Piggy Metrics

Piggy Metrics is a simple financial advisor app built to demonstrate the Microservice Architecture Pattern using Spring Boot, Spring Cloud and Docker. The project is intended as a tutorial, but you are welcome to fork it and turn it into something else!

Functional services

Piggy Metrics is decomposed into three core microservices. All of them are independently deployable applications organized around certain business domains.

Account service

Contains general input logic and validation: incomes/expenses items, savings and account settings.

Method	Path	Description	User authenticated	Available from UI
GET	/accounts/{account}	Get specified account data
GET	/accounts/current	Get current account data	×	×
GET	/accounts/demo	Get demo account data (pre-filled incomes/expenses items, etc)		×
PUT	/accounts/current	Save current account data	×	×
POST	/accounts/	Register new account		×

Statistics service

Performs calculations on major statistics parameters and captures time series for each account. Datapoint contains values normalized to base currency and time period. This data is used to track cash flow dynamics during the account lifetime.

Method	Path	Description	User authenticated	Available from UI
GET	/statistics/{account}	Get specified account statistics
GET	/statistics/current	Get current account statistics	×	×
GET	/statistics/demo	Get demo account statistics		×
PUT	/statistics/{account}	Create or update time series datapoint for specified account

Notification service

Stores user contact information and notification settings (reminders, backup frequency etc). Scheduled worker collects required information from other services and sends e-mail messages to subscribed customers.

Method	Path	Description	User authenticated	Available from UI
GET	/notifications/settings/current	Get current account notification settings	×	×
PUT	/notifications/settings/current	Save current account notification settings	×	×

Notes

Each microservice has its own database, so there is no way to bypass API and access persistence data directly.
MongoDB is used as a primary database for each of the services.
All services are talking to each other via the Rest API

Infrastructure

Spring cloud provides powerful tools for developers to quickly implement common distributed systems patterns -

Config service

Spring Cloud Config is horizontally scalable centralized configuration service for the distributed systems. It uses a pluggable repository layer that currently supports local storage, Git, and Subversion.

In this project, we are going to use native profile, which simply loads config files from the local classpath. You can see shared directory in Config service resources. Now, when Notification-service requests its configuration, Config service responses with shared/notification-service.yml and shared/application.yml (which is shared between all client applications).

Client side usage

Just build Spring Boot application with spring-cloud-starter-config dependency, autoconfiguration will do the rest.

Now you don't need any embedded properties in your application. Just provide bootstrap.yml with application name and Config service url:

spring:
  application:
    name: notification-service
  cloud:
    config:
      uri: http://config:8888
      fail-fast: true

With Spring Cloud Config, you can change application config dynamically.

For example, EmailService bean is annotated with @RefreshScope. That means you can change e-mail text and subject without rebuild and restart the Notification service.

First, change required properties in Config server. Then make a refresh call to the Notification service: curl -H "Authorization: Bearer #token#" -XPOST http://127.0.0.1:8000/notifications/refresh

You could also use Repository webhooks to automate this process

Notes

@RefreshScope doesn't work with @Configuration classes and doesn't ignores @Scheduled methods
fail-fast property means that Spring Boot application will fail startup immediately, if it cannot connect to the Config Service.

Auth service

Authorization responsibilities are extracted to a separate server, which grants OAuth2 tokens for the backend resource services. Auth Server is used for user authorization as well as for secure machine-to-machine communication inside the perimeter.

In this project, I use Password credentials grant type for users authorization (since it's used only by the UI) and Client Credentials grant for service-to-service communciation.

Spring Cloud Security provides convenient annotations and autoconfiguration to make this really easy to implement on both server and client side. You can learn more about that in documentation.

On the client side, everything works exactly the same as with traditional session-based authorization. You can retrieve Principal object from the request, check user roles using the expression-based access control and @PreAuthorize annotation.

Each PiggyMetrics client has a scope: server for backend services and ui - for the browser. We can use @PreAuthorize annotation to protect controllers from an external access:

@PreAuthorize("#oauth2.hasScope('server')")
@RequestMapping(value = "accounts/{name}", method = RequestMethod.GET)
public List<DataPoint> getStatisticsByAccountName(@PathVariable String name) {
	return statisticsService.findByAccountName(name);
}

API Gateway

API Gateway is a single entry point into the system, used to handle requests and routing them to the appropriate backend service or by aggregating results from a scatter-gather call. Also, it can be used for authentication, insights, stress and canary testing, service migration, static response handling and active traffic management.

