Deployment to Production Servers

Deploy Compute for Production

How to deploy Compute to production server environment. To run Compute locally for development and testing, read Running and Debugging Compute Locally.

1. Set up Core-Hour Billing

Follow the “Core-Hour Billing” guide to get set up.

2. Prepare Windows Server

To run Compute you’ll need a server or virtual machine pre-installed with Windows Server 2019.

We’ll assume you’re deploying Compute to one of Amazon’s EC2 instances. There are a few things to pay attention to when setting up the instance – use this as a rough guide if you’re using a virtual machine from another cloud provider or a physical server.

  • Start with the “Microsoft Windows Server 2019 Base” AMI.
  • The t2.medium instance type (2 vCPU, 4 GB RAM) is recommended.
  • Assign a public ip, or better yet use an Elastic IP and Route53.
  • Set a “Name” tag to help keep track of instances.
  • Configure the security group to allow Compute traffic:
    • RDP - 3389 TCP
    • HTTP - 80 TCP
    • HTTPS - 443 TCP

Wait for the virtual machine to spin up… ☕️

3. Install Rhino and Compute

On the virtual machine, copy and paste the command below into a powershell window and hit Enter. You will be asked to enter a few things…

  • EmailAddress - the Rhino download link requires a valid email address
  • ApiKey - the API Key that clients will use when communicating with Compute
  • RhinoToken – the long token that links Rhino to your core-hour billing team
iwr -useb -outfile bootstrap.ps1; .\bootstrap.ps1 -install

At the end of the installation process, Windows will restart to complete the setup. Wait a minute and log back in to check that the compute.geometry service is running.

4. Verify Compute and license usage

  1. Open a browser and go to http://public-dns-or-ip/version. If Compute is working it will return its version and Rhino’s version.
  2. Go to the Licenses Portal (login to your Rhino account if prompted).
  3. Under Team Licenses click your new team.
  4. Verify that Rhino is in use in your core-hour billing team.

5. Next steps

Check out the Rhino Compute AppServer – a node.js server acting as a bridge between client apps and private Compute servers.