Make sure to have the latest containerlab installed.

you can check this with sudo clab version upgrade

Deploy the lab topology with the -c switch - this will always make sure that the lab is loaded from scratch, if you leave that off, it will load whatever the last state was from the configuration artifacts

sudo clab deploy -t topo.yml -c

When the topology finishes booting, you will have the following starting state.

Note that sros will be running in mixed-mode.

  1.1.1.1/32                    2.2.2.2/32
 ┌──────────┐                  ┌──────────┐
 │          │   12.12.12.0/31  │          │  
 │  sros    ├──────────────────┤   srl    │
 │          │.0              .1│          │
 │          │                  │ agg:12/8 │
 └──────────┘                  └──────────┘
                ISIS: 49.0001
                iBGP: 65001

Running in mixed-mode precludes us from using pysros. In order to achieve the goals of this script, I chose to utilize two different applications - pygnmi and netmiko

The cleanest way to install these is to run a python virtual env for this demo script:

python3 -m venv venv
source venv/bin/activate
pip install netmiko
pip install pygnmi

Once you have the prereqs installed and the topology is running in your containerlab, we can run the script.

The script is in the scripts directory, so I am assuming you are in the scripts directory when running the command

python rollback_demo.py

What the script does is very simple -

• netmiko Saves a rollback point in SROS
• pygnmi Changes the ISIS level-capability to 1, causing the ISIS adj to go down since the interfaces are set to level 2
• pygnmi Checks the state of the ISIS adj, and if it is down, calls for a rollback
• netmiko Performs a rollback