.word范文灾难恢复操作方案2一、模拟环境中断主站点与备站点线路,模拟主存储完全损坏二、灾备站点启用服务移除slave存储上reploftest卷与master站点间的映射关系.word范文修改reploftest卷名称(防止主备存储通讯正常后,删除远程复制关系时,系统自动对repleoftest的操作).word范文将修改后的slave_test卷映射给slave数据库服务器。.word范文重启slave数据库服务器,识别新分配的存储卷,并加载文件系统,启动数据库。slave-site-db:~#fdisk-lDisk/dev/mapper/36000d310038186000000000000000029:10GiB,10737418240bytes,20971520sectorsUnits:sectorsof1*512=512bytesSectorsize(logical/physical):512bytes/4096bytesI/Osize(minimum/optimal):2097152bytes/2097152bytesDisklabeltype:dosDiskidentifier:0xcbb2660d.word范文DeviceBootStartEndSectorsSizeIdType/dev/mapper/36000d310038186000000000000000029-part14096209715192096742410G83Linuxslave-site-db:~#mount/dev/mapper/36000d310038186000000000000000029-part1/tttslave-site-db:~#cd/tttslave-site-db:/ttt#lsdatadata.tarlost+foundslave-site-db:/ttt#cpdata.tar1.tar三、计划回切至主站点3.1、停数据库1)、停数据库2)、umount文件系统cd/umount/ttt.word范文3.2、存储回切操作1)、取消映射2)、手工对slave_test创建快照.word范文3)、恢复通讯链路(模拟主存储故障完全解决)确认两台存储是否完全恢复连接4)、删除原先远程复制关系确认test到reploftest关系是否存在,是否存在故障.word范文删除test卷的远程复制关系5)、删除master存储上原来的test卷的映射和卷本身.word范文.word范文6)、建立slave存储到master存储的远程复制关系.word范文.word范文.word范文等待两个存储之间完成数据复制,复制时间受数据量和带宽速度影响。7)、删除slave_tests与replofslave_test的映射关系,但保留master存储上的replofslave_test卷.word范文8)、修改master存储上replofslave_test卷名称,并映射个master数据库集群服务器.word范文9)、修改master存储test卷快照策略.word范文10)、重启2台master数据库服务器识别存储.word范文A、master-db2:~#RebootB、master-db2:~#fdisk-lDisk/dev/mapper/36000d31003818400000000000000002f:10GiB,10737418240bytes,20971520sectorsUnits:sectorsof1*512=512bytesSectorsize(logical/physical):512bytes/4096bytesI/Osize(minimum/optimal):2097152bytes/2097152bytesDisklabeltype:dosDiskidentifier:0xcbb2660dDeviceBootStartEndSectorsSizeIdType/dev/mapper/36000d31003818400000000000000002f-part14096209715192096742410G83LinuxC、修改/etc/multipath.conf文件,将新的卷的UUID36000d31003818400000000000000002f填写到mpath3下(生产数据8T卷为mpath1).word范文vi/etc/multipath.conf,修改mpath3的wwid为36000d31003818400000000000000002dmultipaths{multipath{aliasmpath3wwid36000d31003818400000000000000002d}multipath{aliasmpath2wwid36000d310038184000000000000000015}multipath{aliasmpath1wwid36000d31003818400000000000000001f}}devices{device{failbackimmediateno_path_retryqueuepath_grouping_policymultibusproductCompellentVol.word范文vendorCOMPELNT}}2)、重新启动两台服务器master-db2:~#rebootmaster-db2:~#mount/dev/mapper/mpath3-part1/tttmaster-db2:~#cd/tttmaster-db2:/ttt#ls1.tardatadata.tarlost+found四、重建master存储到slave存储的远程复制关系1)、删除slave存储上slave_test卷.word范文2)、建立master存储test卷到slave存储的远程复制关系.word范文.word范文.word范文