MySQL备份工具mydumper备份

#!/usr/bin/phthon


import os
import time
import commands
import shutil
import threading
from os.path import join, getsize
import MySQLdb as mydb


# 备份目录
baseDir = "/data2/backup/backup_data/"


# ns 或 wx;  备份后是否要压缩(mydumper 自带压缩功能)，要压缩 True，否则 False.
idc = 'ns'; isZip = True


# 备份失败是否重试 ，True 则重试,不重试 设 False, retry_sleep 多久后才开始重试(秒)
is_errRetryBackup = True; retry_sleep = 300


# 备份日期
backup_date = time.strftime("%Y%m%d")


# 备份命令
cmd  = "/usr/local/bin/mydumper -h %s -u root -p password? -P %s %s -t 5 -o %s"


'''
功能描述：
  1. mydumper 远程批量备份, 备份列表由配置文件提供
  2. 可按要求对备份是否压缩(mydumper 自动压缩)
  3. 备份失败允许再尝试备份一次
  4. 备份信息写入数据库
'''
def main():
    thread_pool = []


    # 是否启用压缩
    zip = '-c' if isZip == True else ''


    # 从配置文件读取 ip, 名称, 端口, 并拼凑备份语句
    #f = open('/data2/backup/cnf/other_list.cnf', 'r')
    f = open('/data/other_list.cnf', 'r')
    for lines in f.readlines():
        if (not lines.startswith('#') and len(lines.strip()) > 0):
            str = lines.split()
            host, businessName, port, isMaster = str[0], str[1], str[2], str[3]
            # 业务文件夹不存在则创建
            dir = baseDir + '/' + businessName;
            if (not os.path.exists(dir)):
                 os.makedirs(dir)


            dir += "/%s%s" % (businessName, backup_date)
            # 业务目录: dir , 备份目录: dir/name+备份日期
            strcmd  = cmd % (host, port, zip, dir)


            th = threading.Thread(target = mydumper, args =(strcmd, dir, businessName, host, port, is_errRetryBackup, int(isMaster)))
            thread_pool.append(th)


    if (thread_pool):
        for t in thread_pool:
            t.daemon = True
            t.start()
        for t in thread_pool:
            t.join()


def mydumper(sCmd, backupDir, businessName, host, port, is_Retry, isMaster):
    master_host = ""; backup_host = host; name = businessName; port = port; backup_type = 1; file = "";
    start_time = ""; stop_time = ""; returncode = 0; file_size = 0; slave_statement = ""; std_err = "";


    start_time = time.strftime("%Y%m%d%H%M%S")


    # 清除可能遗留的备份
    if (os.path.exists(backupDir)):
        shutil.rmtree(backupDir)


    # 执行备份
    returncode, std_err = execute(sCmd)


    stop_time = time.strftime("%Y%m%d%H%M%S")


    if (returncode == 0):
        # 备份 std_err 返回不为空也视为出错。
        if (std_err.strip() != ""):
            returncode = 123456
        else:
            # 获取 change master to 信息，再次校验备份是否有效
            returncode, std_err, master_host, slave_statement = statement(backupDir, backup_host, isMaster)


            if (returncode == 0):
                file = backupDir


    if (returncode != 0):
        # 异常备份标记为: 日期 + _ERR
        errDir = backupDir + "_ERR"
        os.rename(backupDir, errDir)
        file = errDir


    # 获取备份大小
    file_size = getDirsize(file)


    if (len(std_err) > 255):
        std_err = std_err[:250] + "..."


    my_args  = [idc, master_host, backup_host, name, port, backup_type, file, start_time, stop_time, returncode, file_size, slave_statement, std_err]


    # 写入数据库
    call_proc(my_args)


    # 备份失败是否需要重备? 重备允许一次.
    if (is_Retry == True and returncode != 0):
        time.sleep(retry_sleep)


        oldfile = sCmd.split('-o')[1]
        pos = oldfile.rfind("/") + 1


        # 获取备份全路径, 备份文件标记为重备 字样
        retry_file = oldfile[:pos] + "ReBackup-" + oldfile[pos:]


        retryCmd = sCmd.replace(oldfile,  retry_file)


        # 重备开始
        mydumper(retryCmd, retry_file.strip(), name, host, port, False, isMaster)


def getDirsize(path):
    # 获取备份文件夹大小
    size = 0L
    for root, dirs, files in os.walk(path):
        size += sum([getsize(join(root, name)) for name in files])
    return (size)


def statement(path, backup_host, isMaster):
    '''
    功能: 从 metadata 读取change master to 信息
    1. 备份过程: 会先生成: metadata.partial, 完成后metadata.partial会重名为: metadata 并写入备份完成时间
    2. metadata 分3段:
       (1) Started dump: 备份开始时间.
       (2) master 的log-file 和 log-pos 信息 (必有); slave 的host、log-file 和 log-pos 信息 （备机是slave 才有)
       (3) Finished dump: 备份结束时间
    3. 返回错码, master_host 和 change master to 信息
    '''
    path += "/metadata"; sMetadata = ""; master_host = ""; er_code = 654321; er_info = "%s not exists !!!" % (path)


    if (os.path.exists(path)):
        if (isMaster != 1):
            # 备机是 slave
            num = 3
            sFinds = "SLAVE STATUS"
        else:
            num = 2
            sFinds = "MASTER STATUS"


        f = open(path, 'r')
        rows = f.readlines(); i = 100; lst =[]
        for s in rows:
            if (s.find(sFinds) > 0):
                i = 1; continue


            if (i <= num):
                lst.append(s.split(':')[1].strip())
                i += 1


        if (isMaster == 1):
            # 备机是 master
            master_host = backup_host
            log_file, log_pos = lst;
        else:
            # 备机是 slave
            master_host, log_file, log_pos = lst;


        er_code = 0
        er_info = ""
        sMetadata = "CHANGE MASTER TO MASTER_HOST='%s',MASTER_LOG_FILE='%s',MASTER_LOG_POS=%s,MASTER_USER='rep_user',MASTER_PASSWORD='meizu.com'" % (master_host, log_file, log_pos )


    return (er_code, er_info, master_host, sMetadata)


def execute(cmd):
    '''
    1.执行 shell 命令
    2.返回执行信息 (returncode = 0 则执行成功, std_err 为报错的错误信息)
    '''
    try:
        returncode, std_err = commands.getstatusoutput(cmd)
        return (returncode, std_err)
    except os.error, e:
        # 异常返回 1001 错误
        return (1001, e)


def call_proc(my_args):
    # 备份信息写入数据库
    try:
        conn = mydb.connect(host = '127.0.0.1', user = 'test', passwd = 'zxc/213?', db = 'meizu_item')
        cur  = conn.cursor()


        cur.callproc('sp_backup_i',[my_args[0], my_args[1], my_args[2], my_args[3], my_args[4], my_args[5], my_args[6], my_args[7], my_args[8], my_args[9], my_args[10], my_args[11], my_args[12]])
        conn.commit()
    except mydb.Error, e:
        pass
        # print "Mysql Error %d: %s" % (e.args[0], e.args[1])
    finally:
        cur.close(); conn.close()


if __name__ == '__main__':
    main()