If you need to run a lab larger than 5 devices, so you can’t run it in the free version of CML that you maybe have running in a VM, you can instead use a free Cisco DevNet CML Sandbox. Click the link below to visit the site.
https://developer.cisco.com/site/sandbox
Select Launch Sandbox
Select Launch on the Cisco Modeling Labs tile.
Sign in using your desired account type.
As a matter of courtesy, and to free up the resource for others to use, select the number of hours you plan to use the sandbox (the default 8 hours) then click Review Summary.
Confirm everything looks good and click Launch Environment.
A timer will display a countdown while the sandbox is being prepared.
2. You will receive an email from devnetsandbox stating ‘We are preparing you Sandbox…’
Once the sandbox lab has launched, you will receive a second email, which contains the connection information specific to your sandbox.
You’ll also see that the right sidebar populates with information about the lab environment.
Establish a VPN connection
This is where things get a bit messy, in my opinion. The email from Cisco recommends using the Cisco AnyConnect VPN client to establish a VPN connection using the connection information that they provide. However, you can’t download the AnyConnect client unless you’re a Cisco enterprise customer. One might think that Cisco would provide a channel for learners to obtain the AnyConnect client, but no. Not that I could find anyway. You might be able to get the software from your university if you’re a student. That’s where I got it.
If you can’t use AnyConnect/SecureClient, Cisco recommends you use the OpenConnect VPN client, and basically wishes you “good luck!”
Using OpenConnect VPN
Cisco’s email graciously provides links to the OpenConnect info for Windows, Ubuntu, and MacOS.
Windows
I had installed OpenConnect on my Windows machine and been able to connect successfully and one point, but when I tried recently I couldn’t get it to work. I may take another stab at it in the near future.
Ubuntu
For Ubuntu, it’s a whole thing that deserves a write-up of its own, if I feel motivated at some point. (I don’t normally run Ubuntu, so…) I’ve provided the link to the instructions here, and a quick read-through gives you an indication of what you’ll be in for.
How to Install OpenConnect VPN Server on Ubuntu 22.04
Installing a list of dependencies (which strangely doesn’t include git which the doc asks you to install later), configuring Uncomplicated Firewall (UFW), then it recommends installing ocserv from source, so installing another list of dependencies, cloning a git repository, generating configuration scripts, building the app, editing the service file, generating SSL certificates… there’s more, but I’m gonna stop there. You get the idea. It’s a whole project all on its own.
MacOS
OpenConnect VPN on MacOS is relatively straightforward. You can either install OpenConnect using Homebrew, or via the OpenConnect GitHub repo.
I opted to use Homebrew, and ran ‘brew install open connect’ and the installer was off and running. The installer will instruct you to run a few commands post-installation, which is presents to you ready to copy and paste into the command line.
Once the OpenConnect installation and post-installation commands are completed, run the following command to connect using the connection info in Cisco’s email. (After reading the OpenConnect help documentation, I determined that the command expects the -u and –server= flags for the username and server name.)
sudo openconnect -u [username] --server=[server_name]Note: The first password prompt is for your sudo password, not the Cisco password
Note: The second password prompt is for the Cisco VPN password from the email.
You should then see the connection status information.
Accessing Cisco Modeling Labs
At this point, open a browser to the lab IP, for example https://10.10.20.161 and sign in using the User ID and Password from the Lab Environment Instructions tab, and you’ll be presented with the CML dashboard.
From here you can follow the usual process to import your own lab environment, or create one from scratch.
