/event-driven-vg

Primary LanguageJava

Event Carried State Transfer – Microservice Design Pattern

We can maintain data consistency across all the microservices using Apache/Confluent Kafka. This approach avoids many unnecessary network calls among microservices, improves the performance of microservices and make the microservices loosely coupled. For ex: Order-service does not have to be up and running when user details are updated via user-service. User-service would be raising an event. Order-service can subscribe to that whenever it is up and running. So that information is not going to be lost! In the old approach, it makes microservices tightly coupled in such a way that all the dependent microservices have to be up and running together. Otherwise it would make the system unavailable.

CREATE TABLE `test`.`users` (
  `id` INT NOT NULL,
  `firstname` VARCHAR(45) NULL,
  `lastname` VARCHAR(45) NULL,
  `email` VARCHAR(45) NULL,
  PRIMARY KEY (`id`));
  
INSERT INTO `test`.`users` (`id`, `firstname`, `lastname`, `email`) VALUES ('1', 'Neha', 'Parate', 'neha.parate@gmail.com');
INSERT INTO `test`.`users` (`id`, `firstname`, `lastname`, `email`) VALUES ('2', 'Aravind', 'Dekate', 'aravind.dekate@gmail.com');
INSERT INTO `test`.`users` (`id`, `firstname`, `lastname`, `email`) VALUES ('3', 'Mayur', 'Devghare', 'mayur.devghare@gmail.com');
INSERT INTO `test`.`users` (`id`, `firstname`, `lastname`, `email`) VALUES ('4', 'Suchita', 'Vinchurkar', 'suchita.vinchurkar@gmail.com');


CREATE TABLE `test`.`product` (
  `id` INT NOT NULL,
  `description` VARCHAR(500) NULL,
  `price` INT NULL,
  `qty_available` INT NULL,
  PRIMARY KEY (`id`));

INSERT INTO `test`.`product` (`id`, `description`, `price`, `qty_available`) VALUES ('1', 'IPad', '300', '10');
INSERT INTO `test`.`product` (`id`, `description`, `price`, `qty_available`) VALUES ('2', 'IPhone', '650', '50');
INSERT INTO `test`.`product` (`id`, `description`, `price`, `qty_available`) VALUES ('3', 'Sony TV', '320', '100');

MySQL

mysql> use test;
Database changed
mysql> show tables;
+----------------+
| Tables_in_test |
+----------------+
| product        |
| users          |
+----------------+
2 rows in set (2.56 sec)

mysql> select * from users;
+----+-----------+------------+------------------------------+
| id | firstname | lastname   | email                        |
+----+-----------+------------+------------------------------+
|  1 | Neha      | Parate     | neha.parate@gmail.com        |
|  2 | Aravind   | Dekate     | aravind.dekate@gmail.com     |
|  3 | Mayur     | Devghare   | mayur.devghare@gmail.com     |
|  4 | Suchita   | Vinchurkar | suchita.vinchurkar@gmail.com |
+----+-----------+------------+------------------------------+
4 rows in set (1.06 sec)

mysql> select * from products;
ERROR 1146 (42S02): Table 'test.products' doesn't exist
mysql> select * from product;
+----+-------------+-------+---------------+
| id | description | price | qty_available |
+----+-------------+-------+---------------+
|  1 | IPad        |   300 |            10 |
|  2 | IPhone      |   650 |            50 |
|  3 | Sony TV     |   320 |           100 |
+----+-------------+-------+---------------+
3 rows in set (1.12 sec)

MongoDB

db.getCollection('purchase_order').find({})
{
    "_id" : ObjectId("60a536412306ab5bdd7a6b06"),
    "user" : {
        "id" : NumberLong(1),
        "firstname" : "Parate",
        "lastname" : "Parate",
        "email" : "neha.parate@gmail.com"
    },
    "product" : {
        "id" : NumberLong(1),
        "description" : "ipad"
    },
    "price" : 300.0,
    "_class" : "com.example.demo.model.PurchaseOrder"
}

Now when user details have been updated, user-service will raise an event and iterested order-service will listen to that event update the information.

For ex: For Neha as user, looking to update her email address from gmail.com to hotmail.com.

POST Request

curl --location --request PUT 'http://localhost:8080/user-service/update' \
--header 'Content-Type: application/json' \
--data-raw '{
    "id" : 1,
    "firstname" : "Neha",
    "lastname" : "Parate",
    "email" : "neha.parate@hotmail.com"
}'

Now MySQL DB has been updated and kafka event have been raise and listen by order-service to update the purchase_order collection. Now check purchase_ordershould have updated details

image

MySQL

mysql> select * from users;
+----+-----------+------------+------------------------------+
| id | firstname | lastname   | email                        |
+----+-----------+------------+------------------------------+
|  1 | Parate    | Parate     | neha.parate@hotmail.com      |
|  2 | Aravind   | Dekate     | aravind.dekate@gmail.com     |
|  3 | Mayur     | Devghare   | mayur.devghare@gmail.com     |
|  4 | Suchita   | Vinchurkar | suchita.vinchurkar@gmail.com |
+----+-----------+------------+------------------------------+
4 rows in set (0.00 sec)

Control Center Confluent Kafka

image

http://localhost:8081/order-service/all

[
    {
        "id": "60a536412306ab5bdd7a6b06",
        "user": {
            "id": 1,
            "firstname": "Parate",
            "lastname": "Parate",
            "email": "neha.parate@hotmail.com"
        },
        "product": {
            "productId": 1,
            "description": "ipad"
        },
        "price": 300.0
    }
]

Architecture

image