存在以下这种情况， 由于误DROP DISKGROUP或者误将DISK DROP 出原Diskgroup，或者因为Bug 13331814: ASM DISKS TURNED INTO FORMER WHILE DISKGROUP IS MOUNTED， 导致ASM DISK的header_status=FORMER，而非正常的MEMBER状态。
诗檀软件专业数据库修复团队
服务热线 ： 13764045638 QQ号:47079569 邮箱：service@parnassusdata.com
举例来说如下面的例子
[oracle@mlab2 ~]$ sqlplus / as sysasm
SQL*Plus: Release 11.2.0.3.0 Production on Tue Nov 19 21:55:09 2013
Copyright (c) 1982, 2011, Oracle. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - 64bit Production
With the Automatic Storage Management option
SQL> create diskgroup maclean external redundancy disk '/dev/asm-disk9';
Diskgroup created.
SQL> select group_number,name,state from v$asm_diskgroup;
GROUP_NUMBER NAME STATE
------------ ------------------------------ -----------
1 DATA MOUNTED
2 MACLEAN MOUNTED
SQL> col path for a40
SQL> select name,path,header_status from v$asm_disk where group_number=2;
NAME PATH
------------------------------ ----------------------------------------
HEADER_STATU
------------
MACLEAN_0000 /dev/asm-disk9
MEMBER
这里我们将diskgroup drop一次：
SQL> drop diskgroup maclean;
Diskgroup dropped.
SQL> alter diskgroup maclean mount;
alter diskgroup maclean mount
*
ERROR at line 1:
ORA-15032: not all alterations performed
ORA-15017: diskgroup "MACLEAN" cannot be mounted
ORA-15063: ASM discovered an insufficient number of disks for diskgroup
"MACLEAN"
SQL> select name,path,header_status from v$asm_disk where path='/dev/asm-disk9';
NAME PATH
------------------------------ ----------------------------------------
HEADER_STATU
------------
/dev/asm-disk9
FORMER
使用kfed查看 ASM disk header metadata
[oracle@mlab2 ~]$ kfed read /dev/asm-disk9 |head -25
kfbh.endian: 1 ; 0x000: 0x01
kfbh.hard: 130 ; 0x001: 0x82
kfbh.type: 1 ; 0x002: KFBTYP_DISKHEAD
kfbh.datfmt: 1 ; 0x003: 0x01
kfbh.block.blk: 0 ; 0x004: blk=0
kfbh.block.obj: 2147483648 ; 0x008: disk=0
kfbh.check: 554377417 ; 0x00c: 0x210b20c9
kfbh.fcn.base: 0 ; 0x010: 0x00000000
kfbh.fcn.wrap: 0 ; 0x014: 0x00000000
kfbh.spare1: 0 ; 0x018: 0x00000000
kfbh.spare2: 0 ; 0x01c: 0x00000000
kfdhdb.driver.provstr: ORCLDISK ; 0x000: length=11
kfdhdb.driver.reserved[0]: 65796 ; 0x008: 0x00010104
kfdhdb.driver.reserved[1]: 1 ; 0x00c: 0x00000001
kfdhdb.driver.reserved[2]: 4206569 ; 0x010: 0x00402fe9
kfdhdb.driver.reserved[3]: 3367865 ; 0x014: 0x003363b9
kfdhdb.driver.reserved[4]: 196018176 ; 0x018: 0x0baf0000
kfdhdb.driver.reserved[5]: 390595073 ; 0x01c: 0x17480201
kfdhdb.compat: 168820736 ; 0x020: 0x0a100000
kfdhdb.dsknum: 0 ; 0x024: 0x0000
kfdhdb.grptyp: 1 ; 0x026: KFDGTP_EXTERNAL
kfdhdb.hdrsts: 4 ; 0x027: KFDHDR_FORMER
kfdhdb.dskname: MACLEAN_0000 ; 0x028: length=12
kfdhdb.grpname: MACLEAN ; 0x048: length=7
kfdhdb.fgname: MACLEAN_0000 ; 0x068: length=12
这里的kfdhdb.hdrsts: 4 ; 0x027: KFDHDR_FORMER 说明了该DISK的状态为FORMER
首先备份对应 ASM DISK的header
[oracle@mlab2 ~]$ mkdir /tmp/asm
[oracle@mlab2 ~]$ dd if=/dev/asm-disk9 of=/tmp/asm/asm-disk9-header bs=1024k count=20
[oracle@mlab2 ~]$ kfed read /dev/asm-disk9 > /tmp/asm/asm-disk9-meta
[oracle@mlab2 ~]$ ls -l /tmp/asm/asm-disk9-meta
-rw-r--r-- 1 oracle oinstall 6597 Nov 20 01:26 /tmp/asm/asm-disk9-meta
修改asm-disk9-meta这个文本中的内容
kfdhdb.hdrsts: 4 ; 0x027: KFDHDR_FORMER
修改为
kfdhdb.hdrsts: 3 ; 0x027: KFDHDR_MEMBER
如下面2图所示的变更：
到
之后可以使用修改好的元数据信息文本来patch ASM DISK了，具体命令如下：
[oracle@mlab2 ~]$ kfed merge /dev/asm-disk9 text=/tmp/asm/asm-disk9-meta
再次确认
[oracle@mlab2 ~]$ kfed read /dev/asm-disk9 |grep hdrsts
kfdhdb.hdrsts: 3 ; 0x027: KFDHDR_MEMBER
最后尝试MOUNT该DISK对应的DISKGROUP:
SQL> alter diskgroup maclean mount;
Diskgroup altered.
SQL> select name,state from v$asm_diskgroup;
NAME STATE
------------------------------ -----------
DATA MOUNTED
MACLEAN MOUNTED