我脚本里的异步非阻塞每个子进程的CPU占用率都很低怎么办，我想都跑满应该怎么做？

Question

我在学习python异步非阻塞的时候遇到了一些问题，处理数据的时候我在进程池创建了64个进程，运行了大概半个多小时，然后进程CPU使用率变为0了，进程全部休眠，只剩下一两个python3存在CPU占用率，所有操作好像都停下了，需要写入的文件末尾不在新增。这是为什么呢？我应该怎么解决。除此之外，就是每个进程的CPU占用率一直比较低，我怎么才能跑满每一核？

# python
# -*- encoding: utf-8 -*-
'''
@File        :read.py
@Time        :2022/02/10 03:30:02
@Author        :charles kiko
@Version        :1.0
@Contact        :[e-mail]
@Desc        :扫描基因组内的重复序列 python read.py lens fasta out_file
'''
import os
import gc
import sys
import csv
import numpy as np
import pandas as pd
from pandas import Series, DataFrame
from Bio import SeqIO
import multiprocessing # Step I : 导入模块
from multiprocessing import cpu_count#读取CPU核心数用于匹配线程数

def count_str(seq):
    str_dic = {'A':0,'T':0,'C':0,'G':0}
    for i in seq:
        str_dic[i] = str_dic[i] + 1
    if str_dic['G'] + str_dic['C'] == 0:
        return 0
    else:
        return (str_dic['G'] + str_dic['C']) / len(seq)

def find(seq,name,start,end):
    global DNA_dict# 全局变量
    global contig
    global contigs_0
    global dup
    # print(seq)
    file = open('out.out', 'a+')
    file.write(name+'_'+str(start)+'_'+str(end)+'\n')
    file.close()
    name1 = name+str(start)+str(end)
    GC = count_str(seq)
    for contig_ in contig:
        if contig_ not in DNA_dict.keys():
            print(contig_,'not in keys!')
            gc.collect()
            # continue
            return 0
        top = str(DNA_dict[contig_])
        left = str(seq)
        for i in range(len(top) - len(left) + 1):
            if i + len(left) >= len(top):
                continue
            if top[i:i + len(left)] == left:
                if contig_ == name and start == i:
                    continue
                name2 = contig_+str(i)+str(i+len(left))
                if name1 not in dup:
                    dup.append(name1)
                    lt = [contig_,name1,start,end,len(left),GC,seq]
                    file = open(sys.argv[3], 'a+')
                    file.write('\t'.join([str(m) for m in lt])+'\n')
                if name2 not in dup:
                    dup.append(name2)
                    lt = [contig_,name2,i,i+len(left),len(left),GC,seq]
                    file = open(sys.argv[3], 'a+')
                    file.write('\t'.join([str(m) for m in lt])+'\n')
                # print(seq)
                file = open('out.out', 'a+')
                file.write(contig_+'*'+str(i)+'*'+str(i+len(left))+'\n')
                file.close()
                del lt
                gc.collect()
                file.close()
        del top,left
        gc.collect()
    return 0

if __name__ == '__main__' :#多进程
    length = 5
    dup=[]
    contigs = {}
    contigs_0 = {}
    for i in open(sys.argv[1],'r'):
        if i != '\n':
            lt = i.strip('\n').split()
            contigs[str(lt[0])] = int(lt[1])
    DNA_dict = SeqIO.to_dict(SeqIO.parse(sys.argv[2], "fasta"))# 提取之后直接返回字典
    for key in DNA_dict.keys():
        DNA_dict[key] = str(DNA_dict[key].seq).upper()
    contigs_ = {k: v for k, v in sorted(contigs.items(), key=lambda item: item[1], reverse=True)}
    contig = list(contigs_.keys())
    # print(contig)
    file = open('out.out', 'w')
    file.close()
    file = open(sys.argv[3], 'w')
    file.write(
        '\t'.join(['contig_name','name','start', 'end','seq_length','GC','SEQ'])+'\n')
    file.close()
    pool = multiprocessing.Pool(processes = 8) # Step II : 进程池
    for i in contig:
        if i not in DNA_dict.keys():
            continue
        for j in range(5,len(DNA_dict[i])):
            for m in range(len(DNA_dict[i]) - j + 1):
                contigs_0[str(DNA_dict[i][m:m + j])+str(m)+str(m+j)] = 1
                pool.apply_async(find, (DNA_dict[i][m:m + j],i,m,m+j,), )  # Step III : 异步（并行）计算
    pool.close() # Step IV : 准备结束
    pool.join() # Step IV : 完全结束

Answer 1

不跑满是件好事。你的CPU全占满了的话，程序运行会特别慢，而且伤风扇和电脑各个配件的使用寿命