第一次安装11gR2遇到很多问题，排查错误也用了好些时间。现将有关问题记录如下：

一、安装grid执行root.sh的时候不成功，如下错误

Errors in file :

ORA-27091: unable to queue I/O

ORA-15081: failed to submit an I/O operation to a disk

ORA-06512: at line 4

PROT-1: Failed to initialize ocrconfig

Command return code of 255 (65280) from command: /oracle/grid/bin/ocrconfig -upgrade grid oinstall

Failed to create Oracle Cluster Registry configuration, rc 255

CRS-2500: Cannot stop resource 'ora.crsd' as it is not running

CRS-4000: Command Stop failed, or completed with errors.

Command return code of 1 (256) from command: /oracle/grid/bin/crsctl stop resource ora.crsd -init

Stop of resource "ora.crsd -init" failed

Failed to stop CRSD

CRS-2673: Attempting to stop 'ora.asm' on 'rac1'

CRS-2677: Stop of 'ora.asm' on 'rac1' succeeded

CRS-2673: Attempting to stop 'ora.ctssd' on 'rac1'

CRS-2677: Stop of 'ora.ctssd' on 'rac1' succeeded

CRS-2673: Attempting to stop 'ora.cssdmonitor' on 'rac1'

CRS-2677: Stop of 'ora.cssdmonitor' on 'rac1' succeeded

CRS-2673: Attempting to stop 'ora.cssd' on 'rac1'

CRS-2677: Stop of 'ora.cssd' on 'rac1' succeeded

CRS-2673: Attempting to stop 'ora.gpnpd' on 'rac1'

CRS-2677: Stop of 'ora.gpnpd' on 'rac1' succeeded

CRS-2673: Attempting to stop 'ora.gipcd' on 'rac1'

CRS-2677: Stop of 'ora.gipcd' on 'rac1' succeeded

CRS-2673: Attempting to stop 'ora.mdnsd' on 'rac1'

CRS-2677: Stop of 'ora.mdnsd' on 'rac1' succeeded

Initial cluster configuration failed.See /oracle/grid/cfgtoollogs/crsconfig/rootcrs_rac1.log for details

解决方法：按照Linux: root.sh Fails to Format Voting disks when Placing OCR/Voting Disks on ASM Using asmlib [ID 955550.1]问题解决(引用ID 955550.1)

In this Document

Applies to:

Oracle Server - Enterprise Edition - Version: 11.2.0.0 to 11.2.0.2 - Release: 11.2 to 11.2

Linux x86

Red Hat Enterprise Linux Advanced Server x86-64 (AMD Opteron Architecture)

x86 64 bit (for Enterprise Linux only)

Linux x86-64

x86 32 bit (for Enterprise Linux only)

Grid Infrastructure, clusterware, CRS, voting disk

Symptoms

Oracle Grid Infrastructure installation with Oracle ASM using ASMlib fails during root.sh execution.

At the completion of Grid Infrastructure installation, root.sh script. executed on the first node in the cluster fails to format the OCR and voting disks which are placed on ASM storage.

Example of the error when root.sh fails:CRS-2676: Start of 'ora.ctssd' on 'auw2k3' succeededASM created and started successfully.DiskGroup DATA created successfully.Errors in file :ORA-27091: unable to queue I/OORA-15081: failed to submit an I/O operation to a diskORA-06512: at line 4PROT-1: Failed to initialize ocrconfigCommand return code of 255 (65280) from command: /u01/app/11.2.0/grid/bin/ocrconfig -upgrade oragrid oinstallFailed to create Oracle Cluster Registry configuration, rc 255CRS-2500: Cannot stop resource 'ora.crsd' as it is not runningCRS-4000: Command Stop failed, or completed with errors.Command return code of 1 (256) from command: /u01/app/11.2.0/grid/bin/crsctl stop resource ora.crsd -initStop of resource "ora.crsd -init" failed

Changes

New installation

Cause

Diskgroup is succesfully created, root.sh fails at generating the OCR keys when invoking ocrconfig because userid used for ASMlib driver differs from grid software owner.

Oracle Grid infrastructure installation generates a log file in $GRID_HOME/cfgtoollogs/crsconfig.

Per log file'rootcrs_.log':

