Spring & Spring Boot stuff

Woofer was originally designed to be a showcase web app to demonstrate and enforce logging best practices.

But actually it is also a quite advanced Spring Boot-based microservices project sample that could be used as a bootstrap to create a microservices application.

This page highlights Spring & Spring Boot modules used in Woofer.

Woofer's microservices

Woofer heavily relies on Spring Boot, and Spring Cloud to implement a microservices architecture.

It involves the following components:

service-registry: the service registry server, basically implemented by Eureka Server.
router: a router and load balancer service, basically implemented by Zuul; only required if you want to load balance the woofer-webfront service.
woofer-webfront: this is the web front-end application, that the end-user will use to connect, read and post woofs.
woofer-backend: this is the backend service, a pure JSON/Rest API service, that is consumed by the woofer-webfront and possibly mobile apps (not part of the demo).
woofer-notifier: this is a notification service, that is asynchronously triggered when domain events occur (a user posts a woof, or subscribes/unsubscribes to a mate). It is called asynchronously by woofer-backend.
database: a basic MariaDB instance
logs centralization system: an ELK instance
distributed tracing system: a basic Zipkin server.

Woofer's profiles

By default, when launching Woofer services locally with no profile, it will use a memory database (h2), output Java and Http logs to the console.

But Woofer has several Spring Boot profiles to adapt to several environments.

Profile	Description	Required environment variables
logstash	used to ship Java and Http logs to a Logstash server in JSON native format	`$LOGSTASH_HOST`, `$LOGSTASH_PORT` and optionally `$MD_PROJECT` (project ID metadata)
h2	uses h2 as database (mem or file)	`$H2_DATASOURCE_URL` (default in memory), `$H2_USER` (default `sa`) and `$H2_PASSWORD` (default none)
mysql	uses MySQL or MariaDB as database	`$MYSQL_DATASOURCE_URL` (default on `localhost`), `$MYSQL_USER` (default `root`) and `$MYSQL_PASSWORD` (default none)
zipkin	activates distributed tracing with a Zipkin server	`$ZIPKIN_URL` and `$ZIPKIN_SAMPLING` (a ratio)
openshift	adapts the Woofer configuration to an OpenShift v3 environment (work in progress)	none
jmx	exposes Spring Boot Actuator metrics through JMX	none

Spring Data JPA

Spring Data JPA is used to implement our Object-Relational Mapping (ORM).

It allows creating Data Access Objects (called Respositories in Spring terminology) simply by annotations and interfaces.

See:

Spring Data Rest

Spring Data Rest is used to build hypermedia-driven REST web services on top of Spring Data repositories.

Woofer also uses the Projections feature.

See:

Spring Cloud Netflix

Spring Cloud Netflix is a great contribution from Netflix, that embeds several features to build cloud applications.

Eureka

Eureka server & client provides a very cheap and robust way of implementing services registration and discovery in a microservices architecture (more info).

Its use is quite straightforward as it only requires adding the right dependency to your pom.xml, and a basic annotation to your Sprint Boot application (either @EnableEurekaClient or @EnableEurekaServer).

Feign

Feign allows implementing declarative REST clients (i.e. through interfaces & annotations).

It is used in woofer-webfront to implement the woofer-backend REST client.

See:

Ribbon

Ribbon provides client side load balancer.

Its use is quite straightforward as it only requires adding the right dependency to your pom.xml.

Spring Boot Actuator

Spring Boot Actuator provides a set of tools to monitor and manage your application.

Apart from default management endpoints, woofer-webfront and woofer-backend publish some business metrics to count active sessions, woofs and subscriptions.

See:

Spring REST Docs

Spring REST Docs is a tool for generating part of REST API documentation.

It is used in woofer-backend project to generate the reference API documentation.

See:

Spring Cloud Sleuth

Spring Cloud Sleuth is used for distributed tracing & logs correlation (see Logging - Code page for more details).