OracleBlog.cn

新版本的opatch使用lsinventory报错

小荷 — Mon, 25 Jan 2010 17:13:17 +0800

今天在给一个省打一个9208侦听内存泄漏的patch，打的时候，没有报错，但是在opatch lsinventory的时候，却报错了：

oracle@tj_dc01:/oracle/app/oracle/product/9.2.0 > opatch lsinventory

Oracle Interim Patch Installer version 1.0.0.0.63
Copyright (c) 2009 Oracle Corporation. All Rights Reserved..

Oracle recommends you to use the latest OPatch version
and read the OPatch documentation available in the OPatch/docs
directory for usage. For information about the latest OPatch and
other support-related issues, refer to document ID 293369.1
available on My Oracle Support (https://myoraclesupport.oracle.com)

Oracle Home : /oracle/app/oracle/product/9.2.0
Oracle Home Inventory : /oracle/app/oracle/oraInventory
Central Inventory : /oracle/app/oracle/oraInventory
   from : /var/opt/oracle/oraInst.loc
OUI location : /oracle/app/oracle/oui
OUI shared library : /oracle/app/oracle/oui/bin/hpunix/liboraInstaller.so
Java location : /oracle/app/oracle/jre/1.3.1/bin/java
Log file location : /oracle/app/oracle/product/9.2.0/.patch_storage/<patch ID>/*.log

Creating log file "/oracle/app/oracle/product/9.2.0/.patch_storage/LsInventory__01-25-2010_16-25-28.log"

Exception in thread "main" java.lang.NoClassDefFoundError: javax/xml/transform/Source
   at opatch.LsInventory.getOneOffList(LsInventory.java:675)
   at opatch.LsInventory.main(LsInventory.java:247)
Result:

There is no Interim Patch

OPatch succeeded.

这个opatch是从metalink上下载的最新的opatch，而且也是对应了HP IA的os平台，opatch的版本应该不会错。如果错了，opatch apply的时候就应该报了，怎么还会到lsinventory的时候才报错。

上metalink上查了这个问题(ID 417275.1)，原来最新版本的opatch（p6880880_101000_HPUX-IA64.zip）要求需要在OUI 10.1.0.x和JRE 1.4.2，而这2个是需要在9205或者9206升级的时候才会安装，在基线版本9202，以及9203，9204的版本中只有OUI 2.2.0.x和JRE 1.3.1，9207和9208这2个版本是不含OUI和JRE的升级的。所以当我从HP IA的基线版本9202升级到9208，再使用opatch打完补丁，用opatch lsinventory检查的时候，就会因为OUI的问题出现上述的报错。

oracle上给的解决办法是：

1.下载patch 3540563----很可惜，这个patch没有HP IA的平台。
2.降级9208，重新升级到9205，再升级到9208。因为在升级到9205的时候，就升级了OUI和JRE，所以不会报错了----扯淡啊！我怎么可能为了这点小事去downgrate。

这里我提供2个方法来进行验证是否正确安装了patch：

1.在patch目录，用opatch query -verify：

oracle@tj_dc01:/opt/ora_inst/ora_patch9208/5576565 > opatch query -verify

Oracle Interim Patch Installer version 1.0.0.0.63
Copyright (c) 2009 Oracle Corporation. All Rights Reserved..

Oracle recommends you to use the latest OPatch version
and read the OPatch documentation available in the OPatch/docs
directory for usage. For information about the latest OPatch and
other support-related issues, refer to document ID 293369.1
available on My Oracle Support (https://myoraclesupport.oracle.com)

Query ...

Please enter the patch location:
/opt/ora_inst/ora_patch9208/5576565

---------- Query starts ------------------

Patch ID: 5576565

The contents of the file: /var/opt/oracle/oraInst.loc

#Oracle Installer Location File Location

#Fri Jan 22 10:16:22 GMT+08:00 2010

inventory_loc=/oracle/app/oracle/oraInventory

inst_group=dba

Parsing /oracle/app/oracle/oraInventory/ContentsXML/comps.xml

Found "oracle.swd.oui" version "2.2.0.14.0"
on "/oracle/app/oracle/oui"

Verifying patch...
Verifying that patch ID is in Oracle Home inventory.

