ASM中磁盘组权限设置

aix平台11gr2单库使用使用grid和oracle用户分别部署gi和db,在添加磁盘的时候，使用设置磁盘所属用户和组为grid与oinstall,设置权限为755.添加磁盘成功后，数据库直接crash.
asm添加磁盘操作

SQL>  alter diskgroup DATA add disk '/dev/rhdisk15' 
NOTE: Assigning number (2,7) to disk (/dev/rhdisk15)
NOTE: requesting all-instance membership refresh for group=2
NOTE: initializing header on grp 2 disk DATA_0007
NOTE: requesting all-instance disk validation for group=2
Wed Apr 03 22:09:03 2013
NOTE: skipping rediscovery for group 2/0xa026f7ec (DATA) on local instance.
NOTE: requesting all-instance disk validation for group=2
NOTE: skipping rediscovery for group 2/0xa026f7ec (DATA) on local instance.
NOTE: initiating PST update: grp = 2
Wed Apr 03 22:09:03 2013
GMON updating group 2 at 21 for pid 17, osid 22610284
NOTE: PST update grp = 2 completed successfully 
NOTE: membership refresh pending for group 2/0xa026f7ec (DATA)
GMON querying group 2 at 22 for pid 13, osid 20643916
NOTE: cache opening disk 7 of grp 2: DWDATAGRP_0007 path:/dev/rhdisk15
GMON querying group 2 at 23 for pid 13, osid 20643916
SUCCESS: refreshed membership for 2/0xa026f7ec (DATA)
NOTE: starting rebalance of group 2/0xa026f7ec (DATA) at power 1
SUCCESS:  alter diskgroup DATA add disk '/dev/rhdisk15'
Starting background process ARB0
Wed Apr 03 22:09:07 2013
ARB0 started with pid=22, OS id=14155890 
NOTE: assigning ARB0 to group 2/0xa026f7ec (DATA) with 1 parallel I/O
NOTE: Attempting voting file refresh on diskgroup DATA
Wed Apr 03 22:09:19 2013
SQL>  alter diskgroup DATA add disk '/dev/rhdisk11' 
Wed Apr 03 22:09:20 2013
NOTE: stopping process ARB0
NOTE: rebalance interrupted for group 2/0xa026f7ec (DATA)
NOTE: Assigning number (2,8) to disk (/dev/rhdisk11)
NOTE: requesting all-instance membership refresh for group=2
NOTE: initializing header on grp 2 disk DATA_0008
NOTE: requesting all-instance disk validation for group=2
NOTE: skipping rediscovery for group 2/0xa026f7ec (DATA) on local instance.
NOTE: requesting all-instance disk validation for group=2
NOTE: skipping rediscovery for group 2/0xa026f7ec (DATA) on local instance.
NOTE: initiating PST update: grp = 2
Wed Apr 03 22:09:23 2013
GMON updating group 2 at 24 for pid 17, osid 22610284
NOTE: PST update grp = 2 completed successfully 
NOTE: membership refresh pending for group 2/0xa026f7ec (DATA)
GMON querying group 2 at 25 for pid 13, osid 20643916
NOTE: cache opening disk 8 of grp 2: DATA_0008 path:/dev/rhdisk11
GMON querying group 2 at 26 for pid 13, osid 20643916
SUCCESS: refreshed membership for 2/0xa026f7ec (DATA)
NOTE: starting rebalance of group 2/0xa026f7ec (DATA) at power 1
SUCCESS:  alter diskgroup DATA add disk '/dev/rhdisk11'
Starting background process ARB0
Wed Apr 03 22:09:26 2013
ARB0 started with pid=22, OS id=22872116 
NOTE: assigning ARB0 to group 2/0xa026f7ec (DATA) with 1 parallel I/O
NOTE: Attempting voting file refresh on diskgroup DATA
Wed Apr 03 22:14:41 2013
NOTE: ASM client xifenfei:xifenfei disconnected unexpectedly.
NOTE: check client alert log.
NOTE: Trace records dumped in trace file /u01/diag/asm/+asm/+ASM/trace/+ASM_ora_15073468.trc
Wed Apr 03 22:16:53 2013
NOTE: client xifenfei:xifenfei registered, osid 20709378, mbr 0x0
Wed Apr 03 22:20:33 2013
NOTE: client xifenfei:xifenfei deregistered

这里可看到增加磁盘操作正常并且开始做rebalance，但是也看到关于client xifenfei异常断开连接(本质就是数据库crash)

crash时的alert日志

Wed Apr 03 22:00:00 2013
Setting Resource Manager plan SCHEDULER[0x318B]:DEFAULT_MAINTENANCE_PLAN via scheduler window
Setting Resource Manager plan DEFAULT_MAINTENANCE_PLAN via parameter
Wed Apr 03 22:00:00 2013
Starting background process VKRM
Wed Apr 03 22:00:00 2013
VKRM started with pid=31, OS id=22413426 
Wed Apr 03 22:09:06 2013
ORA-15025: could not open disk "/dev/rhdisk15"
ORA-27041: unable to open file
IBM AIX RISC System/6000 Error: 13: Permission denied
Additional information: 11
Wed Apr 03 22:09:06 2013
SUCCESS: disk DATA_0007 (7.2092304189) added to diskgroup DATA
Wed Apr 03 22:09:26 2013
ORA-15025: could not open disk "/dev/rhdisk15"
ORA-27041: unable to open file
IBM AIX RISC System/6000 Error: 13: Permission denied
Additional information: 11
Wed Apr 03 22:09:26 2013
SUCCESS: disk DATA_0008 (8.2092304190) added to diskgroup DATA
Wed Apr 03 22:14:40 2013
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_dbw0_17367438.trc:
ORA-15080: synchronous I/O operation to a disk failed
WARNING: failed to write mirror side 1 of virtual extent 1 logical extent 0 of file 261 in 
group 2 on disk 7 allocation unit 464 
KCF: read, write or open error, block=0x6a online=1
        file=1 '+DATA/xifenfei/datafile/system.261.788373447'
        error=15081 txt: ''
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_dbw0_17367438.trc:
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_dbw0_17367438.trc:
ORA-63999: data file suffered media failure
ORA-01114: IO error writing block to file 1 (block # 106)
ORA-01110: data file 1: '+DATA/xifenfei/datafile/system.261.788373447'
ORA-15081: failed to submit an I/O operation to a disk
ORA-15081: failed to submit an I/O operation to a disk
DBW0 (ospid: 17367438): terminating the instance due to error 63999

