Develop a Food Ordering System using Microservices architecture and Spring Cloud

This is a system designed in Spring Boot/Data/Cloud to provide REST APIs for food ordering/delivery system.

Development environment

• Mongo DB is running on docker, image specified in docker-compose.yml file, and started by docker-compose.
• Spring Data Rest is used for quick DB access and verification.
• Spring Boot is used for fast REST API development and independant deployment.
• Spring Boot Actuator is used to provide monitoring information. (/health, /metrics... etc.)
• HAL Browser is used for quick repository exploration.
• Lombok is used to eliminate constructors and getter/setter implementation for cleaner coding style.
• Spring Cloud is used to provide infrastructure services.
• Eureka is used for microservices registration and discovery.
• Hystrix is used as circuit breaker when failure to prevent avalanche in system.
• RabbitMQ is used to decouple microservices.
• WebSocket is used to send message to UI.
• RestTemplate is used to communicate between microservices.

Features

• User can search a restaurant based on restaurant name.
• User can order food by choosing different menu item, quantity and add a note about his/her diet restrictions and etc.
• User can fill in the delivery address.
• After user places an order, the order should contain the food items user ordered, quantity, price and order time.
• User needs to pay for his/her order by providing credit card number, expiration date and security code.
• After payment is made successfully, it should return payment ID, timesatmp and then the order is considered as completed so the user can see the estimated delivery time.
• Estimated delivery time is a random time between 5 minutes ~ 1 hour.

REST APIs

Restaurant Service

Search restaurant by name

GET localhost:8001/api/restaurants?name=<restaurant_name>
Return
{
    "id": <restaurant_id>,
    "name": <restaurant_name>
}

Create a restaurant

POST localhost:8001/api/restaurants
Request Body
{
    "name": <restaurant_name>
}
Return HttpStatus.CREATED

Get all menu items by restaurant id

GET localhost:8001/api/restaurants/{restaurantId}/menuItems
Return
[
    {
        "id": <menu_item_id>,
        "restaurantId": <restaurantId>,
        "name": <menu_item_name>,
        "description": <menu_item_description>,
        "price": <menu_item_price>
    },
    ...
]

Create menu item

POST localhost:8001/api/restaurants/{restaurantId>/menuItems
Request Body
{
        "restaurantId": <restaurantId>,
        "name": <menu_item_name>,
        "description": <menu_item_description>,
        "price": <menu_item_price>
}
Return HttpStatus.CREATED

Bulk upload menu items

POST localhost:8001/api/restaurants/bulk/menuItems
Request Body
[
    {
        "restaurantId": <restaurantId>,
        "name": <menu_item_name>,
        "description": <menu_item_description>,
        "price": <menu_item_price>
    }
]
Return HttpStatus.CREATED

Order Service

Create an order

POST localhost:8002/api/restaurants/{rid}/orders
{
    "restaurantId": <restaurant_id>,
    "items":
    [
        {
            "name": <menu_item_name>,
            "price": <menu_item_price>,
            "quantity": <# of items>
        },
        ...
    ],
    "userInfo":
    {
        "firstName": <customer_first_name>,
        "lastName": <customer_last_name>,
        "phone": <customer_phone>,
        "address": <customer_address>
    },
    "specialNote": <special_note>
}
Return:
HttpStatus.CREATED
{
    "id": <order_id>,
    "restaurantId": <restaurant_id>,
    "items":
    [
        {
            "name": <menu_item_name>,
            "price": <menu_item_price>,
            "quantity": <# of items>
        },
        ...
    ],
    "userInfo":
    {
        "firstName": <customer_first_name>,
        "lastName": <customer_last_name>,
        "phone": <customer_phone>,
        "address": <customer_address>
    },
    "specialNote": <special_note>,
    "totalPrice": <total_price>,
    "orderTime": <order_time_in_milliseconds>
}

Payment Distribution Service

Payment distribution

POST localhost:8003/api/payments
{
    "orderId": <order_id>,
    "amount": <payment_amount>,
    "creditCardInfo": 
    {
        "firstName": <first_name>,
        "lastName": <lastName>,
        "expiredMonth": <month>,
        "expiredYear": <year>,
        "securityCode": <security_code>
    }
}

Note: Since payment process can be slow and become a bottleneck in a busy environment, this API will send the payment information to message queue for Payment Service to process.

Payment Service

Process payment

POST localhost:8004/api/payments
{
    "orderId": <order_id>,
    "amount": <payment_amount>,
    "creditCardInfo": 
    {
        "firstName": <first_name>,
        "lastName": <lastName>,
        "expiredMonth": <month>,
        "expiredYear": <year>,
        "securityCode": <security_code>
    }
}
Return
HttpStatus.CREATED

Note: **Hystrix** fallbackMethod is added in case processPayment() fails.
Note: this API will store the payment into DB, find the matching order and update the order paymentId, deliveryTime, and notify Order Complete Service with RestTemplate.

