Serverless

Introduction#

PaaS deploy applications. Serverless goes even further and enables the deployment of individual functions.

Thereby, Serverless allows even smaller deployments than with PaaS. A REST service can be divided into a variety of functions, one per HTTP method and resource.

The advantages of Serverless are similar to those of PaaS:

• A high degree of abstraction and thus relatively simple deployment.
• Serverless functions are only activated when a request is made so that no costs are incurred if no requests are processed.
• They can also scale very flexibly.

Serverless technologies include:

• AWS Lambda
• Google Cloud Functions
• Azure Functions
• Apache OpenWhisk. OpenWhisk allows you to install a Serverless environment in your own data center.

As with a PaaS, Serverless includes support for operation. Metrics and log management are provided by the cloud provider.

REST with AWS Lambda and the API gateway#

With AWS Lambda, REST services can be implemented. The API gateway in the AWS cloud can call Lambda functions.

A separate function can be implemented for each HTTP operation. Instead of a single PaaS application, there are many Lambda functions.

A technology like Amazon SAM can make the large number of Lambda functions quite easy to use.

Even if there are a lot of functions to deploy, it is barely any more effort for the developers than if they implement the REST methods in a class. Often Lambda functions can significantly reduce the cost of running a solution.

Glue code#

In another area, Lambda functions are helpful; in response to an event in the Amazon Cloud, a Lambda function can be called.

S3 (Simple Storage Service) offers storage for large files in the Amazon Cloud. When a new file is uploaded, a Lambda function can convert it to another format.

However, these are not real microservices, but rather glue code to complement functionalities in Amazon services.

                                                 Q U I Z  

We’ll conclude this chapter in the next lesson.