这里可以看到数据库异常crash是因为/dev/rhdisk15没有权限去操作该文件，导致dbw0进程异常，从而出现该数据库crash

尝试重启数据库(asm重启正常)

SQL> startup
ORACLE instance started.

Total System Global Area 1.2827E+10 bytes
Fixed Size                  2233480 bytes
Variable Size            1711278968 bytes
Database Buffers         1.1073E+10 bytes
Redo Buffers               40894464 bytes
Database mounted.
ORA-01113: file 1 needs media recovery
ORA-01110: data file 1: '+DATA/xifenfei/datafile/system.261.788373447'

这里提示file 1需要恢复,查看alert日志,出现以下错误

Completed: ALTER DATABASE   MOUNT
Wed Apr 03 22:17:02 2013
ALTER DATABASE OPEN
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_ora_11534798.trc:
ORA-27041: unable to open file
IBM AIX RISC System/6000 Error: 13: Permission denied
Additional information: 3
Additional information: 4
Additional information: 4194306
WARNING: Write Failed. group:2 disk:8 AU:462 offset:16384 size:16384
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_ora_11534798.trc:
ORA-15080: synchronous I/O operation to a disk failed
WARNING: failed to write mirror side 1 of virtual extent 0 logical extent 0 of file 261 in 
group 2 on disk 8 allocation unit 462 
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_ora_11534798.trc:
ORA-27041: unable to open file
IBM AIX RISC System/6000 Error: 13: Permission denied
Additional information: 3
Additional information: 4
Additional information: 4194306
WARNING: Write Failed. group:2 disk:8 AU:690 offset:16384 size:16384
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_ora_11534798.trc:
ORA-27041: unable to open file
IBM AIX RISC System/6000 Error: 13: Permission denied
Additional information: 3
Additional information: 4
Additional information: 4194306
WARNING: Write Failed. group:2 disk:8 AU:918 offset:16384 size:16384
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_ora_11534798.trc:
ORA-15080: synchronous I/O operation to a disk failed
WARNING: failed to write mirror side 1 of virtual extent 0 logical extent 0 of file 263 in 
group 2 on disk 8 allocation unit 918 
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_ora_11534798.trc:
ORA-15080: synchronous I/O operation to a disk failed
WARNING: failed to write mirror side 1 of virtual extent 0 logical extent 0 of file 262 in 
group 2 on disk 8 allocation unit 690 
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_ora_11534798.trc:
ORA-01110: data file 3: '+DATA/xifenfei/datafile/undotbs1.263.788373475'
ORA-01114: IO error writing block to file 3 (block # 1)
ORA-15081: failed to submit an I/O operation to a disk
ORA-15081: failed to submit an I/O operation to a disk
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_ora_11534798.trc:
ORA-01110: data file 2: '+DATA/xifenfei/datafile/sysaux.262.788373463'
ORA-01114: IO error writing block to file 2 (block # 1)
ORA-15081: failed to submit an I/O operation to a disk
ORA-15081: failed to submit an I/O operation to a disk

recover database 操作

SQL> recover database;
ORA-00283: recovery session canceled due to errors
ORA-01201: file 1 header failed to write correctly

Wed Apr 03 22:18:49 2013
ALTER DATABASE RECOVER  database  
Media Recovery Start
 started logmerger process
Wed Apr 03 22:18:50 2013
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_pr00_12714126.trc:
ORA-27041: unable to open file
IBM AIX RISC System/6000 Error: 13: Permission denied
Additional information: 3
Additional information: 4
Additional information: 4194306
WARNING: Write Failed. group:2 disk:8 AU:462 offset:16384 size:16384
Errors in file /oracle/diag/rdbms/xifenfei/xifenfei/trace/xifenfei_pr00_12714126.trc:
ORA-27041: unable to open file
IBM AIX RISC System/6000 Error: 13: Permission denied
Additional information: 3
Additional information: 4
Additional information: 4194306

依然是这里的提示依然是因为磁盘无读写权限从而出现数据库无法写数据文件问题,修改刚刚加入的磁盘文件权限问为660(4读2写1执行),表明与oinstall相同组的oracle用户对该磁盘也有读写权限.
这个事故是一个很简单,而且随着11g中asm使用grid和oracle用户的客户越来越多，相关的事故也越来越多，因为大多数使用人习惯直接给某个文件授权为755,而在这样的grid和oracle分开安装的系统中,将出现增加磁盘后,数据库crash,而且不能起来(因为oracle用户对磁盘只有读权限,无写权限),一种比较好的规范:在11gr2的asm系统中(grid和oracle用户),建议设置磁盘为grid.oinstall,权限设置为660