Order Complete Updater Service

Order complete

POST localhost:8005/api/orders
{
    "id": <order_id>,
    "items":
    [
        {
            "name": <menu_item_name>,
            "price": <menu_item_price>,
            "quantity": <# of items>
        },
        ...
    ],
    "userInfo":
    {
        "firstName": <customer_first_name>,
        "lastName": <customer_last_name>,
        "phone": <customer_phone>,
        "address": <customer_address>
    },
    "specialNote": <special_note>,
    "totalPrice": <total_order_price>,
    "orderTime": <order_time>,
    "deliveryTime": <food_delivery_time>,
    "paymentId": <payment_id>
}

Note: This API will serialize the order to WebSocket channel: "topic/orders", UI can subscribe to this channel to receive message and display to user.

Getting started

Start MongoDB, RabbitMQ on Docker

docker-compose up -d

Check data in MongoDB

Find mongodb container id

docker ps

Enter mongodb container by typing the first 3 charactters of the container id (ex: '9cd'), then type mongo inside the container to use mongodb shell command.

docker exec -it 9cd bash
# mongo                             // open mongo shell
> use test                          // Spring boot use test db as default
> show collections                  // show all collections inside test db
> db.restaurant.find().pretty()     // show all data inside restaurant table
> exit                              // quit mongo shell
> exit                              // exit container shell

Installation

mvn clean install

Start Eureka

sh ./start-eureka.sh

Start Hystrix

sh ./start-hystrix.sh

Start Restaurant Service

sh ./start-restaurant-service.sh

Start Order Service

sh ./start-order-service.sh

Start Payment Distribution

sh ./start-payment-distribution.sh

Start Payment Service

sh ./start-payment-service.sh

Start Order Complete Updater

sh ./start-order-complete-updater.sh

Upload Test Menu Items

cd restaurant-service
sh ./upload-menu-items.sh

Note: default restaurant id for testing: "11111111-1111-1111-11111111111111111".

Explore by HAL Browser

http://localhost:8001/browser/index.html

port: 8001 can be changed for different services.

Investigate registerd services in Eureka

http://localhost:8761

Investigate message queue in RabbitMQ

http://localhost:15762
Go to Queues -> binder.payments

Checkout application metrics

http://localhost:8005/health
http://localhost:8005/env
http://localhost:8005/metrics
http://localhost:8005/mappings

Note: 8005 can be changed for different services.

Test work flow with PostMan

Create an order

POST localhost:8002/api/restaurants/11111111-1111-1111-11111111111111111/orders
{
    "restaurantId": "11111111-1111-1111-11111111111111111",
    "items": [
        {
            "name": "menuItem 1",
            "price": 11,
            "quantity": 2
        },
        {
            "name": "menuItem 2",
            "price": 12,
            "quantity": 3
        }
    ],
    "userInfo": {
        "firstName": "first1",
        "lastName": "last1",
        "phone": "14081234567",
        "address": "123 stree1 ave, San Jose, CA 95123"
    }
}
Returns:
Returns:
{
    "id": "5903e81327b884525eb9a5be",
    ...
    "totalPrice": 58,
    ...
}

Post a payment for the order

POST localhost:8003/api/payments
{
    "amount": 58,
    "orderId": "5903e81327b884525eb9a5be",
    "creditCardInfo": {
        "firstName": "first 1",
        "lastName": "last 1",
        "expiredMonth": "02",
        "expiredYear": "2019",
        "securityCode": "231"
    }
}

Check RabbitMQ, you will see a message is queued by payment-distribution and consumed by payment-service.

At the log of order-complet-updater, you will see the following message: "Receive order = Order(id=5903e81327b884525eb9a5be, ..." "WebSocketSession[1 current WS(1)-HttpStream(0)-HttpPoll(0), 3 total..."

So the message has been sent out through WebSocket by order-complete-updater.

To see the message with test UI:

localhost:8005
Click on "Subscribe to Order Complete Updates" to subscribe the channel.
Click on "Send Test Message" to send out the test JSON object.

See the message received at the buttom text box.

Subscribed to /topic/orders
sendMessage triggered
{
  "id": "5903e81327b884525eb9a5be",
  "restaurantId": "11111111-1111-1111-11111111111111111",
  "items": [
     {
       "name": "menuItem 1",
       "price": 11,
       "quantity": 2
     },
     {
       "name": "menuItem 2",
       "price": 12,
       "quantity": 3
     }
  "totalPrice": 58,
  "orderTime": 1493428243933,
  "specialNote": "",
  "deliveryTime": 1493486808730,
  "paymentId": "",
  "userInfo": {
     "id": "",
     "firstName": "first1",
     "lastName": "last1",
     "phone": "14081234567",
     "address": "123 stree1 ave, San Jose, CA 95123"
  }
}

Test fallback

Open Hystrix dashboard

localhost:7979

Monitor order-complete-updater

http://localhost:8004/hystrix.stream

Stop order-complete-updater. Post a payment to payment-service. Watch processPayment error rate jump from 0.0 to 100%. Watch error message from fallback method in log. Keep posting the same payment again and again, eventually, see the Circuit status changed to "Opened" from "Closed."

LICENSE

MIT