多线程读取mysql中一张2.4T的表并dump为文件

最近公司一个mysql集群要关闭，需要将一张表中的数据dump下来，八个节点，每个节点300g，总共2.4t，由于没有服务器密码只能通过脚本的方式将数据dump下来

from multiprocessing.dummy import Pool as TPool
import traceback
import MySQLdb
import pandas as pd
from dbutils.pooled_db import PooledDB


class LoadFromMysql():

    def __init__(self):
        self.mysql_pool = PooledDB(
            creator=MySQLdb, mincached=5, maxcached=10,
            host='******', port=****, user='*****', passwd='*****', charset='utf8'
        )
        self.db = self.mysql_pool.connection()

    # 将数据库查询出的元组转为df
    def get_df_from_db(self, sql):
        cursor = self.db.cursor()
        cursor.execute(sql)
        data = cursor.fetchall()
        columnDes = cursor.description  # 获取连接对象的描述信息
        columnNames = [columnDes[i][0] for i in range(len(columnDes))]
        df = pd.DataFrame([list(i) for i in data], columns=columnNames)
        return df

    def extract_data(self, wtid, protocolid):
        sql_pro = (
            f'select descrcn from config.propaths where transtype =2 and descrcn regexp "平均风速|最小风速|最大风速|平均有功功率|最小有功功率|最大有功功率|平均无功功率|最小无功功率|最大无功功率" and'
            ' protocolid = {protocolid} order by offsets ASC'.format(protocolid=protocolid))
        pro_df = self.get_df_from_db(sql_pro)
        if not pro_df.empty:
            cloums = ['wfid', 'wtid', 'rectime']
            for descrcn in pro_df["descrcn"].values:
                cloums.append(descrcn)
            # 得到列长度
            cloums_size = len(cloums)
            statisticdata_sql = 'select wfid,wtid,rectime,id0,id1,id2,id3,id4,id5,id6,id7,id8,id9,id10,id11,id12 from dbo.statisticdata where wtid = {wtid}'.format(
                wtid=wtid)
            sta_df = self.get_df_from_db(statisticdata_sql)
            # 抽取需要的列
            extract_df = sta_df.iloc[:, 0:cloums_size]
            # 转换列名
            extract_df.columns = cloums
            return (extract_df, wtid)

    def write_data(self, df, wtid):
        # 写出为文件 /data/wff_algo/data_file/
        #/home/lijiapeng/projects/mysql_cluster_load/data/
        df.to_csv(path_or_buf='E:\\{wtid}.csv'.format(wtid=wtid), sep=str(','), index=False, na_rep='',
                  header=True, encoding="utf_8_sig")

    def load_from_mysql(self):
        # 获取所有wtid以及对应的协议id
        wtinfo_sql = 'select wtid,protocolid from config.wtinfo'
        wtinfo_df = self.get_df_from_db(wtinfo_sql)
        pool = TPool(6)
        for wtid, protocolid in wtinfo_df.values:
            try:
                res = pool.apply_async(self.extract_data,(wtid, protocolid))
                kw_tup = res.get()
                if kw_tup:
                    pool.apply_async(self.write_data, (kw_tup[0], kw_tup[1]))
                    #print(f"执行完毕{wtid}".format(wtid=wtid))
            except Exception as e:
                print("写入文件时发生异常，异常wtid为{wtid}".format(wtid=wtid))
                traceback.print_exc()
        pool.close()
        pool.join()


if __name__ == '__main__':
    load = LoadFromMysql()
    load.load_from_mysql()
    #pass