Automate your bare-metal fleet from code

The API is designed so that nothing requires the panel. Common automations:

• Provisioning: order servers with a plan, location, OS image, hostname and SSH keys, then poll until the server is active.
• Rebuilds: reinstall a server from a clean image as part of a deploy, or on a schedule for CI runners.
• Power and recovery: reboot, power-cycle or boot into rescue mode when a health check fails.
• Networking: set reverse DNS when a mail or web server comes online.
• Inventory: list your servers and their status to feed monitoring, a CMDB or cost reports.

Treat ordering calls with care: creating a server starts a monthly charge, so put an approval or a plan review in front of any automation that can order hardware. Keep tokens out of source control, store them in your CI or secret manager, and log the calls your automation makes. The full list of endpoints is in the API reference.

The Terraform provider maps FlexDedi servers to resources, so bare metal follows the same workflow as the rest of your infrastructure: write the desired state, run terraform plan to see what will change, and apply it after review.

resource "flexdedi_server" "db" {
  plan     = "pro-32"
  location = "fra"
  os       = "ubuntu-24.04"
  ssh_keys = ["ops-laptop"]
}

Physical servers are billed monthly and take minutes rather than seconds to deliver, so read plans carefully: replacing a resource means a new server or a reinstall. Add prevent_destroy lifecycle rules to databases and other stateful machines, and keep state in a remote backend with locking.

Keep one state per environment, such as staging and production, and combine the provider with the ones you already use for DNS and monitoring, so a new server gets its records and checks in the same apply. Pass server addresses from Terraform outputs to Ansible or your configuration tool.