Getting Started with Containerlab

What is Containerlab?

Containerlab is an open-source tool that simplifies the deployment of container-based networking labs. It uses Docker and containerized network operating systems (NOS) like FRR, Nokia SR Linux, Arista cEOS, Cisco NX-OS, and SONiC to create realistic network environments.

Why Use Containerlab?

Lightweight – Uses containers instead of full VMs
Faster Deployment – Start up a lab in seconds
Multi-Vendor Support – Cisco, Arista, Nokia, Juniper, and more
Reproducibility – YAML-based configuration for easy sharing
Integration with Automation – Works well with Ansible, Netconf, and gRPC

---

Installing Containerlab

Containerlab requires Docker and sudo/root privileges to run.

Step 1: Install Docker

Containerlab runs on top of Docker, so ensure you have it installed:

sudo apt update && sudo apt install -y docker.io

Verify Docker is running:

docker --version

Step 2: Install Containerlab

The easiest way to install Containerlab is via the official script:

bash -c "$(curl -sL https://get.containerlab.dev)"

Alternatively, install manually using deb or rpm packages:

curl -sL https://github.com/srl-labs/containerlab/releases/latest/download/containerlab-linux-amd64 -o /usr/local/bin/containerlab
chmod +x /usr/local/bin/containerlab

Step 3: Verify Installation

After installation, check if Containerlab is working:

containerlab version

---

Understanding Containerlab Topology Files

Containerlab uses a YAML-based topology definition file. This file defines nodes (devices), links, and management settings.

Basic Structure of a Containerlab Topology File

name: my-first-lab

topology:
  nodes:
    router1:
      kind: vr-ros
      image: vrnetlab/vr-routeros:latest

    router2:
      kind: vr-ros
      image: vrnetlab/vr-routeros:latest

  links:
    - endpoints: ["router1:eth1", "router2:eth1"]

Key Sections:

1. name – Name of the lab
2. topology – Defines all devices and links
3. nodes – Lists the devices (e.g., routers, switches)
4. kind – Specifies the device type (e.g., vr-ros for RouterOS, ceos for Arista EOS)
5. image – The Docker image used
6. links – Defines connections between devices

---

Deploying Your First Lab

Now, let’s deploy and manage a simple lab with two routers.

Step 1: Define the Topology

Create a new directory for your lab and navigate into it:

mkdir ~/containerlab-labs && cd ~/containerlab-labs

Create a new YAML file (mylab.clab.yml) and add the following:

name: simple-lab

topology:
  nodes:
    r1:
      kind: vr-ros
      image: vrnetlab/vr-routeros:latest

    r2:
      kind: vr-ros
      image: vrnetlab/vr-routeros:latest

  links:
    - endpoints: ["r1:eth1", "r2:eth1"]

Step 2: Deploy the Lab

Run the following command to deploy the lab:

sudo containerlab deploy -t mylab.clab.yml

Once deployed, list running labs:

containerlab inspect

To see the running Docker containers:

docker ps

---

Logging into Nodes

Once the lab is deployed, you can access the network devices using different methods.

Method 1: Using docker exec

Each node runs as a Docker container. To access a node’s shell:

docker exec -it r1 vtysh

or for a Bash shell:

docker exec -it r1 /bin/bash

For routers running Cisco’s IOS-XE or Arista EOS:

docker exec -it r1 Cli

Method 2: Using containerlab inspect

Get the management IP of a node:

containerlab inspect -t mylab.clab.yml

Find the IP under mgmt_ipv4 and SSH into it:

ssh admin@<MGMT-IP>

Example:

ssh admin@172.20.20.2

Method 3: Directly Connecting via Terminal

For NOS images like Arista cEOS, Cisco NX-OS, or Nokia SR Linux, they expose a default SSH port. You can SSH using:

ssh -p 22 admin@$(docker inspect -f '' r1)

For containers exposing console ports via telnet, check with:

docker ps | grep r1

Then connect using:

telnet 127.0.0.1 <PORT>

---

Stopping and Deleting Labs

Once you’re done, clean up the environment with:

sudo containerlab destroy -t mylab.clab.yml

This will stop and remove the containers while keeping the YAML file intact.