Netflix opensourced such an edge service and Spring Cloud allows to use it with a single @EnableZuulProxy annotation. In this project, we use Zuul to store some static content (the UI application) and to route requests to appropriate the microservices. Here's a simple prefix-based routing configuration for the Notification service:

zuul:
  routes:
    notification-service:
        path: /notifications/**
        serviceId: notification-service
        stripPrefix: false

That means all requests starting with /notifications will be routed to the Notification service. There is no hardcoded addresses, as you can see. Zuul uses Service discovery mechanism to locate Notification service instances and also Circuit Breaker and Load Balancer, described below.

Service Discovery

Service Discovery allows automatic detection of the network locations for all registered services. These locations might have dynamically assigned addresses due to auto-scaling, failures or upgrades.

The key part of Service discovery is the Registry. In this project, we use Netflix Eureka. Eureka is a good example of the client-side discovery pattern, where client is responsible for looking up the locations of available service instances and load balancing between them.

With Spring Boot, you can easily build Eureka Registry using the spring-cloud-starter-eureka-server dependency, @EnableEurekaServer annotation and simple configuration properties.

Client support enabled with @EnableDiscoveryClient annotation a bootstrap.yml with application name:

spring:
  application:
    name: notification-service

This service will be registered with the Eureka Server and provided with metadata such as host, port, health indicator URL, home page etc. Eureka receives heartbeat messages from each instance belonging to the service. If the heartbeat fails over a configurable timetable, the instance will be removed from the registry.

Also, Eureka provides a simple interface where you can track running services and a number of available instances: http://localhost:8761

Load balancer, Circuit breaker and Http client

Ribbon

Ribbon is a client side load balancer which gives you a lot of control over the behaviour of HTTP and TCP clients. Compared to a traditional load balancer, there is no need in additional network hop - you can contact desired service directly.

Out of the box, it natively integrates with Spring Cloud and Service Discovery. Eureka Client provides a dynamic list of available servers so Ribbon could balance between them.

Hystrix

Hystrix is the implementation of Circuit Breaker Pattern, which gives us a control over latency and network failures while communicating with other services. The main idea is to stop cascading failures in the distributed environment - that helps to fail fast and recover as soon as possible - important aspects of a fault-tolerant system that can self-heal.

Moreover, Hystrix generates metrics on execution outcomes and latency for each command, that we can use to monitor system's behavior.

Feign

Feign is a declarative Http client which seamlessly integrates with Ribbon and Hystrix. Actually, a single spring-cloud-starter-feign dependency and @EnableFeignClients annotation gives us a full set of tools, including Load balancer, Circuit Breaker and Http client with reasonable default configuration.

Here is an example from the Account Service:

@FeignClient(name = "statistics-service")
public interface StatisticsServiceClient {

	@RequestMapping(method = RequestMethod.PUT, value = "/statistics/{accountName}", consumes = MediaType.APPLICATION_JSON_UTF8_VALUE)
	void updateStatistics(@PathVariable("accountName") String accountName, Account account);

}

Everything you need is just an interface
You can share @RequestMapping part between Spring MVC controller and Feign methods
Above example specifies just a desired service id - statistics-service, thanks to auto-discovery through Eureka

Monitor dashboard

In this project configuration, each microservice with Hystrix on board pushes metrics to Turbine via Spring Cloud Bus (with AMQP broker). The Monitoring project is just a small Spring boot application with the Turbine and Hystrix Dashboard.

Let's see observe the behavior of our system under load: Statistics Service imitates a delay during the request processing. The response timeout is set to 1 second:


`0 ms delay`	`500 ms delay`	`800 ms delay`	`1100 ms delay`
Well behaving system. Throughput is about 22 rps. Small number of active threads in the Statistics service. Median service time is about 50 ms.	The number of active threads is growing. We can see purple number of thread-pool rejections and therefore about 40% of errors, but the circuit is still closed.	Half-open state: the ratio of failed commands is higher than 50%, so the circuit breaker kicks in. After sleep window amount of time, the next request goes through.	100 percent of the requests fail. The circuit is now permanently open. Retry after sleep time won't close the circuit again because a single request is too slow.

