Direct Access to External Storage
In this lesson, you will learn how to remove the gatekeeper role from the Lambda functions to get away from traditional architecture.
This chapter shows how to start shifting the core responsibilities of typical servers to the AWS platform, and how to benefit from serverless architectures to make your application faster and cheaper. You’ll also learn about the options to let client devices directly use AWS resources on your behalf.
In the previous chapters, you deployed a simple application that could be ready for millions of users if necessary but doesn’t cost anything if nobody is using it. AWS operates and monitors the application, so most of the ‘ops’ part of DevOps is effectively included in the price. This is quite a nice outcome for such little effort, mostly achieved because you didn’t have to write infrastructural code. With Lambda, API Gateway, and S3, all that infrastructural stuff comes out of the box.
Three-tier server setup #
Our application still looks very similar to a typical three-tier server setup. In such an architecture, an application server usually contains the business logic, but also plays two more roles. It works as a gatekeeper and a workflow orchestration system. The application server approves or rejects client requests, talks to back-end storage, schedules asynchronous processing and passes back results to client devices. For highly available and scalable web applications, the code running on a client device usually talks to the application server ...