Migrate to Terraform

Automatically generate Terraform code from your existing AWS and GCP infrastructure

Overview

Infra.new's cloud resource import feature helps you bring your existing cloud infrastructure under Terraform management. Whether you've been managing resources manually through the console or using other tools, our AI-powered import process automatically generates clean, maintainable Terraform code for your existing resources.

This feature supports both AWS and Google Cloud Platform resources, allowing you to:

  • Discover all resources in your cloud accounts
  • Select specific resources to import
  • Generate production-ready Terraform configurations
  • Configure state backend management
  • Provide context to guide the AI's code generation

How It Works

Infra.new uses read-only credentials to find resources in your cloud accounts and generate Terraform code for them.

1

Select Cloud Provider

Choose between AWS or Google Cloud Platform to begin the import process.

2

Configure Credentials

Select existing credentials or create new ones with appropriate permissions to read your cloud resources.

3

Discover & Select Resources

Infra.new scans your account and presents all resources. Select which ones to import into Terraform.

4

Configure State Backend

Choose where to store your Terraform state - S3, GCS, or locally for testing.

5

Generate Terraform Code

Provide additional context to guide the AI, then generate production-ready Terraform configurations.

Note: The import process is read-only and will not modify any existing resources. It only generates Terraform code that represents your current infrastructure state.

Setup Instructions

Select your cloud provider then follow the steps to import your cloud resources into Terraform:

Before you can import AWS resources, you need to configure AWS credentials with the appropriate permissions.

Configure AWS Credentials

Follow our step-by-step guide to create and configure IAM credentials

Step 1: Enable AWS Resource Explorer

AWS Resource Explorer must be enabled to discover resources in your account:

  1. Navigate to AWS Resource Explorer in the console
  2. Click Turn on Resource Explorer
  3. Select Quick setup
  4. Choose an aggregator region (typically us-east-1)
  5. Click Turn on

Note: Initial indexing takes 15-30 minutes to complete.

Step 2: Add Credentials to Infra.new

Once you have your cloud credentials, add them to Infra.new to start the import process:

  1. Navigate to Credentials in the Infra.new dashboard
  2. Click Import Cloud Resources
  3. Select your cloud provider (AWS or GCP)
  4. Choose to either:
    • Use existing credentials if you've already added them
    • Add new credentials using the access keys or JSON file from Step 1

Step 3: Discover and Select Resources

After adding credentials, Infra.new will scan your cloud account and display all discovered resources:

Resource Discovery

  • The discovery process typically takes 30-60 seconds depending on the number of resources
  • Resources are organized by service type (e.g., EC2, S3, RDS for AWS)
  • Each resource shows key information like name, type, region, and current state

Selecting Resources to Import

  • Use the checkbox next to each resource to select it for import
  • Use the service-level checkboxes to select all resources of a type
  • The search bar helps you find specific resources by name or ID
  • Filter by region, service type, or resource state

Best Practices for Resource Selection

  • Start with a small set of related resources for your first import
  • Group resources by application or environment
  • Consider dependencies between resources (they'll be handled automatically)
  • Review the resource count before proceeding - large imports may take longer to generate

Step 4: Configure State Backend

Choose where to store your Terraform state file. This is crucial for managing your infrastructure going forward:

Recommended

Remote State Backends

For production use, we strongly recommend using a remote backend:

S3 Backend
  • • Ideal for AWS-based infrastructure
  • • Supports state locking with DynamoDB
  • • Can create a new bucket or use existing
GCS Backend
  • • Best for GCP infrastructure
  • • Built-in state locking
  • • Can create a new bucket or use existing

When configuring a remote backend, you'll specify the bucket name, prefix/key for organizing state files, and the region where the bucket is located.

Local State Backend

For testing or development purposes, you can use local state storage:

  • State is stored on your local machine
  • No remote infrastructure required
  • Not recommended for team environments or production

⚠️ Local state files are not shared between team members and can be accidentally deleted

Step 5: Review and Generate Code

The final step allows you to review your selections and provide context to guide the AI code generation:

Summary Review

Before generating code, you'll see a summary of:

  • Total number of resources selected for import
  • Resources grouped by service type
  • Selected state backend configuration
  • Any detected dependencies between resources

Provide Additional Context

To generate optimal Terraform code, you can provide additional context about your infrastructure:

Example context you might provide:

