Cloud Native Architectures

Scaling

Serverless applications are able to scale by design, without the requirement of having to understanding individual server capacity units. Instead, consumption units are the important factor for these services, typically throughput or memory, which can be automatically updated to have more or less capability as the application requires. In addition, for functions (such as AWS Lambda), since they are short lived (less than five minutes), they do not auto scale like an application deployed to a cloud server instance when the CPU is heavily used. Functions are designed to execute as an event so every time that event happens they trigger the function, which executes exactly once and dies off. Hence, the scaling capabilities of functions ...