3層主-從-從從

從裡面15張表被有計劃的刪除了幾千行資料，共15張表致中，同事意外乾的，導致主從中斷。主庫更新頻繁，要求其中被刪掉資料在所有資料庫中都刪掉，主庫中所有的更新全部都同步的到後面的從庫中去，因是資料採集類的該15張表要求一致性不嚴格。

所以偷懶解決

1.跳過主從複製中的錯誤

2.停掉爬蟲更新程式

3.從dump15張表，寫入主

1. 跳過主從複製中的錯誤的指令碼

#!/usr/bin/python
# -*- coding: UTF-8 -*-
import pymysql
import time

ip53="10.18.141.53"

# # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
mydb53 = pymysql.connect(ip53,"dba","dba.","crdw" )
mycursor53 = mydb53.cursor()
mycursor53.execute('SET CHARACTER SET utf8;')
mycursor53.execute('SET NAMES utf8;')
mycursor53.execute('SET character_set_connection=utf8;')

Slave_SQL_Running53 = ''
for i in range(1,180000):
    time.sleep(1)
    mycursor53.execute('show slave status')
    results = mycursor53.fetchall()
    for row in results:
        Slave_IO_Running53 = row[10]
        Slave_SQL_Running53 = row[11]
        Last_IO_Errno53 = row[34]
        Last_IO_Error53 = row[35]
        Last_SQL_Errno53 = row[36]
        Last_SQL_Error53 = row[37]

    if(Slave_SQL_Running53 == 'No' and Slave_IO_Running53 == 'Yes' ) :
        Esql = '''
                STOP SLAVE;
                SET GLOBAL SQL_SLAVE_SKIP_COUNTER = 1;
                START SLAVE;'''

        mycursor53.execute(Esql)

        print 'times:' ,i
        print Last_IO_Errno53
        print Last_IO_Error53
        print Last_SQL_Errno53
        print Last_SQL_Error53
    time.sleep(0.1)

mydb53.close()

print 'please check again ....'