RunPod Terraform Provider

Official Terraform provider for RunPod - manage GPU cloud infrastructure as code.

Features

• Pod Management: Create, read, and delete GPU/CPU pods with flexible configuration
• Network Volumes: Manage persistent storage volumes across pods
• Serverless Endpoints: Deploy and manage serverless GPU endpoints
• Data Sources: Query existing resources including pods, volumes, endpoints, and templates

Supported Resources

Resources

• runpod_pod - Manage GPU/CPU pods
• runpod_network_volume - Manage persistent network storage
• runpod_endpoint - Manage serverless endpoints

Data Sources

• runpod_pods - List all pods
• runpod_network_volumes - List all network volumes
• runpod_endpoints - List all serverless endpoints
• runpod_templates - List available templates

Requirements

• Terraform >= 1.0
• Go >= 1.23
• RunPod API key (get one at runpod.io)

Quick Start

1. Set your API key:

   export RUNPOD_API_KEY="your-api-key-here"

2. Create a basic configuration:

   terraform {
     required_providers {
       runpod = {
         source = "decentralized-infrastructure/runpod"
       }
     }
   }
   
   provider "runpod" {
     # API key from environment variable RUNPOD_API_KEY
   }
   
   # Create a GPU pod
   resource "runpod_pod" "example" {
     name              = "my-gpu-pod"
     image_name        = "runpod/pytorch:2.1.0-py3.10-cuda11.8.0-devel"
     gpu_type_ids      = ["NVIDIA GeForce RTX 4090", "NVIDIA A40"]
     data_center_ids   = ["US-CA-2", "US-TX-3"]
     
     gpu_count            = 1
     cloud_type           = "COMMUNITY"
     support_public_ip    = true
     volume_in_gb         = 20
     container_disk_in_gb = 20
     
     ports = ["8888/http", "22/tcp"]
   }

3. Deploy:

   terraform init
   terraform plan
   terraform apply

Examples

See the examples/ directory for comprehensive usage examples.

Building The Provider

1. Clone the repository
2. Enter the repository directory
3. Build the provider:

go build -o terraform-provider-runpod

Adding Dependencies

This provider uses Go modules.

To add a new dependency:

go get github.com/author/dependency
go mod tidy

Developing the Provider

If you wish to work on the provider, you'll first need Go installed on your machine.

To compile the provider, run go build. This will build the provider and put the provider binary in the current directory.

For development testing, you can use a .terraformrc file to override the provider location:

provider_installation {
  dev_overrides {
    "registry.terraform.io/decentralized-infrastructure/runpod" = "/path/to/your/go/bin"
  }
  direct {}
}

Documenting the Provider

In order to generate documentation for the provider, the following command can be run:

go get github.com/hashicorp/terraform-plugin-docs/cmd/tfplugindocs
tfplugindocs generate

License

This project is licensed under the Mozilla Public License 2.0 - see the LICENSE file for details.