Problems caused by Gil problems and Solutions

python Because of its global interpreter lock GIL We can't achieve real parallel computing through threads . Need to be solved  IO intensive and   Computationally intensive   The parallel operation problem of .

IO intensive ： Read the file , Read network Socket frequent . 

Computationally intensive ： A lot of consumption CPU Mathematical and logical operations of , That is what we call parallel computation here .

IO intensive ： The solution is synergy , frequently-used greenlet 、gevent、ayncio Wait to deal with

Simple understanding See connection below ：

10 Minute quick understanding python asynchronous asyncio_ Xiaosheng listens to Yuyuan's blog -CSDN Blog _asyncio python

See you further   Link below

 

Computationally intensive ： Commonly used concurrent.futures Module to handle

concurrent.futures modular , You can use multiprocessing To achieve true parallel computing .

The core principle is ：concurrent.futures In the form of subprocesses , Run multiple... In parallel python Interpreter , So that python Programs can take advantage of multicore CPU To speed up execution . Because subprocesses are separated from the main interpreter , So their global interpreter locks are also independent of each other . Each subprocess can use a complete CPU kernel .

python concurrent.futures_weixin_30394981 The blog of -CSDN Blog

Python in concurrent.futures Module instruction _ Tuanzi Da Yuan Shuai's blog -CSDN Blog _concurrent.futures