falsk + gunicorn achieve concurrency

-Gunicorn does not support WIN, all are done under Linux.

-Just run the script like this if __name__ == "__main__":
    #app.run(debug=False, host='0.0.0.0', port=8886, threaded=True)
    app.run()

- pip install gunicorn 

- gunicorn -w 4 -b 0.0.0.0:8080 main:app

-w is the number of work, set four -b to specify the port, main is the name of your script, app is the name of your flask application.

Figure 4 process

Video memory is naturally 4 times.

 

 supplement:

pip install vent

New configuration file: vim gunicorn.py

import os
import gevent.monkey
gevent.monkey.patch_all()
import multiprocessing

# backlog = 512    #监听队列
# debug = True  # debug开启
# threads = 2 # 指定每个进程开启的线程数
# daemon = True # 表示开启后台运行,默认False
loglevel = 'debug'
bind = "0.0.0.0:5001" # 绑定的ip和端口号

# /var/log/WEB_APP 目录必须存在
# pidfile = "/var/log/WEB_APP/gunicorn.pid"
# accesslog = "/var/log/WEB_APP/access.log"
# access_log_format = '%(t)s %(p)s %(h)s "%(r)s" %(s)s %(L)s %(b)s %(f)s" "%(a)s"' 

#设置守护进程
#daemon=True

#设置超时时间120s。默认为30s
timeout=120

#设置访问日志与错误信息日志路径
errorlog = "./error.log"

# 启动的进程数
#workers = multiprocessing.cpu_count() # 这里取的是CPU的数量
workers = 4   # 进程数，

worker_class = 'gevent'   # worker_class是指开启的每个工作进程的模式类型，默认为sync模式，这里使用gevent模式

x_forwarded_for_header = 'X-FORWARDED-FOR'

Run gunicorn manager02:app -c gunicorn.py (manager02 is the name of the script, app is the name of the application in flask) really high concurrency (until the CPU bottleneck is reached?)

 

## 2021 01 25 

- fastapi

- pip install fastapi 
pip install uvicorn

uvicorn fast_api:app --workers 3 --port 5000