Verifying archive files.

Comparing "/opt/ora_inst/ora_patch9208/5576565/files/lib32/libnl9.a/nlpc.o" and "/oracle/app/oracle/product/9.2.0/.patch_storage/verify/lib32/libnl9.a/nlpc.o"

Comparing "/opt/ora_inst/ora_patch9208/5576565/files/lib32/libnl9.a/snlpc.o" and "/oracle/app/oracle/product/9.2.0/.patch_storage/verify/lib32/libnl9.a/snlpc.o"

Comparing "/opt/ora_inst/ora_patch9208/5576565/files/lib/libnl9.a/nlpc.o" and "/oracle/app/oracle/product/9.2.0/.patch_storage/verify/lib/libnl9.a/nlpc.o"

Comparing "/opt/ora_inst/ora_patch9208/5576565/files/lib/libnl9.a/snlpc.o" and "/oracle/app/oracle/product/9.2.0/.patch_storage/verify/lib/libnl9.a/snlpc.o"

Comparing "/opt/ora_inst/ora_patch9208/5576565/files/lib32/libnl9.a/nlpc.o" and "/oracle/app/oracle/product/9.2.0/.patch_storage/verify/lib32/libnl9.a/nlpc.o"

Comparing "/opt/ora_inst/ora_patch9208/5576565/files/lib32/libnl9.a/snlpc.o" and "/oracle/app/oracle/product/9.2.0/.patch_storage/verify/lib32/libnl9.a/snlpc.o"

Comparing "/opt/ora_inst/ora_patch9208/5576565/files/lib/libnl9.a/nlpc.o" and "/oracle/app/oracle/product/9.2.0/.patch_storage/verify/lib/libnl9.a/nlpc.o"

Comparing "/opt/ora_inst/ora_patch9208/5576565/files/lib/libnl9.a/snlpc.o" and "/oracle/app/oracle/product/9.2.0/.patch_storage/verify/lib/libnl9.a/snlpc.o"

OPATCH_VERIFY_OK: Patch has been applied.

---------- Query ends -------------------

Do you want to check out another patch?
Please respond Y|N >
N
Query stops.

OPatch succeeded.
oracle@tj_dc01:/opt/ora_inst/ora_patch9208/5576565 >

2.用一个老版本的opatch检查也是可以的。

oracle@tj_dc01:/opt/ora_inst/ora_patch9208/OPatch > opatch lsinventory

Oracle Interim Patch Installer version 1.0.0.0.53
Copyright (c) 2005 Oracle Corporation. All Rights Reserved..

We recommend you refer to the OPatch documentation under
OPatch/docs for usage reference. We also recommend using
the latest OPatch version. For the latest OPatch version
and other support related issues, please refer to document
293369.1 which is viewable from metalink.oracle.com

Oracle Home = /oracle/app/oracle/product/9.2.0
Location of Oracle Universal Installer components = /oracle/app/oracle/oui
Location of OraInstaller.jar = "/oracle/app/oracle/oui/lib"
Oracle Universal Installer shared library = /oracle/app/oracle/oui/bin/hpunix/liboraInstaller.so
Location of Oracle Inventory Pointer = /var/opt/oracle/oraInst.loc
Location of Oracle Inventory = /oracle/app/oracle/oraInventory
Path to Java = /oracle/app/oracle/jre/1.3.1/bin/java
Log file = /oracle/app/oracle/product/9.2.0/.patch_storage/<patch ID>/*.log

Creating log file "/oracle/app/oracle/product/9.2.0/.patch_storage/LsInventory__01-25-2010_16-25-14.log"

Result:

Installed Patch List:
=====================
1) Patch 5576565 applied on Sun Jan 24 10:34:11 GMT+08:00 2010
[ Base Bug(s): 5576565 ]

