how to bring up this application

1. clone repository
2. Before using these restful APIs, please setup database as mentioned below.
3. run npm i
4. edit ormconfig.json and change your database configuration (you can also change a database type, but don't forget to install specific database drivers).
5. Note that if the customers table doesn't exists before running the application , the application will create the table during startup process. If you want to create the customers table manually, please use the DDL in the init.sql and don't specify any constraints. This seems to be an issue of typeorm. see typeorm/typeorm#2376. Lastly , you can change the synchronize flag in ormconfig.json to false.
6. use the DML in the init.sql located in sql/ to initialize sample data.
7. run npm start
8. use curl or postman to inovke the restful API, for example, to test login, you can send post request with email and password included in x-wwww-for-urlencoded body to the url http://localhost:3000/login.
9. As for logout, just send post request to http://localhost:3000/logout to destroy session object associated with the current user.
10. As for retriving customer's personal info, send a get request to http://localhost:3000/customers/${id} to get the customer with his id specified in the end of the url.

How to setup DB

1. install docker service in your computer.
2. use the command docker run --name MyPostgres -d -p 5432:5432 -e POSTGRES_USER=admin -e POSTGRES_PASSWORD='12345' postgres to make a postgresql databse up and running.
3. If you use the above command, don't forget to change ormconfig.json accordingly.
4. create test database in the newly created postgres service.

Criteria

1. The database design

+ Please see src/entity/customers.ts

2. The API design

+ We follow standard restful url and http method convention for API operations. Please see point 8, 9 ,10 listed above or controller files under src/controller/

3. Any application and network level security consideration

+ We can: 
+ 1. use IP whiteList
+ 2. use HTTPS with API keys or user credentials
+ 3. use JWT but make sure to clean all possible places in client side, renew token periodically.

4. Explanation on how to scale up and down both API and database to handle variable traffic

+ We can:
+ 1. add index to the table columns which is used as query condition
+ 2. split customers data into different tables based on some criteria or algorithm such as range of IDs or hash value of a ID
+ 3. Based on user daily activities, warm up cache for heavy users.  

5. Considerations on how to enable real time monitoring and alerting onto the service for production support

+ We can:
+ 1. monitor CPU usage, disk I/O, network bandwidth on application servers, database and network devices.
+ 2. In database, make some aggregation report on slow sql operation
+ 3. In applicatiom, make profile on critical operations and log them. use ELK stack to collect such information
+ 4. leverage third party agent.

6. An automatable testing strategy for this service

+ use Mocha and Chia to write some unit tests 

7. instructions to bring up the service

+ Please see the first section

8. Deployment Strategy

+ For application, We can build a Docker file which is made on top of some OS, say ubuntu, and install git, node js in it. After that, we clone the source repository and build source code, bring up the application, and expose port number we want.
+ For database, make a docker file which is based on a postgres and create a database on it. Finally, just expose postgres service.