User Management System

By: Dicky Novanto

How to install the project

I. User Management System project Installation

3. Install MySQL server
4. Install Redis: create new Redis container by docker-compose up -d
5. Generate 10 millions of user data and insert all the data into table (see section III)
6. For generating the user data, python script is used.
   • Hash algorithm for password: pbkdf2.
   • To get pbkdf2 hash algorithm on Go: go get -u golang.org/x/crypto/pbkdf2
7. For importing the generated user data into SQL into test_user user and test database to table User:

   • mysql --local-infile test -u test_user -p, with password: test_user

   • LOAD DATA LOCAL INFILE 'user.txt' INTO TABLE User;

8. Clone this project’s repository
9. Setup go driver sql: go get -u github.com/go-sql-driver/mysql
10. Setup profiling library: go get -u github.com/pkg/profile
11. Build and run these following servers:
    • picture server (in pictureserver directory)
    • HTTP server (in web directory)
    • TCP server (in tcpserver directory)
12. How to build and run each server:
    • Go to each server directory and enter these commands:
      • go install (this will create an executable binary file with the name of the server's directory and place the binary file to GOBIN)
      • directory name
13. Browse: localhost:8080 to show home page, and the system can be used

II. WRK (HTTP benchmarking tool) Installation:

1. Install WRK: brew install wrk
2. Refer to section How to run performance test to run WRK
3. Details of how to use WRK is provided in the README of WRK Repository (https://github.com/wg/wrk)

III. Generating 10 millions of user data

1. Go to folder scripts
2. Run python script with command: python generateUser.py

How to start User Management System

1. Run TCP server first by entering command: tcpserver
2. Run Picture server with command: pictureserver
3. Run HTTP server with command: httpserver

How to run performance test

1. Go to scripts directory, open terminal
2. Enter this command: wrk -t100 -c1000 -d10s -s post.lua http://localhost:8080/authenticate
   • WRK will be running with the number of thread = 100, the number of keep-alive connection = 1000, duration = 10 seconds and running script post.lua that exist inside the directory, and the endpoint that is tested is http://localhost:8080/authenticate
3. The result provides many information about the performance test, for example the number of request per second.