. an array as string then using split in Terraform to re-create the array.

This lead me to thinking “There has to be a better way”. Naturally one option is to create a .tfvars.json file and then substitute the variables using the same technique I used in my previous article Azure Pipelines – Parameters + JSON File Substitution. This would work for the most part but would not solve using array types.

I started thinking, could I get a parameter into a script and then somehow workout if it was a complex object and then write code to extract the value into something useful like JSON. This lead me to a community post that mentioned a function convertToJson.

A Solution

Based on using convertToJson and combing the technique from my previous article I came up with a step to create a HCL formatted .auto.tfvars file. The only thing is that for objects the colons ‘:’ need converting to equals ‘=’.

- ${{ each item in parameters }}:
     - script: echo '${{ item.key }}=${{ replace(convertToJson(item.value), ':', '=')}}' >> parameters.auto.tfvars
       displayName: "JsonVar ${{ item.key }}"

The .auto.tfvars file is automatically loaded by Terraform which removes the need to specify any -var or -var-file options.

Example Pipeline

For my example pipeline I have used an object for Tags and an array for a list of Network addresses for use with a Network Security Group.
The initial pipeline setups up the parameters and the Azure Storage Account for my Terraform state files.

trigger: none
pr: none
parameters:
  - name: resourceGroup
    displayName: Resource Group
    type: string
    default: 'terraform-test'
  - name: resourceLocation
    displayName: Resource Location
    type: string
    default: 'uksouth'
  - name: projectName
    displayName: Project Tag Name
    type: string
    default: 'Demo'
  - name: tags
    type: object
    default:
      Project: "Demo"
      Environment: "Dev"
  - name: network_source_addresses
    displayName: Network Address List
    type: object
    default:
      - "192.168.1.20"
      - "192.168.1.254"
variables:
  subscription: 'My Subscription'
  terraformVersion: '0.14.6'
  terraformResourceGroup: 'test-deployment'
  terraformStorageName: 'demoterraformstore'
  terrformStorageSku: Standard_LRS
  terraformContainerName: 'terraform'
  terraformStateFilename: test.tfstate
pool:
  vmImage: "ubuntu-latest"
steps:
- task: AzureCLI@2
  displayName: 'Azure CLI'
  inputs:
    azureSubscription: $(subscription)
    scriptType: bash
    scriptLocation: inlineScript
    inlineScript: |
      az group create --location ${{ parameters.resourceLocation }} --name $(terraformResourceGroup)
      az storage account create --name $(terraformStorageName) --resource-group $(terraformResourceGroup) --location ${{ parameters.resourceLocation }} --sku $(terrformStorageSku) --tags "project=${{ parameters.projectName }}"
      az storage container create --name $(terraformContainerName) --account-name $(terraformStorageName)
    addSpnToEnvironment: false
- template: deploy.yml
  parameters:
    resourceGroup: ${{ parameters.resourceGroup }}
    resourceLocation: ${{ parameters.resourceLocation }}
    tags: ${{ parameters.tags }}
    network_source_addresses: ${{ parameters.network_source_addresses }}
    secret_value: $(secret_value)

I separated the Terraform parts into a template so that the loop only uses the parameters that are needed for Terraform and not any others that are in the main pipeline.
Note: I use the Microsoft Terraform Azure Pipelines Extension to deploy the Terraform scripts.

parameters:
  - name: resourceGroup
    type: string
  - name: resourceLocation
    type: string
  - name: tags
    type: object
  - name: network_source_addresses
    type: object
  - name: secret_value
    type: string
steps:
- ${{ each item in parameters }}:
     - script: echo '${{ item.key }}=${{ replace(convertToJson(item.value), ':', '=')}}' >> parameters.auto.tfvars
       displayName: "JsonVar ${{ item.key }}"
- bash: |
    cat parameters.auto.tfvars
  displayName: "Debug show new file"
- task: TerraformInstaller@0
  displayName: 'Install Terraform'
  inputs:
    terraformVersion: $(terraformVersion)
  
- task: TerraformTaskV1@0
  displayName: 'Terraform Init'
  inputs:
    backendServiceArm: $(subscription)
    backendAzureRmResourceGroupName: '$(terraformResourceGroup)'
    backendAzureRmStorageAccountName: '$(terraformStorageName)'
    backendAzureRmContainerName: '$(terraformContainerName)'
    backendAzureRmKey:  '$(terraformStateFilename)'
- task: TerraformTaskV1@0
  displayName: 'Terraform Plan'
  inputs:
    command: plan    
    environmentServiceNameAzureRM: $(subscription)
- task: TerraformTaskV1@0
  displayName: 'Terraform Apply'
  inputs:
    command: apply
    commandOptions: -auto-approve
    environmentServiceNameAzureRM: $(subscription)

Conclusion

I think this is a nice technique for using complex types in Azure Pipelines for use with Terraform deployments or anything else that would benefit from this idea. This was a fun problem to try and solve and I hope that sharing this helps others who have encountered the same problem.

Coding With Taz