April Wrap-up - Serverless Buch goes Alexa, Chatbot, Spring and Java EE

May 5, 2017

Tags: #wrapup #summary #digest #serverless #serverlessbuch #book #aws #cloud #alexa #chatbot #javaee #spring

As promised in march wrap-up, I’ll write some words about the serverless demo apps for my Serverless Computing Buch. I played with an Alexa skill, a Slack chatbot and running Spring and Java EE APIs on AWS Lambda.

Spring and Java EE goes Serverless

So you thought, you have to tie yourself to the AWS Lambda Java API if you want to develop functions running on it? Lucky us, AWS Labs created a library with which you can run existing Java APIs on AWS Lambda: awslabs/aws-serverless-java-container

I started with a Spring application firing up on Serverless. The API Gateway routes all requests to the Lambda function where special SpringLambdaContainerHandler handles the origin request and dispatches it to the Spring classes. For configuration there’s a @Configuration annotated class, like you know it from your Spring app.

Dependency Injection with autowiring works like a charm. Also handling all requests with Spring MVC. Just use your well known annotations and you’re done. Additionally I’ve put up the example with Spring Data DynamoDB. You can find the repository at serverlessbuch/lambda-spring.

The Java EE example is very analog to the Spring one. The library supports running JAX-RS with Jersey (reference implementation) on AWS Lambda, so you have a JerseyLambdaContainerHandler instead of the SpringLambda… one. Pure JAX-RS was too boring for me, so I added some CDI with Weld (also reference implementation) to have the resources automatically injected into my classes. For validating the inputs, I also added Bean Validation (JSR-303) to the example. It works pretty well! Find the repository at serverlessbuch/lambda-jaxrs-cdi.

Alexa Skill

This was really fun - Alexa loves AWS Lambda!
It’s so easy creating an Alexa skill (for Amazon Echo and Echo Dot). Just define a few Intents and Slots for getting the spoken user input as data. Forward it to a AWS Lambda function and create a response. Alexa will read it out to you!

Link to the repository: serverlessbuch/alexaskill

Slack Chatbot

Can’t get enough of talking to machines? Create a serverless Slack Chatbot!

There are various Slack APIs for usage. While AWS Lambda (or serverless in general) is stateless, no real-time API (based on websockets) can be used. But there is the Event API which will notify you about certain events whith a HTTP(s) request and you can “respond” asynchronously via the Web API with messages. Another interesting option is using the so called Slash commands where the user enters something like /calc 2+2 and your created Slack chatbot will answer 2+2 = 4.

The repository contains both, the slash commands and the Event/Web API examples: serverlessbuch/slack-chatbot

Read all the details and more in my upcoming book: serverlessbuch.de and follow the book itself on Twitter: @serverlessbuch.

So long, stay tuned and if you have any questions, don’t hesitate to ask. As always, I’m happy to answer!

« March Wrap-up - All Serverless - Book and JavaLand May Wrap-up - Jax Conference and Serverless Book »