为什么使用python的多线程mysql查询花费这么多时间

Question

有一个sql，查询时间为3s。我认为，如果我创建3个线程，则每个线程几乎同时运行此sql，整个时间应几乎为3s。但实际上需要8秒为什么？

这样的代码：

import pymysql
import threading
import time
from config.db import db_config

def get_connection():
    return pymysql.connect(**db_config)

def test(sql):
    print('start query...')
    st = time.time()
    conn = get_connection()
    cur = conn.cursor()
    cur.execute(sql)
    print('end query.. timeuse: ', time.time() - st)

def get_threads(num=3):
    threads = []
    for i in range(num):
        threads.append(threading.Thread(target=test, args=(sql,)))
    return threads

sql = "a sql cost 3s"

def main(num):
    threads = get_threads(num)
    for thread in threads:
        thread.start()
    for thread in threads:
        thread.join()

运行结果：

>>> main(1)
start query...
end query.. timeuse:  3.0793652534484863

>>> main(3)
start query...
start query...
start query...
end query.. timeuse:  7.348235368728638
end query.. timeuse:  7.411351919174194
end query.. timeuse:  8.129048585891724

备注：mysql引擎是innodb，max_connections是12384