Boot Kubeless

Spring Boot project for Kubeless Custom Runtime

A simple Spring Boot Application to demonstrate how to build and run Custom Java Runtime on Kubeless for HTTP and Kafka Triggers.

It has two applications:

1. boot-kubeless-http: A simple Spring Boot app for Kubeless HTTP trigger. It provides a GET and POST endpoint to add/list simple domain entity to demonstrate the use of various Spring projects.
2. boot-kubeless-kafka: A simple Spring Boot app as Kafka Consumer for Kubeless Topic/Kafka trigger.

Tech

boot-kubeless uses a number of open source projects:

• Spring Boot - An opinionated framework for building production-ready Spring applications. It favors convention over configuration and is designed to get you up and running as quickly as possible.
• Kubeless - The Kubernetes Native Serverless Framework.
• Spring Data JPA - Spring Data JPA, part of the larger Spring Data family, makes it easy to easily implement JPA based repositories.
• Spring Boot Actuator - Actuator endpoints allow you to monitor and interact with your application.
• Spring for Kafka - The Spring for Apache Kafka project applies core Spring concepts to the development of Kafka-based messaging solutions.

Installation and Configuration

$ git clone https://github.com/RawSanj/boot-kubeless.git

$ cd boot-kubeless

Build and Run the applications:

Build and run the boot-kubeless-http application:

$ cd boot-kubeless-http

$ mvn clean package

$ mvn spring-boot:run

Build and run the boot-kubeless-kafka application:

Note: Download and Install Kafka. Set the Kafka Server Address and Topic name in application.yml file.

$ cd boot-kubeless-kafka

$ mvn clean package

$ mvn spring-boot:run

Run in Docker

Build and run the boot-kubeless-http locally in Docker:

Build the JAR file:

$ cd boot-kubeless-http

$ mvn clean package

Build docker image:

$ mvn docker:build

Run docker image:

$ docker run -d -p 8080:8080 \
$ rawsanj/boot-kubeless-http

Build and run the boot-kubeless-kafka locally in Docker:

Build the JAR file:

$ cd boot-kubeless-kafka
$ mvn clean package

Build docker image:

$ mvn docker:build

You would need to run Kafka in Docker. Try this Kafka docker image. Run docker image:

$ docker run -d -p 8080:8080 \
$ rawsanj/boot-kubeless-kafka --link <Kafka_Container>:alias

Deploy on Kubernetes as FaaS using kubeless cli

Setup and Install Kubeless.

Follow the Steps here.

Deploy the boot-kubeless-http as Kubeless Function with HTTP Trigger

Deploy the Spring Boot app boot-kubeless-http as a Function using docker image available in Docker Hub:

$ kubeless function deploy --runtime-image rawsanj/boot-kubeless-http --trigger-http boot-kubeless-http

Test the function:

Calls the GET method of Spring Boot App

$ kubeless function call boot-kubeless-http

Calls the POST method of Spring Boot App

$ kubeless function call boot-kubeless-http --data '{"language":"Go","rate":1}'

Deploy the boot-kubeless-kafka as Kubeless Function with Kafka/Topic Trigger

Before deploying the function, lets create kafka topic:

$ kubeless topic create boot-topic

Deploy the Spring Boot app boot-kubeless-kafka as a Function using docker image available in Docker Hub:

$ kubeless function deploy boot-kubeless-kafka --runtime-image rawsanj/boot-kubeless-kafka --trigger-topic boot-topic --env KUBELESS_KAFKA_SVC=kafka --env  KUBELESS_KAFKA_NAMESPACE=kubeless

Note: If you have your own Kafka Cluster running on K8S, update the namespace and service name in the environment variables above.

Test the function:

Publish message onto the Kafka topic

$ kubeless topic publish --topic boot-topic --data "Hello From Kafka"

Ssh into boot-kubeless-kafka Pods to check the Messages getting consumed.

$ kubectl get pods | grep boot-kubeless-kafka   // Note the Pod name.

$ kubectl logs boot-kubeless-kafka-68b.... // Above pod name.

You can also test the Function from Kubeless UI.