Douyin Project

Introduction

This is a backend project about simple-douyin of Bytedance, using gin gorm and so on.

APIS

• POST /douyin/user/login
• POST douyin/user/register
• GET /douyin/user/
• GET /douyin/feed

How to develop

Core

The core code is in /app, which has four parts:

• controller: it serves as apis' handlers.
• service: it's main business logic.
• models: it's the database models and functions(dao) to operate data.
• define: it's the data structure for each request and response.

From top to bottom layers: Router -> Controller -> Service(define) -> Dao -> Model

Database

Database is initialized in /init/database.go.

Mysql

The main database is mysql, storing all persistent data. Its control is integrated in gorm.

Tips: you can create your datatables in database by gorm.AutoMigrate(...)

Redis

The cache database is redis in which data react quickly. But it's just used in api GET /douyin/user.

You can run benchmark in /tests/redis_test.go to see the performance.

Config

You can edit the config.yaml to add another setting, and use in project like this:

// learn how to use viper in go
viper.Get("new_config")

Test

Feature Test

For each finished apis, it needs to be tested features in tests folder.

For each utils, it needs to be tested basicly in the same level folder.

The standards are as follows:

• Test all error code's examples which the apis may response.
• Write notes above this line or subtest including name to explain this test.

Benchmark

For some modules, you can write benchmark in test files. Such as redis and mysql benchmark.

goos: linux
goarch: amd64
pkg: douyin/tests
cpu: Intel(R) Xeon(R) Platinum 8255C CPU @ 2.50GHz
BenchmarkRedis
BenchmarkRedis-4   	   10000	    112403 ns/op	     188 B/op	       7 allocs/op
BenchmarkMysql
BenchmarkMysql-4   	    3570	    300697 ns/op	    9013 B/op	     134 allocs/op
PASS
ok  	douyin/tests	2.291s

Pressure Test

You can test the QPS by some tools, such as ab/wrk/...

The folder wrk is several scripts of lua, used to presssure tests.

Run the command in bash: wrk -t12 -c400 -d10s -T30 -s ./wrk/login.lua http://127.0.0.1:8080

Here is the result:

Running 10s test @ http://127.0.0.1:8080
  12 threads and 400 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency    45.82ms   19.46ms 229.79ms   80.51%
    Req/Sec   733.62    134.87     1.19k    81.42%
  87715 requests in 10.09s, 14.89MB read
Requests/sec:   8694.54
Transfer/sec:      1.48MB