# Deploy Deadline Cloud farms with Terraform
<a name="examples-terraform"></a>

The [terraform](https://github.com/aws-deadline/deadline-cloud-samples/tree/mainline/terraform) directory in the deadline-cloud-samples repository on the GitHub website contains Terraform configurations that deploy Deadline Cloud farms with `awscc_deadline_*` resources from the AWSCC provider. For a comparison of Terraform with CloudFormation and the AWS CDK, see [Manage Deadline Cloud infrastructure as code](infrastructure-as-code.md).

The [starter\_farm](https://github.com/aws-deadline/deadline-cloud-samples/tree/mainline/terraform/farm_templates/starter_farm) configuration on the GitHub website is the Terraform equivalent of the CloudFormation starter farm ([Deploy a starter Deadline Cloud farm with CloudFormation](examples-cfn-starter-farm.md)). It deploys a farm with a production queue and a package build queue. The farm includes up to three service-managed fleets: CPU Linux, CPU Windows, and CUDA Linux. A conda queue environment uses a private conda channel on an Amazon S3 bucket you provide, plus the `deadline-cloud` channel.

To deploy the starter farm, you need Terraform 1.0 or later, AWS credentials, an Amazon S3 bucket for job attachments and the conda channel, and a Deadline Cloud monitor. Set the required `job_attachments_bucket_name` variable, then run:

```
cd terraform/farm_templates/starter_farm
terraform init
terraform apply
```

The [knfsd\_xregion\_cache](https://github.com/aws-deadline/deadline-cloud-samples/tree/mainline/terraform/farm_templates/knfsd_xregion_cache) configuration on the GitHub website deploys a service-managed fleet that reads a distant NFS filer through a KNFSD read cache over a VPC resource endpoint. Use it as a starting point when your fleet caches reads from an on-premises or otherwise-distant filer.