Terraform Integration

FaultRay can import infrastructure models directly from Terraform configuration files and state, enabling resilience evaluation before deployment.

Setup

No additional dependencies are required. Terraform integration is included in the base FaultRay installation.

Import from Terraform files

HCL configuration files

faultray tf-import --dir ./terraform --output tf-model.json

Terraform state

faultray tf-import --state terraform.tfstate --output tf-model.json

Remote state (S3)

faultray tf-import --state s3://my-bucket/terraform.tfstate --output tf-model.json

Supported Resources

FaultRay parses the following Terraform resource types:

Provider	Resources
AWS	aws_instance, aws_db_instance, aws_lb, aws_elasticache_cluster, aws_s3_bucket, aws_cloudfront_distribution, aws_ecs_service, aws_lambda_function
GCP	google_compute_instance, google_sql_database_instance, google_compute_forwarding_rule, google_redis_instance
Azure	azurerm_virtual_machine, azurerm_sql_server, azurerm_lb, azurerm_redis_cache
Kubernetes	kubernetes_deployment, kubernetes_service, kubernetes_ingress

CI/CD Usage

Gate Terraform applies on resilience score:

# In your CI pipeline
resource "null_resource" "resilience_check" {
  provisioner "local-exec" {
    command = "faultray tf-import --dir . --output /tmp/model.json && faultray evaluate -m /tmp/model.json --threshold 70"
  }
}

Example: Pre-apply validation

# Plan changes
terraform plan -out=tfplan

# Import current + planned state
faultray tf-import --dir . --output current.json
faultray tf-import --plan tfplan --output planned.json

# Compare resilience impact
faultray diff current.json planned.json --only-regressions

If regressions are detected, the diff command exits with code 2, allowing CI/CD pipelines to block the apply.