认真就输

          在客户现场升级数据库，随便给客户解决一个srvctl注册的实例，切通过srvctl命令不能对实例进行管理

         环境：OS:AIX 6.1  DB:10.2.0.4.12  CRS:10.2.0.4.0

故障现象：

查看资源的情况
[oracle@uam_db1]#crs_stat -t
Name Type Target State Host
------------------------------------------------------------
ora....B1.lsnr application ONLINE ONLINE uam_db1
ora....db1.ons application ONLINE ONLINE uam_db1
ora....db1.vip application ONLINE ONLINE uam_db1
ora....B2.lsnr application ONLINE ONLINE uam_db2
ora....db2.ons application ONLINE ONLINE uam_db2
ora....db2.vip application ONLINE ONLINE uam_db2
ora.uamdb.db application ONLINE ONLINE uam_db1
ora....b1.inst application ONLINE ONLINE uam_db1
ora....b2.inst application OFFLINE OFFLINE
这里发现节点2的实例是offline
查看数据库的状态：
SQL> select instance_name,status from gv$instance;

INSTANCE_NAME STATUS
---------------- ------------
uamdb1 OPEN
uamdb2 OPEN  这里看到数据库的状态是正常的

故障处理结果：

     通过处理后，crs能正常管理节点的信息。crs_stat -t出来的结果也是正常的。

故障分析过程：  

        在crsd.log中可以发现下面的信息：

2013-06-26 03:50:08.097: [ CRSRES][11855]32Attempting to start `ora.uamdb.uamdb2.inst` on member `uam_db2`
2013-06-26 03:52:16.204: [ CRSAPP][11855]32StartResource error for ora.uamdb.uamdb2.inst error code = 1

   在imonuamdb.log中可以发现下面的信息：文件所在的路径是$ORACLE_HOME/log/uam_db1/racg/下面，srvctl来对数据库的关闭和开启都的信息都会记录在imon{db_name}.log这个文件里面。另外srvctl能管理的一些资源的日志也是在这个目录下面

2013-06-26 02:05:34   : Could not make directory /oracle/orabase/product/10g/db_home/racg/init/uam_db2/uamdb/ reason: Permission denied
Oracle Database 10g RACG Release 10.2.0.4.0 Production Copyright 1996, 2005, Oracle.  All rights reserved.
2013-06-26 03:50:08   : Could not make directory /oracle/orabase/product/10g/db_home/racg/init/uam_db2/uamdb/ reason: Permission denied
Oracle Database 10g RACG Release 10.2.0.4.0 Production Copyright 1996, 2005, Oracle.  All rights reserved.
2013-06-26 03:50:11   : Could not make directory /oracle/orabase/product/10g/db_home/racg/init/uam_db2/uamdb/ reason: Permission denied
Oracle Database 10g RACG Release 10.2.0.4.0 Production Copyright 1996, 2005, Oracle.  All rights reserved.
2013-06-26 03:50:23   : Could not make directory /oracle/orabase/product/10g/db_home/racg/init/uam_db2/uamdb/ reason: Permission denied
Oracle Database 10g RACG Release 10.2.0.4.0 Production Copyright 1996, 2005, Oracle.  All rights reserved.
2013-06-26 03:51:11   : Could not make directory /oracle/orabase/product/10g/db_home/racg/init/uam_db2/uamdb/ reason: Permission denied
Oracle Database 10g RACG Release 10.2.0.4.0 Production Copyright 1996, 2005, Oracle.  All rights reserved.
2013-06-26 03:52:16   : Could not make directory /oracle/orabase/product/10g/db_home/racg/init/uam_db2/uamdb/ reason: Permission denied
Oracle Database 10g RACG Release 10.2.0.4.0 Production Copyright 1996, 2005, Oracle.  All rights reserved.
2013-06-26 03:52:19   : Could not make directory /oracle/orabase/product/10g/db_home/racg/init/uam_db2/uamdb/ reason: Permission denied
Oracle Database 10g RACG Release 10.2.0.4.0 Production Copyright 1996, 2005, Oracle.  All rights reserved.
2013-06-26 03:52:31   : Could not make directory /oracle/orabase/product/10g/db_home/racg/init/uam_db2/uamdb/ reason: Permission denied

      这里报权限不够，查看目录权限

[oracle@uam_db2]#cd /oracle/orabase/product/10g/db_home/racg/init/
[oracle@uam_db2]#ls -lrt
total 0
drwxr-----    3 oracle   oinstall        256 Jan 18 04:55 uam_db1
drwxr-----    3 root     system          256 Jan 18 05:28 uam_db2

发现用户目录的权限root:system，对比两个节点的信息，发现这样不错。

故障处理方案：

  手动修改目录的owner.group后，crs能正常管理instance等信息

uam_db2@root[/oracle/orabase/product/10g/db_home/racg/init]chown oracle:oinstall uam_db2