Log analysis

Centralized logging can be very useful while attempting to identify problems in a distributed environment. Elasticsearch, Logstash and Kibana stack lets you search and analyze your logs, utilization and network activity data with ease.

Distributed tracing

Analyzing problems in distributed systems can be difficult, especially trying to trace requests that propagate from one microservice to another.

Spring Cloud Sleuth solves this problem by providing support for the distributed tracing. It adds two types of IDs to the logging: traceId and spanId. spanId represents a basic unit of work, for example sending an HTTP request. The traceId contains a set of spans forming a tree-like structure. For example, with a distributed big-data store, a trace might be formed by a PUT request. Using traceId and spanId for each operation we know when and where our application is as it processes a request, making reading logs much easier.

The logs are as follows, notice the [appname,traceId,spanId,exportable] entries from the Slf4J MDC:

2018-07-26 23:13:49.381  WARN [gateway,3216d0de1384bb4f,3216d0de1384bb4f,false] 2999 --- [nio-4000-exec-1] o.s.c.n.z.f.r.s.AbstractRibbonCommand    : The Hystrix timeout of 20000ms for the command account-service is set lower than the combination of the Ribbon read and connect timeout, 80000ms.
2018-07-26 23:13:49.562  INFO [account-service,3216d0de1384bb4f,404ff09c5cf91d2e,false] 3079 --- [nio-6000-exec-1] c.p.account.service.AccountServiceImpl   : new account has been created: test

appname: The name of the application that logged the span from the property spring.application.name
traceId: This is an ID that is assigned to a single request, job, or action
spanId: The ID of a specific operation that took place
exportable: Whether the log should be exported to Zipkin

Infrastructure automation

Deploying microservices, with their interdependence, is much more complex process than deploying a monolithic application. It is really important to have a fully automated infrastructure. We can achieve following benefits with Continuous Delivery approach:

The ability to release software anytime
Any build could end up being a release
Build artifacts once - deploy as needed

Here is a simple Continuous Delivery workflow, implemented in this project:

In this configuration, Travis CI builds tagged images for each successful git push. So, there are always the latest images for each microservice on Docker Hub and older images, tagged with git commit hash. It's easy to deploy any of them and quickly rollback, if needed.

Let's try it out

Note that starting 8 Spring Boot applications, 4 MongoDB instances and a RabbitMq requires at least 4Gb of RAM.

Before you start

Install Docker and Docker Compose.
Change environment variable values in .env file for more security or leave it as it is.
Build the project: mvn package [-DskipTests]

Production mode

In this mode, all latest images will be pulled from Docker Hub. Just copy docker-compose.yml and hit docker-compose up

Development mode

If you'd like to build images yourself, you have to clone the repository and build artifacts using maven. After that, run docker-compose -f docker-compose.yml -f docker-compose.dev.yml up

docker-compose.dev.yml inherits docker-compose.yml with additional possibility to build images locally and expose all containers ports for convenient development.

If you'd like to start applications in Intellij Idea you need to either use EnvFile plugin or manually export environment variables listed in .env file (make sure they were exported: printenv)

Important endpoints

http://localhost:80 - Gateway
http://localhost:8761 - Eureka Dashboard
http://localhost:9000/hystrix - Hystrix Dashboard (Turbine stream link: http://turbine-stream-service:8080/turbine/turbine.stream)
http://localhost:15672 - RabbitMq management (default login/password: guest/guest)

Contributions are welcome!

PiggyMetrics is open source, and would greatly appreciate your help. Feel free to suggest and implement any improvements.

daizf/piggymetrics

Piggy Metrics

Functional services

Account service

Statistics service

Notification service

Notes

Infrastructure

Config service

Client side usage

With Spring Cloud Config, you can change application config dynamically.

Notes

Auth service

API Gateway

Service Discovery

Load balancer, Circuit breaker and Http client

Ribbon

Hystrix

Feign

Monitor dashboard

Log analysis

Distributed tracing

Infrastructure automation

Let's try it out

Before you start

Production mode

Development mode

Important endpoints

Contributions are welcome!