This project is a demo project designed and developed using spring framework and docker+swarm with considering microservices architecture pattern in mind.
Technologies used in this project :
- Spring Cloud
- Spring Fiegn
- Spring Ribbon [Maintenance mode]
- Netflix Eureka
- Circuit Breaker Pattern using Hystrix [Maintenance mode]
- Netflix Zuul [Maintenance mode]
- Spring Sleuth
- Spring Boot 2.0
- Rabbit MQ 3.2
- MySQL 5.6
- Spring Data JPA
- Zipkin
Make sure that you have installed docker on your machine and docker swarm is up and running with at least 1 master and 2 worker.
Then Follow Below Steps to run it in docker environment using docker-swarm .
- Checkout the project using command
# git clone https://github.com/rajeevshukla/spring-microservices-with-docker.git - Navigate inside it
# cd spring-microservices-with-docker - Deploy it on docker-swarm cluster
# docker stack deploy --compose-file docker-compose-swarm.yml mystackand enter. - Scale up the service?
# docker service scale <serviceid>=<noOfServices>This will spawn up all the services and deploy it on all connected node.
When you try to spawn up entire microservice infrastrucure (usually we don't do that in production) and deploy all services using single command what we are doing in this project. It creates problem to maintain the container order. You don't know if you java based service will started first or database on which your java service dependent. To fix that problem we are using wait-for-it.sh which pings dependent container and if that dependent container is up then it starts conatiner initialization otherwise it halts for a moment untill the dependent is not up. Please check docker-compose-swarm.yml file where wait-for-it.sh is being pass in the entry-point.
Monitoring is one of the cross cutting concern that every microservice infra structure has to handle. There are various container monitoring tools available in the market. One of the best open source monitoring tool is Prometheous.
But this won't be enough as it provides information in textual statistics. So we need to represent this information in UI dashbaord format. To do this we need Grafana .
We also need cAdviser which will analyze and exposes resource usage and performance data from running containers. cAdvisor exposes Prometheus metrics out of the box. We will using Grafana as UI dashboard which allows us to query, visualize, alert on and understand your metrics no matter where they stored. You can create, customize, dashbaord with your team and foster a data driven culture.
So we will using this monitoring tool in below flow-
cAdvisor --> Prometheus --> Grafana
| Service Name | Port | Addtional Info |
|---|---|---|
| Cloud Config Server | 8888 | |
| User Service | 8080 | |
| Email Service | 8089 | |
| Eureka Discovery Service | 8762 | |
| Zuul API Gateway | 8889 | |
| Zipkin Server | 9411 | |
| RabbitMQ | 15672 | guest/guest |
| MySQL | 3306 | root/password |
| Prometheus | 9090 | |
| Grafana | 3000 |