  • "This is a production environment for our e-commerce application"
  • "Organize resources into modules by service type"
  • "Include variables for environment-specific configuration"
  • "Add comments explaining the purpose of each resource"
  • "Use naming conventions that follow our company standards"

Code Generation

Once you click Generate Terraform Code, Infra.new will:

  1. Analyze all selected resources and their configurations
  2. Detect relationships and dependencies
  3. Generate clean, modular Terraform code
  4. Include the import blocks needed to adopt existing resources
  5. Configure the specified state backend
  6. Apply any context or preferences you provided

After Generation

Once your Terraform code is generated, you can:

  • Review and edit the generated code directly in Infra.new
  • Run terraform plan to verify the import
  • Apply the configuration to complete the import process
  • Continue managing your infrastructure as code
  • Track changes to your infrastructure over time

Supported Resources

Infra.new supports importing most commonly used cloud resources. Here's a partial list by provider:

AWS Resources

Compute

EC2 Instances, Lambda Functions, ECS Services, EKS Clusters

Storage

S3 Buckets, EBS Volumes, EFS File Systems

Database

RDS Instances, DynamoDB Tables, ElastiCache Clusters

Networking

VPCs, Subnets, Security Groups, Load Balancers

IAM

Roles, Policies, Users, Groups

GCP Resources

Compute

Compute Instances, Cloud Run Services, GKE Clusters

Storage

Cloud Storage Buckets, Persistent Disks

Database

Cloud SQL Instances, Firestore, Bigtable

Networking

VPC Networks, Subnets, Firewalls, Load Balancers

IAM

Service Accounts, IAM Policies, Custom Roles

This is not an exhaustive list. Infra.new continuously adds support for more resource types. If you encounter an unsupported resource, it will be flagged during the discovery process.

Limitations and Considerations

  • State File Size: Very large infrastructures may result in large state files. Consider importing resources in logical groups.
  • Sensitive Data: Some resources may contain sensitive data in their configuration. Review generated code and use variables for sensitive values.
  • Resource Dependencies: Complex dependency chains are automatically detected, but manual review is recommended for critical resources.
  • Naming Conflicts: Imported resource names are sanitized to be Terraform-compliant. You may want to refactor names after import.
  • Default Resources: Some default or system-managed resources are filtered out during discovery as they shouldn't be managed by Terraform.

Next Steps

After successfully importing your cloud resources:

Connect State Backends

Set up continuous monitoring of your Terraform state for change tracking

Create Blueprints

Use the generated code as a blueprint for future infrastructure

After Generation

After Generation

Once your Terraform code is generated, you can:

  • Review and edit the generated code directly in Infra.new
  • Run terraform plan to verify the import
  • Apply the configuration to complete the import process
  • Continue managing your infrastructure as code
  • Track changes to your infrastructure over time

Supported Resources

Infra.new supports importing most commonly used cloud resources. Here's a partial list by provider:

AWS Resources

Compute

EC2 Instances, Lambda Functions, ECS Services, EKS Clusters

Storage

S3 Buckets, EBS Volumes, EFS File Systems

Database

RDS Instances, DynamoDB Tables, ElastiCache Clusters

Networking

VPCs, Subnets, Security Groups, Load Balancers

IAM

Roles, Policies, Users, Groups

GCP Resources

Compute

Compute Instances, Cloud Run Services, GKE Clusters

Storage

Cloud Storage Buckets, Persistent Disks

Database

Cloud SQL Instances, Firestore, Bigtable

Networking

VPC Networks, Subnets, Firewalls, Load Balancers

IAM

Service Accounts, IAM Policies, Custom Roles

This is not an exhaustive list. Infra.new continuously adds support for more resource types. If you encounter an unsupported resource, it will be flagged during the discovery process.

Limitations and Considerations

  • State File Size: Very large infrastructures may result in large state files. Consider importing resources in logical groups.
  • Sensitive Data: Some resources may contain sensitive data in their configuration. Review generated code and use variables for sensitive values.
  • Resource Dependencies: Complex dependency chains are automatically detected, but manual review is recommended for critical resources.
  • Naming Conflicts: Imported resource names are sanitized to be Terraform-compliant. You may want to refactor names after import.
  • Default Resources: Some default or system-managed resources are filtered out during discovery as they shouldn't be managed by Terraform.

Next Steps

After successfully importing your cloud resources:

Connect State Backends

Set up continuous monitoring of your Terraform state for change tracking

Create Blueprints

Build reusable templates from your imported infrastructure