CML Installation Guide

This guide provides instructions for deploying a new instance of Cisco Modeling Labs (CML) or for upgrading an existing CML instance.

Attention

CML is designed to be a lab, training, and automated testing tool. As such, it should be deployed on an internal, trusted network. This recommendation is especially true when CML is being used as an automated testing platform, such as part of a CI/CD pipeline, or as an extension of physical lab gear. In these cases, CML becomes more critical to the infrastructure, and unauthorized access can lead to further network compromise.

Standalone vs. Cluster

When deploying CML, you should first decide whether to deploy a standalone CML server or a CML cluster. With a standalone instance, you deploy a single instance of CML. If you purchased CML-Personal or CML-Personal Plus, you will deploy a standalone CML instance. Clustering is only available in the non-personal CML products. With a cluster, you deploy multiple CML instances, where one CML instance acts as the controller for one or more computes. Clustering allows for horizontal scaling across multiple servers while still providing a single user interface to manage labs running on the CML cluster.

Bare Metal vs. Virtual Machine (VM)

There are two main options for deploying a new CML instance: virtual machine and bare metal installation. Customers who are familiar with VMware generally find that the VM deployment option is faster and makes the CML server easier to manage over time. For the bare metal installation option, you must dedicate a server to CML. The performance of a CML bare metal instance is marginally better than a CML VM deployment on the equivalent hardware. When you run a network topology simulation in CML, each node in the topology runs as a VM on the CML server. When CML is deployed as a VM, it uses nested virtualization to run the VMs for the simulated nodes.

Cloud Deployment on AWS and Azure

CML now supports running instances on cloud platforms such as AWS (Amazon Web Services) and Azure Deployments in these cloud environments are facilitated through a specialized toolchain, which is currently in BETA. The BETA status indicates that the toolchain is undergoing refinement and may evolve based on feedback and development progress. To ensure a smooth setup process, it is recommended to review the requirements and prerequisites detailed in the README file. For specific cloud platform instructions, refer to the CML on AWS documentation or the CML on Azure documentation.

In-Place Upgrades

If you already have an existing CML 2.x installation, you may instead perform an in-place upgrade of your existing system. You will use the same upgrade process no matter which deployment option, VM or bare metal server, you originally used for your CML server. In-place upgrades are not supported for all releases. See the Release Notes for Cisco Modeling Labs for details on which existing, installed CML versions can be upgraded to the current release. Upgrades from CML 1.x to CML 2.x are not supported. If an in-place upgrade from your existing CML server version is not supported, you will need to migrate to a new CML installation to move to the new release.

VM Images for CML Labs

CML provides VM images that you can use to simulate nodes in a network topology simulation. Each node in your CML labs uses one of these Cisco reference platform VM images or third-party VM images. We recommend using the latest reference platforms with any new CML installation. However, if you are upgrading an existing CML instance to the new release, then we recommend continuing to use the original reference platform ISO from the older release even after upgrading.

Note

Please note that Cisco images provided as part of the Cisco Modeling Labs reference platform ISO file are licensed only for use within Cisco Modeling Labs. Downloading the images and using them outside Cisco Modeling Labs without a proper license is prohibited.

Related Information