Automate Board Provisioning
This guide will walk you through how to set up you remote proxy machine to automate provisioning boards.
Pre-requisites
Before we start this tutorial, you will need:
An AMD x86-64 or ARM device to act as a remote proxy device
A board to be provisioned
USB cable to connect the board to the remote proxy device
A Remote Proxy Fleet
Enroll a Remote Proxy Device
Navigate to the Remote Proxies tab on the Fleet panel. If you haven't already, create a Remote Proxy Fleet to manage all of your upcoming proxy device enrollments. Make sure that the Proxy Fleet checkbox is ticked. Detailed instructions on how to create a fleet and enroll devices can be found at Creating a Fleet and Enrolling Machines guides respectively.
Several NVIDIA-required packages must be present on the Linux host serving as the Remote Proxy. Install them by running the following command once in a terminal.
curl -sfL https://cdn.chassy.io/test/provision-dependencies | bash -Set Up a Provisioning Workflow
Create a new Workflow to set up your provisioning details. Utilize the Find Package and Import Package steps to upload files into Chassy Index when needed.
Add a Provision Setup step to select your remote proxy device, and provide details on how you want the connected boards to be set up. Run the workflow when completed.
Connect a Board to your Remote Proxy
Going back to the Remote Proxies tab on the Fleet panel, you should see your remote proxy machine and its status.
Put your board into recovery mode, then connect it to your remote proxy machine using a USB cable. The provisioning process will automatically begin. On the Chassy Console, you can see which provisioning phase the connected board is on: Attached, Flashing, Configuring, Verifying, Succeeded, Error, Terminated, or Timed out.
For connected boards that got Error , Terminated , or Timed out , you can restart the provisioning process by clicking on the restart button.
Last updated