How many subnets should be created for a highly available web application using a public load balancer across all 3 ADs?

In designing a highly available web application using a public load balancer across all three Availability Domains (ADs), creating two subnets—one regional public and one regional private—is a strategic approach. The public subnet accommodates the load balancer, which is necessary for distributing incoming traffic to the web servers, while the private subnet hosts the actual web servers that require protection from direct internet access.

The public subnet allows external users to access the load balancer, while the use of a private subnet for the web servers enhances security by ensuring they are not accessible directly from the internet. This setup not only promotes security best practices but also adheres to the principles of high availability by leveraging resources spread across multiple availability domains.

Moreover, having separate subnets for public and private resources improves the manageability of the network architecture and allows for clearer traffic segregation and organization. This configuration aligns with the best practices for deploying secure and highly available applications in cloud environments, making it the appropriate choice for such a scenario.