2009-10-21 12:57:26: Querying for existing CSS voting disks2009-10-21 12:57:26: Performing initial configuration for cluster2009-10-21 12:57:28: Start of resource "ora.ctssd -init" Succeeded2009-10-21 12:57:28: Configuring ASM via ASMCA2009-10-21 12:57:28: Executing as oragrid: /u01/app/11.2.0/grid/bin/asmca -silent -diskGroupName DATA -diskList ORCL:ASMD40,ORCL:ASMD41 -redundancy EXTERNAL -configureLocalASM2009-10-21 12:57:28: Running as user oragrid: /u01/app/11.2.0/grid/bin/asmca -silent -diskGroupName DATA -diskList ORCL:ASMD40,ORCL:ASMD41 -redundancy EXTERNAL -configureLocalASM2009-10-21 12:57:28:   Invoking "/u01/app/11.2.0/grid/bin/asmca -silent -diskGroupName DATA -diskList ORCL:ASMD40,ORCL:ASMD41 -redundancy EXTERNAL -configureLocalASM" as user "oragrid"2009-10-21 12:58:02: Creating or upgrading OCR keys2009-10-21 12:58:04: Command return code of 255 (65280) from command: /u01/app/11.2.0/grid/bin/ocrconfig -upgrade oragrid oinstall2009-10-21 12:58:04: Failed to create Oracle Cluster Registry configuration, rc 2552009-10-21 12:58:04: Exiting exclusive mode2009-10-21 12:58:04: Command return code of 1 (256) from command: /u01/app/11.2.0/grid/bin/crsctl stop resource ora.crsd -init2009-10-21 12:58:04: Stop of resource "ora.crsd -init" failed

In above example the grid software owner isoragridbut the ASMlib device driver has been configured with owneroracle(different user) despite both being part of dba group

% /usr/sbin/oracleasm configureORACLEASM_ENABLED=trueORACLEASM_UID=oracleORACLEASM_GID=dbaORACLEASM_SCANBOOT=trueORACLEASM_SCANORDER=""ORACLEASM_SCANEXCLUDE=""

Solution

Please note all tasks below are done as the root user!1. To Resolve this problem, first deconfigure Grid.Run the following script. on any node that had root.sh executed on :

% $GRID_HOME/crs/install/rootcrs.pl -deconfig -force

2. Reconfigure ASMlib

Delete the ASMlib device/s used for OCR and voting ASM diskgroup, in example above the devices used are:

ORCL:ASMD40

ORCL:ASMD41

% /usr/sbin/oracleasm deletedisk ASMD40% /usr/sbin/oracleasm deletedisk ASMD41

3.  Re-configure ASMlib driver to use the correct userid, in this case "oragrid" user is the software owner for Grid Infrastructure installation, ensure followings are done on all nodes:

% /usr/sbin/oracleasm configure -u oragrid% /usr/sbin/oracleasm configure

4.  stop/start ASMlib on all nodes

% /usr/sbin/oracleasm exit% /usr/sbin/oracleasm init% /usr/sbin/oracleasm scandisks

5.  Create the ASMlib disks again used for Grid Infrastructure installation

% oracleasm createdisk ASMD40 /dev/sdh1% oracleasm createdisk ASMD41 /dev/sdi1

Then run the scandisks again on all nodes ->

% /usr/sbin/oracleasm scandisks

Now ASMlib devices should be correctly configured, this can be confirmed by running 'oracleasm configure' or  'ls -la /dev/oracleasm/disks/ '.

6.  Re-run 'root.sh'

二、在配置asm磁盘组的时候，通过asmca配置界面找不到磁盘。

解决方法：将init+ASM1.ORA中的asm_diskstring禁用掉，重新执行asmca进行配置完成

三、执行dbca建库时识别不到ASM磁盘组

解决方法：由于未执行usermod -G dba,asmdba,oper oracle而导致不能识别出ASM磁盘组。

四、在安装过程中，会提示一个错误“尝试安装Oracle RAC的系统不属于有效群集”

这是11gR2的一个bug，原因是inventory.xml里没有CRS=TRUE等标识.

修改/home/oracle/oracle11g/oraInventory/ContentsXML/inventory.xml里的值(两节点)

1、oracle1,oracle2进入$GRID_HOME/oui/bin

2、./runInstaller -attachHome ORACLE_HOME="/oracle_crs/product/11.2/crs" ORACLE_HOME_NAME="ORADB_11_2"

3、./runInstaller -updateNodeList ORACLE_HOME="/oracle_crs/product/11.2/crs" ORACLE_HOME_NAME="ORADB_11_2" "CLUSTER_NODES={oracle1,oracle2}" CRS=TRUE

修改完两节点的inventory.xml的值后，重新运行db的安装,一步一步的下去，便可完成数据库软件的安装(在此过程中我们选择只安装数据库软件，建库的工作用dbca来完成)

五、oracle base和software install目录

$ORACLE_BASE与$ORACLE_HOME目录不要存在上下级关系,否则在安装时会报错。

[INS-32026]The software Location specified should not be under Oracle base location

六、./root.sh

……(省略)

/oracle/grid/bin/clscfg.bin: error while loading shared libraries: libcap.so.1: cannot open shared object file: No such file or directory

Failed to create keys in the OLR, rc = 127, 32512

解决方法：

从系统光盘上安装libcap软件包即可。在很多RAC安装指导文档中都会说要将安装系统时要将所有的开发包都要装。

再次运行root.sh时，会提示你已经配置过一次了，要删掉之前的配置

执行$ORACLE_HOME/install/roothas.pl -deconfig -force