The third and final cloud native principle is about automation.  Throughout this chapter, the other CNMM principles have been discussed in detail and explained, particularly why using cloud native services and application centric design enable cloud native architectures to achieve scale.  However, these alone do not allow a system to really take advantage of the cloud. If a system were designed using the most advanced services available, but the operational aspects of the application were done manually, it would have a hard time realizing its intended purpose. This type of operational automation is often referred to as Infrastructure as Code, and there is an evolution in maturity to achieve a sophisticated cloud native architecture. Cloud vendors typically develop all of their services to be API endpoints, which allows for programmatic calls to create, modify, or destroy services. This approach is the driver behind Infrastructure as Code, where previously an operations team would be responsible for the physical setup and deployment of components in addition to the infrastructure design and configuration.

With Infrastructure as Code automation, operations teams can now focus on the application-specific design and rely on the cloud vendor to handle the undifferentiated heavy lifting of resource deployment. This Infrastructure as Code is then treated like any other deployment artifact for the application, and is stored in source code repositories, versioned and maintained for long-term consistency of environment buildouts. The degree of automation still evolves on a spectrum with the early phases being focused on environment buildout, resource configuration, and application deployments. As a solution matures, the automation will evolve to include more advanced monitoring, scaling, and performance activities, and ultimately include auditing, compliance, governance, and optimization of the full solution. From there, automation of the most advanced designs use artificial intelligence and machine and deep learning techniques to self-heal and self-direct the system to change its structure based on the current conditions.

Automation is the key to achieving the scale and security required by cloud native architectures: