ARM vs Bicep vs Terraform decision in Azure - Visual Side-by-Side Comparison

Process Flow - ARM vs Bicep vs Terraform decision

Start: Need to deploy Azure infrastructure

↓

Choose tool: ARM, Bicep, or Terraform?

↓

ARM Template

↓

Write JSON

↓

Deploy via Azure

↓

Infrastructure created in Azure

↓

End

This flow shows the decision process for choosing ARM, Bicep, or Terraform to deploy Azure infrastructure, and how each tool works to create resources.

Execution Sample

Azure

resource storageAccount 'Microsoft.Storage/storageAccounts@2021-04-01' = {
  name: 'mystorage'
  location: 'eastus'
  kind: 'StorageV2'
  sku: {
    name: 'Standard_LRS'
  }
}

This Bicep code defines a storage account resource in Azure with basic properties.

Process Table

Step	Tool	Action	Input	Output	Notes
1	Decision	Choose deployment tool	Need Azure infra	Select ARM, Bicep, or Terraform	Start of process
2	ARM	Write JSON template	JSON syntax	ARM template file	Manual JSON, verbose
3	Bicep	Write Bicep code	Simpler syntax	Bicep file	Easier than ARM JSON
4	Bicep	Compile Bicep	Bicep file	ARM JSON template	Bicep compiles to ARM
5	Terraform	Write HCL code	Terraform syntax	Terraform config files	Multi-cloud support
6	ARM/Bicep	Deploy template	ARM JSON template	Azure resources created	Uses Azure native deployment
7	Terraform	Run terraform apply	Terraform config	Azure resources created	Terraform manages state
8	End	Infrastructure ready	Deployed resources	Resources available in Azure	Deployment complete

💡 Infrastructure deployed successfully using chosen tool

Status Tracker

Variable	Start	After Step 2	After Step 4	After Step 6	Final
Template Code	None	ARM JSON template created	ARM JSON template from Bicep	ARM JSON deployed	Resources deployed
Bicep Code	None	None	Bicep code written	Compiled to ARM JSON	ARM JSON deployed
Terraform Config	None	None	None	Terraform config written	Resources deployed

Key Moments - 3 Insights

Why do we need to compile Bicep code before deployment?

Because Azure only understands ARM JSON templates, Bicep code must be converted to ARM JSON before deployment, as shown in execution_table step 4.

Can Terraform deploy resources outside Azure?

Yes, Terraform supports multiple clouds, unlike ARM and Bicep which are Azure-specific, as noted in execution_table step 5.

Why is ARM template considered verbose compared to Bicep?

ARM templates use JSON which is more detailed and harder to write, while Bicep uses simpler syntax that compiles to ARM JSON, shown in steps 2 and 3.

Visual Quiz - 3 Questions

Look at the execution_table, at which step does Bicep code get converted to ARM JSON?

A Step 3

B Step 4

C Step 5

D Step 6

💡 Hint

Check the 'Action' column for 'Compile Bicep' in execution_table row 4

According to variable_tracker, which variable changes after Step 4?

A Template Code

B Terraform Config

C Bicep Code

D None

💡 Hint

Look at 'Template Code' row and its value after Step 4 in variable_tracker

If you want to manage infrastructure across multiple clouds, which tool is best based on execution_table?

A ARM

B Bicep

C Terraform

D None

💡 Hint

See notes in execution_table step 5 about multi-cloud support

Concept Snapshot

ARM, Bicep, and Terraform are tools to deploy Azure infrastructure.
ARM uses JSON templates, which are verbose.
Bicep is simpler syntax that compiles to ARM JSON.
Terraform uses HCL and supports multiple clouds.
Choose based on complexity, multi-cloud needs, and ease of use.

Full Transcript

This visual execution shows how to decide between ARM, Bicep, and Terraform for deploying Azure infrastructure. First, you pick a tool. ARM requires writing JSON templates directly. Bicep lets you write simpler code that compiles into ARM JSON. Terraform uses its own language and can manage resources across clouds. The execution table traces each step from writing code to deploying resources. Variable tracking shows how code changes form and get deployed. Key moments clarify why Bicep compiles to ARM and Terraform's multi-cloud ability. The quiz tests understanding of these steps and tool differences.