Chapter 1 What is Network Automation?

 Application of standards or conventions:

o This can be a standard Network Time Protocol (NTP) server for all network devices or

ensuring the native VLAN is set correctly on all trunk ports.  Correcting errors:

o If there are known problems on the network that have been put off because of the scale of the

change, lack of priority, tedium, understanding of the amount of manual effort required, lack of resources, or waiting for an automated solution to arrive.  Net-new projects:

o Draw a line in the sand and try to commit to automating all net-new project requirements that

require changes to the network.

As part of the repository, have a place for one-time tactical Ansible playbooks. These playbooks will become part of the larger network automation framework and are available for cloning and repurposing. Examples of successful plays helps others develop their own playbooks and code. Chapter 7 has several use cases of one-time tactical plays that can be used as a foundation to solve unique challenges.

At this point the organization should be comfortable with using network automation for information collection and making changes to the network. While these are both great capabilities that will completely change network operations, the real power of network automation is full configuration management coverage. Every device is configured by the network automation engine using dynamic templates powered by abstracted data models. Essentially, converting the network configurations to code. This code is intent-driven data models in human-readable format, comprised of variables and dynamic templates that contain the configurations. Version control can now be implemented using Git. The compiled configurations are then pushed to the network through automated Ansible playbooks. Chapter 8 covers this in more detail.

Creating a CI/CD pipeline is the last step to fully automating the network. Automated builds, tests and releases are now available to network operations. As part of the branching strategy, builds can be triggered automatically whenever pull requests merge code into the master branch (CI). Releases can be scheduled and deployed to the network much like a patch to software is released to applications. Scheduled releases can occur at any frequency – hourly, nightly, weekly, monthly. Approved changes are merged into the master branch in the repository which automatically triggers the deployment during the next scheduled release (CD). This is the full power of network automation unleashed.

Impact on IT

The impact on IT should not be underestimated. There is a lot of change and learning involved. It is not just the tools or new file types or data formats, but the very mindset of IT itself, which can be very hard to change. Acceptance of innovation often faces risk adversity and a natural resistance to change. Network automation will absolutely lead to changes in core processes for change management and network operations. Network maintenance will be automated, and this could mean some restructuring or refocusing of staff.

The only thing that can break network automation is manual intervention and therefore it is imperative that everyone involved with the network fully understands and is onboard with the new direction and processes. Manual changes are in direct contrast with intent-based networks and will be overwritten by