Step 1: Copy the curl Command from the UI
On the Restore Backup page for your MongoDB instance, click { REST API }. You’ll copy a command such as:

curl -v \
  -H "Authorization: Bearer <YOUR_JWT_TOKEN>" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json" \
  -d '{
    "bu_id":          "<BACKUP_ID>",
    "autostart":      true,
    "restore_option": "this",
    "target_id":      "<MONGO_HOST_ADDRESS>:<MONGO_HOST_PORT>",
    "restore_time":   null,
    "rest_api":       true
  }' \
  <API_BASE_URL>/restore_mongo_node/

Step 2: Parse the curl Command

• Authorization header:
  Extract the JWT token after Bearer.
• Content-Type & Accept:
  Both must be application/json.
• Payload fields:
  • bu_id (string, unique backup identifier)
  • autostart (boolean, whether to start node after restore)
  • restore_option (string, typically "this")
  • target_id (string, <host>:<port> of MongoDB node to restore to)
  • restore_time (nullable/timestamp, point-in-time restore; null for latest/full backup)
  • rest_api (boolean)
• Endpoint:
  • Base URI: <API_BASE_URL>
  • Resource path: /restore_mongo_node/

Step 3: Translate into Terraform

• Create directory & file:

integration/
└── terraform/
    └── restore_backup/
        └── main.tf

• Provider block

// Terraform configuration for automating MongoDB node restore from backup on 123cluster
terraform {
  required_providers {
    restapi = {
      source  = "Mastercard/restapi"
      version = "1.19.1"
    }
  }
}

/*
  REST API provider configuration:
  - uri: Base API endpoint for 123cluster.
  - headers: JWT and content type.
  - write_returns_object: Output as object.
  - debug: Enables logging for troubleshooting.
  - HTTP methods: POST for all actions.
*/
provider "restapi" {
  uri                  = "<API_BASE_URL>"
  write_returns_object = true
  debug                = true
  headers = {
    Authorization = "Bearer <YOUR_JWT_TOKEN>" // Use a valid, secure JWT token
    Content-Type  = "application/json"
    Accept        = "application/json"
  }
  create_method  = "POST"
  update_method  = "POST"
  destroy_method = "POST"
}

• Variable declarations

/*
  Use variables for all backup IDs, target nodes, and restore options to ensure flexibility and automation.
*/
variable "backup_id" {
  description = "Unique identifier of the MongoDB backup to restore from"
  type        = string
  default     = "<BACKUP_ID>"
}

variable "mongo_host_address" {
  description = "Target MongoDB node address (IP or FQDN)"
  type        = string
  default     = "<MONGO_HOST_ADDRESS>"
}

variable "mongo_host_port" {
  description = "Target MongoDB node port"
  type        = string
  default     = "<MONGO_HOST_PORT>"
}

variable "autostart" {
  description = "Whether to start the MongoDB node after restore"
  type        = bool
  default     = true
}

variable "restore_option" {
  description = "Restore option (typically 'this')"
  type        = string
  default     = "this"
}

variable "restore_time" {
  description = "Point-in-time restore timestamp (null for latest/full backup)"
  type        = string
  default     = null
}

• Resource definition

/*
  Resource for restoring a MongoDB node from backup via REST API.
  All fields are parameterized for portability and automation.
*/
resource "restapi_object" "restore_node" {
  path = "/restore_mongo_node/"
  data = jsonencode({
    bu_id          = var.backup_id
    autostart      = var.autostart
    restore_option = var.restore_option
    target_id      = "${var.mongo_host_address}:${var.mongo_host_port}"
    restore_time   = var.restore_time
    rest_api       = true
  })
}

• Output block‍

/*
  Output the full API JSON response after triggering the restore.
  Use for logs, CI/CD tracking, or downstream automation.
*/
output "restore_node_response" {
  description = "Raw JSON response from restore_mongo_node"
  value       = restapi_object.restore_node.data
}

Step 4: Initialize & Apply

cd integration/terraform/restore_backup

# Initialize the Terraform working directory and download necessary providers
terraform init

# Apply the configuration, review the planned actions, and confirm execution
terraform apply

# Output the API response for logging or integration with other tools
terraform output restore_node_response

‍

Additional Guidance & Best Practices

• Parameterization: Use variables for all backup IDs, target node addresses, and restore options to ensure safe, portable, and environment-agnostic automation.
• Security: Always store JWT tokens and sensitive data securely—use environment variables, secret managers, or mark variables as sensitive in Terraform.
• CI/CD Integration: Integrate this workflow into your CI/CD pipelines for reliable, auditable, and automated restore operations.
• API Versioning: Regularly check the 123cluster API documentation to ensure compatibility with any changes to restore endpoints or payloads.
• Logging & Outputs: Capture and log API responses as outputs for compliance, monitoring, and integration with follow-up jobs (such as health checks or notifications).
• Point-in-Time Recovery: For production, document and test the use of the restore_time parameter for point-in-time recovery scenarios.
• Environment Isolation: Use Terraform workspaces to avoid conflicts between different environments (dev, staging, prod).
• Error Handling: Implement checks for API errors and ensure your automation triggers alerts or retries on failures.
• Confirmation: Add manual approvals or reviews for destructive or production restores to minimize risk.

‍