超级复杂困难之Oracle数据库大恢复

【 tulaoshi.com - 编程语言 】

　　昨天，一个朋友公司的数据库崩溃。

　　这再次印证了我反复提到的一个命题：数据库也需要休息。

　　每逢节假日，数据库也经常会自我选择放假。

　　以前我说：年终难终 进入数据库事故多发期，一年一度今又是，记得另外一个圣诞节，我还和Biti一起在北京的时候，同样遇到一个上海的朋友数据库崩溃，我们远程指导这位朋友恢复了数据。

　　这次的事情是这样的。

　　1.首先主机宕机，磁盘出错

　　看到以下这类错误，一般你的数据都很危险了

　　Dec 24 13:52:13 kernel: sda5: rw=0, want=18298437640, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=10384710304, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=8756273744, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=5023902272, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=6730428824, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=8884660792, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=9182513808, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=5002858800, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=6730428824, limit=163846872
Dec 24 13:52:13 kernel: attempt to access beyond end of device
Dec 24 13:52:13 kernel: sda5: rw=0, want=15872410168, limit=163846872

　　2.数据文件大量损坏

　　当然这次也不例外，大量文件损坏，dbv大量如下错误：

[oracle@stat datafile]$ dbv file=o1_mf_system_29448mn7_.dbf blocksize=8192
　　DBVERIFY: Release 10.2.0.2.0 - Production on Thu Dec 25 22:17:52 2008
　　Copyright (c) 1982, 2005, Oracle. All rights reserved.
　　DBVERIFY - Verification starting : FILE = o1_mf_system_29448mn7_.dbf
Page 40 is influx - most likely media corrupt
Corrupt block relative dba: 0x00400028 (file 1, block 40)
Fractured block found during dbv:
Data in bad block:
type: 6 format: 2 rdba: 0x00400028
last change scn: 0x0000.18990f0e seq: 0x1 flg: 0x06
spare1: 0x0 spare2: 0x0 spare3: 0x0
consistency value in tail: 0xbc120601
check value in block header: 0xc0cb
computed block checksum: 0xb003
　　Page 232 is influx - most likely media corrupt
Corrupt block relative dba: 0x004000e8 (file 1, block 232)
Fractured block found during dbv:
Data in bad block:
type: 6 format: 2 rdba: 0x004000e8
last change scn: 0x0000.18991b98 seq: 0x1 flg: 0x06
spare1: 0x0 spare2: 0x0 spare3: 0x0
consistency value in tail: 0x6c440601
check value in block header: 0x8d7f
computed block checksum: 0x77dc

　　3.控制文件损坏

　　启动数据库出现如下错误：

　　Wed Dec 24 17:08:52 2008
ALTER DATABASE MOUNT
Wed Dec 24 17:08:56 2008
Errors in file /opt/oracle/admin/stat/udump/stat_ora_4630.trc:
ORA-00600: internal error code, arguments: [kccpb_sanity_check_2], [11258908], [10375171], [0x0], [], [], [], []
Wed Dec 24 17:08:57 2008
ORA-600 signalled during: ALTER DATABASE MOUNT...
Wed Dec 24 17:09:01 2008
Starting ORACLE instance (normal)
Wed Dec 24 17:16:22 2008
Corrupt block 1 found during reading backup piece, file=/opt/oracle/product/db10g/dbs/snapcf_stat.f, corr_type=2

　　4.经过反复确认，这个环境Over了

　　5.不完全的备份

　　以前的备份机制使得我可以从远程主机找到一系列备份集，但是没有控制文件。

　　通过备份集、dbms_backup_restore等手段，首先恢复出来数据文件，然后尝试启动数据库

　　6.强制打开

　　通过强制resetlogs手段打开数据库，出现ORA-600 4000错误

　　Wed Dec 24 18:56:00 2008
Errors in file /opt/oracle/admin/stat/udump/stat_ora_21479.trc:
ORA-00600: internal error code, arguments: [4000], [15], [], [], [], [], [], []
Wed Dec 24 18:56:01 2008
Errors in file /opt/oracle/admin/stat/udump/stat_ora_21479.trc:
ORA-00704: bootstrap process failure
ORA-00704: bootstrap process failure
ORA-00600: internal error code, arguments: [4000], [15], [], [], [], [], [], []

　　7.通过BBED解决ORA-600 4000错误

　　这个没说的，只能通过BBED搞定了，修复有问题的数据块，再次尝试打开数据库

　　8.遇到ORA-600 2662错误

　　这个错误就好解决了，通过我网站上的示例就可以解决：

　　Wed Dec 24 21:13:17 2008
Errors in file /opt/oracle/admin/stat/udump/stat_ora_28316.trc:
ORA-00600: internal error code, arguments: [2662], [0], [412717646], [0], [412772634], [8389633], [], []
ORA-01092: ORACLE instance terminated. Disconnection forced
ORA-00600: internal error code, arguments: [2662], [0], [412717644], [0], [412772634], [8389633], [], []
Wed Dec 24 21:13:18 2008
Errors in file /opt/oracle/admin/stat/udump/stat_ora_28316.trc:
ORA-00600: internal error code, arguments: [2662], [0], [412717647], [0], [412772634], [8389633], [], []
ORA-00600: internal error code, arguments: [2662], [0], [412717646], [0], [412772634], [8389633], [], []
ORA-01092: ORACLE instance terminated. Disconnection forced
ORA-00600: internal error code, arguments: [2662], [0], [412717644], [0], [412772634], [8389633], [], []
Wed Dec 24 21:13:18 2008
Errors in file /opt/oracle/admin/stat/udump/stat_ora_28316.trc:
ORA-00600: internal error code, arguments: [2662], [0], [412717647], [0], [412772634], [8389633], [], []
ORA-00600: internal error code, arguments: [2662], [0], [412717646], [0], [412772634], [8389633], [], []
ORA-01092: ORACLE instance terminated. Disconnection forced
ORA-00600: internal error code, arguments: [2662], [0], [412717644], [0], [412772634], [8389633], [], []

　　9.解决ORA-600 4097号错误

　　接下来继续出现ORA-600 4097号错误，这个也好解决，搞定UNDO表空间就Ok了

　　Wed Dec 24 21:18:12 2008
Errors in file /opt/oracle/admin/stat/bdump/stat_j000_28723.trc:
ORA-00600: internal error code, arguments: [4097], [], [], [], [], [], [], []

　　10.解决一些其他小问题

　　此处省略10000字，终于搞定了用户数据库！

　　这个案例所用到的所有知识在我的网站上全都有详细介绍，不过要能把所有知识综合运用才能解决这次的故障。