Architecture Prototype of a Spring Cloud Application on GCP - Blog Application

The blog application is a fictitious blog system demonstrating microservices as well as domain-driven-design.

The system provides the functionality of creating blogs, posts and comments as well as user authentification and authorisation. Furthermore, statistics over activities are kept.

Architecture Overview

The following diagram shows an overview of the blog microservices, datastores and the interaction between these components, as well as the used technologies of each.

Microservices and components running in a Kubernetes Cluster:

• Config Service

  The Config Service backend is a Spring Cloud Config server. It provides the configuration for most of the microservices. On any time, they can pull their configuration.

• Blog Service

  The Blog Service backend is a Spring Boot application. It is the core domain and manages the blogs and posts. On startup it loads the configuration of the Config Service. To store the blog data, it uses the Blog Database, a Google Cloud Datastore and caches the blog data in Blog Cache a Google Cloud Memorystore for Redis. It also interacts with the Event Bus, a Google Cloud PubSub, where other microservices can subscribe to.

• Auth Service

  The Auth Service backend is a Spring Boot application. It is responsible for the creation and verification of user accounts over JSON Web Tokens (JWT). On startup it loads the configuration of the Config Service. To store the user data, it uses the User Database, a PostgreSQL database in Google Cloud SQL.

• Comment Service

  The Comment Service backend is a Spring Boot application. It is responsible for the comments of the blog posts. On startup it loads the configuration of the Config Service. To store the comment data, it uses the Blog Database.

• Statistic Service

  The Statistic Service backend is a Spring Boot application. It is responsible for keeping statistic over blog activities, such as viewing a blog post. On startup it loads the configuration of the Config Service and subscribes to the Event Bus to get the events. To store the statistic data, it uses the Statistic Database, a Google Cloud Datastore.

• Gateway / Router

  The Gateway / Router backend is a Spring Cloud Gateway server. It is responsible for the routing to the Frontend as well to the microservices. On startup it loads the configuration of the Config Service. The Gateway / Router is accessible through the HTTP Router, a Google Cloud Ingress.

• Frontend

  The Frontend is a static Angular webpage with Bootstrap, served by a NGINX webserver.

Getting Started

The detailed instruction for each service can be found in the README file of the associated subfolder.

Prequisites for cloud build

• Docker installed and runnable without sudo (https://docs.docker.com/install/linux/linux-postinstall/)
• Google Cloud SDK installed
• Google Cloud Account (You'll get a 300$ test budget if you connect a new account to GCP)
• Java 8 installed
• Node (npm) installed

Configure Google Cloud SDK

Authorize gcloud:

gcloud auth login
gcloud auth application-default login

Install kubectl

gcloud components install kubectl

Configure docker

gcloud auth configure docker

Build

Make sure that the prequisites are fullfilled

Open variables.properties and change GCP_PROJECT_NAME to a unique name, which isn't used in any other google cloud project.

Local Build

NOTE: The setup-local script may take up to 5 minutes.

1. Run setup-local
2. The script will open the webbrowser for your project to activate payment for this project. Press any key to continue.
3. Navigate to http://localhost:8181 to open the Blog-Application.

The local build will generate following:

• Google Cloud Project with name and zone configured in variables.properties and payment enabled
• App Engine
• Enable datastore.googleapis.com API
• Run projects locally

The Google Cloud Project is on the local build only needed for the Datastore.

Cloud Build

NOTE: The setup-cloud script may take up to 20 minutes.

1. Run setup-cloud.bat or setup-cloud.sh
2. The script will open the webbrowser for your project to activate payment for this project. Press any key to continue when payment is activated.
3. Go to https://console.cloud.google.com to your project (you may must reload the whole page to see kubernetes cluster). Navigate to Kubernetes Engine > Services and click on the gateway-microservice url to open the Blog-Application.

The cloud build will generate following:

• Google Cloud Project with name and zone configured in variables.properties and payment enabled
• App Engine
• Enable compute.googleapis.com API
• Enable container.googleapis.com API
• Enable datastore.googleapis.com API
• Enable storage-component.googleapis.com API
• Enable pubsub.googleapis.com API
• Enable redis.googleapis.com API
• Enable sqladmin.googleapis.com API
• Create Kubernetes Cluster application
• Deploy all microservices

Cleanup

To prevent the project to constantly create you costs, you can simply delete the whole project and you won't have any further costs:

gcloud projects delete [GCP_PROJECT_NAME]

Devtools

If you are inerested in what development tools could be used, have a look at the Devtools Folder.

Contributors

• Moritz Habegger
• Micha Schena

License

This project is made available under the MIT License. See the LICENSE file for the full license.

The logos used in the diagrams are registered trademarks of Google LLC (Google Cloud), Pivotal Software Inc (Spring), Docker Inc. (Docker), Salvatore Sanfilippo (Redis) and Igor Sysoev (nginx).