This Terraform stack provisions a small Azure VM, assigns a static public IP with an Azure-managed DNS hostname, installs Docker, clones the repo, and starts the existing production terminal stack from terminal/docker-compose.prod.yml.

What it creates

resource group wiring compatible with Azure CLI auth
virtual network, subnet, public IP, NIC, and network security group
one Ubuntu VM
cloud-init bootstrapping for Docker, repo clone, .env, and systemd

Why this fixes the current outage

The current public terminal endpoint uses terminal.egeuysal.com, which is serving an invalid certificate chain to browsers. This stack uses an Azure-managed hostname like:

https://cyberminds-terminal-20260621-ncus.northcentralus.cloudapp.azure.com

That hostname terminates through the repo's Caddy config and should obtain a valid certificate automatically once the VM is up and ports 80/443 are open.

Prerequisites

az login already completed
terraform installed
a resource group created, for example:

az group create --name cyberminds-terminal-ncus-rg --location northcentralus

Usage

cd terminal/infra/azure-vm
cp terraform.tfvars.example terraform.tfvars

Edit terraform.tfvars and set:

subscription_id
admin_ssh_public_key
optionally domain_name_label if the default is already taken
optionally ssh_allowed_source_addresses if you want public SSH access

Then run:

terraform init
terraform apply

After apply:

terraform output public_fqdn
terraform output terminal_health_url

Verify health:

curl -I "$(terraform output -raw terminal_health_url)"

SSH access:

terraform output -raw ssh_command

Public SSH is disabled unless you set ssh_allowed_source_addresses.

Notes

This deploy path uses the Azure hostname directly, so it does not require moving DNS for terminal.egeuysal.com.
The frontend terminal default API origin should match the public_fqdn output if the DNS label changes.