Oracle ASM错误之--ORA-15033

系统环境：

操作系统： AIX 5.3.8

Oracle：   Oracle 10gR2

故障现象：

SQL> create diskgroup datafile
  2      normal redundancy
    failgroup fg1 disk '/dev/rlv_data1'
  3    4    failgroup fg2 disk '/dev/rlv_data2';
create diskgroup datafile
*
ERROR at line 1:
ORA-15018: diskgroup cannot be created
ORA-15033: disk '/dev/rlv_data2' belongs to diskgroup "RECOVER"
ORA-15033: disk '/dev/rlv_data1' belongs to diskgroup "DATA"
SQL> select group_number,name,state from v$asm_diskgroup;
GROUP_NUMBER NAME                           STATE
------------ ------------------------------ -----------
           0 REDO                           DISMOUNTED
           0 ARCH                           DISMOUNTED
           0 DATA                           DISMOUNTED
           0 RECOVER                        DISMOUNTED

故障原因：  

[oracle@aix213 dbs]$lsvg -l asmvg
asmvg:
LV NAME             TYPE       LPs     PPs     PVs  LV STATE      MOUNT POINT
lv_data1            jfs        64      64      1    open/syncd    N/A
lv_data2            jfs        64      64      1    open/syncd    N/A
asmredo2            jfs        32      32      1    open/syncd    N/A
[oracle@aix213 dbs]$lspv 
hdisk1          00c6d0ef0395eef2                    rootvg          active
hdisk0          0006816a72287441                    asmvg           active

应该是在AIX的ODM中，rlv_data1和rlv_data2，已经属于另外的磁盘组。

解决方法：

1）清理raw头部信息

[root@aix213 /]#dd if=/dev/zero of=/dev/rlv_data1 bs=8k count=256
256+0 records in
256+0 records out
[root@aix213 /]#dd if=/dev/zero of=/dev/rlv_data2 bs=8k count=256
256+0 records in
256+0 records out

2）重新建立ASM DISKGROUP

[oracle@aix213 dbs]$sqlplus / as sysdba
SQL*Plus: Release 10.2.0.1.0 - Production on Thu Jul 24 11:44:29 2014
Copyright (c) 1982, 2005, Oracle.  All rights reserved.
Connected to:
Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - 64bit Production
With the Partitioning, OLAP and Data Mining options
SQL> select name ,state from v$asm_diskgroup;
NAME                           STATE
------------------------------ -----------
REDO                           DISMOUNTED
ARCH                           DISMOUNTED
清理raw头部信息后data和recover diskgroup自动被删除 ！
SQL> create diskgroup datafile
  2      normal redundancy
  3      failgroup fg1 disk '/dev/rlv_data1'
  4    failgroup fg2 disk '/dev/rlv_data2';
Diskgroup created.
磁盘组创建成功 ！

3）验证

SQL> select name ,state from v$asm_diskgroup;
NAME                           STATE
------------------------------ -----------
REDO                           DISMOUNTED
ARCH                           DISMOUNTED
DATAFILE                       MOUNTED
SQL> alter diskgroup redo mount;
Diskgroup altered.
SQL> alter diskgroup arch mount;
Diskgroup altered.
SQL> select name ,state from v$asm_diskgroup;
NAME                           STATE
------------------------------ -----------
REDO                           MOUNTED
ARCH                           MOUNTED
DATAFILE                       MOUNTED

@至此，问题被解决 ！