Use GitHub Codespaces

You can run BGP labs in (free1) GitHub codespaces; all you need is a GitHub account:

Select Lab Devices

The BGP labs repository uses Cumulus Linux 4.4 as the default device. To change the device settings, edit the defaults.yml file with vi or nano.

Tip

  • It’s best to use network devices with free-to-download container images (Cumulus Linux, FRR, Nokia SR Linux, Vyos).
  • Use FRR to use the 2 CPU/8 GB codespaces VM with larger labs.
  • Codespaces have persistent storage; you can download and install other containers.
  • To use containers that have to be downloaded from the vendors’ website, download them onto your laptop, drag-and-drop them into the Folders, and install them like you would on a local netlab instance.
  • While it looks like the Codespaces environment supports nested virtualization, the environment does not include the kvm kernel module, so we cannot get virtual machines to work (yet). That also precludes running VMs in virtual machines (the vrnetlab approach).

Start a Lab

Once you have the codespaces up and running:

  • Click on the desired lab exercise in the README.md preview window to select the exercise folder.
  • Right-click on the exercise folder and select “Open in Integrated Terminal” to launch a bash session in the desired directory.
  • Execute netlab up to start the lab.
  • Expand the exercise folder in the Explorer tab.
  • Right-click on the README.md file and select “Open Preview” to open the rendered version of the file.
  • Click the link in the README.md file to get the exercise description in the preview pane.
  • Connect to your devices with the netlab connect command executed in the Terminal pane.

Cleanup and Shutdown

Finally, don’t forget to shut down the lab with netlab down and stop your codespace after you’re done:

  • Click on the blue “Codespaces” button in the bottom-left corner of the browser window.
  • Select “Stop Current Codespace”. You should also adjust idle timeout and default retention period in your codespaces settings.

  1. You get 120 free core hours per month or pay for more. ↩