OPatch succeeded.
oracle@tj_dc01:/opt/ora_inst/ora_patch9208/OPatch >

附件为这个老版本的opatch：点击下载：p2617419_10102_generic_opatch

在hp pa 11.23上安装9206patch

小荷 — Sun, 17 Jan 2010 18:09:43 +0800

hp pa算是一个比较老版本的hpux操作系统，目前hp已经不出hp pa的版本，只出hp ia的版本。目前ia的os版本为11.23和11.31，在之前的hp pa的主要os版本是11.00和11.11。
这次在某省安装的机器比较奇怪，是hp pa的机器，却是安装11.23的操作系统版本。在正常安装完成9201的基线版本后在升级9206的patch，遇到报错了：

$ cd D*1
$ ls
install, patchnote.htm, response, runInstaller, stage
$ export DISPLAY=192.168.0.4:0.0
$ ./runInstaller
Starting Oracle Universal Installer...

Checking installer requirements...

Checking operating system version: must be B.11.00 or B.11.11. Actual B.11.23
Failed <<<<

Exiting Oracle Universal Installer, log for this session can be found at /opt/acle/oraInventory/logs/installActions2010-01-10_03-38-08PM.log
$

根据log去做进一步的检查了：

$ vi installActions2010-01-10_03-38-08PM.log
"installActions2010-01-10_03-38-08PM.log" 9 lines, 385 characters
Using paramFile: /usr/ora_inst/orapatch_9206/Disk1/install/oraparam.ini

Checking installer requirements...

Checking operating system version: must be B.11.00 or B.11.11. Actual B.11.23
Failed <<<<

Exiting Oracle Universal Installer, log for this session can be found at /opt/oracle/oraInventory/logs/installActions2010-01-10_03-3
8-08PM.log
~

在这里我们看到在做系统环境的检查时，调用了/usr/ora_inst/orapatch_9206/Disk1/install/oraparam.ini这个参数文件，因此我们去看这个参数文件：

$ cat /usr/ora_inst/orapatch_9206/Disk1/install/oraparam.ini
[Oracle]
DISTRIBUTION=TRUE
SOURCE=../stage/products.xml
LICENSE_LOCATION=
JRE_LOCATION=../stage/Components/oracle.swd.jre/1.4.2.0.0/1/DataFiles
JRE_MEMORY_OPTIONS=" -mx96m"
DEFAULT_HOME_LOCATION=
DEFAULT_HOME_NAME=OUIHome
NO_BROWSE=/net
NLS_ENABLED=TRUE
BOOTSTRAP=TRUE
OUI_VERSION=10.1.0.3.0
AUTOOUI=ALWAYS
#SHOW_HOSTNAME=ALWAYS_SHOW shows the hostname panel always
#SHOW_HOSTNAME=NEVER_SHOW does not the hostname panel
#SHOW_HOSTNAME=CONDITION_SHOW shows the hostname panel on condition
SHOW_HOSTNAME=NEVER_SHOW
#THIN_JDBC_FILENAME is optional and defaults to classes12.jar
#The value specified for this should be packaged with OUI, and should
#be relative to <OUI expanded stagedir>/jlib/
THIN_JDBC_FILENAME=classes12.jar
#JRE_OSDPARAM is to set OS dependent param for JRE ( mainly for native VM in 1.3.1)
#JRE_OSDPARAM is optional and should be set to -native for the JRE's
#that support native VM ( mainly for Unix platforms ), in JRE 1.3.1
#For JRE 1.4.1 this should be set to empty or the type of VM that is
#supported client/server. The default value is -native in UNIX platforms
#that supports native VM
#Unix supporting native - JRE_OSDPARAM="-native"
#Unix NOT supporting native and 1.4.1 - JRE_OSDPARAM=""
JRE_OSDPARAM=""
#CLUSTERWARE={"oracle.crs","10.1.0.2.0"}
#RUN_OUICA specifies the batch script name that needs to be run
#The script is ouica.bat for win32, and ouica.sh for solaris.
#If the value is not specified, then the OUICA script is not run
RUN_OUICA=ouica.sh

[Certified Versions]
#You can customise error message shown for failure through CERTIFIED_VERSION_FAILURE_MESSAGE
Solaris=5.6,5.7,5.8,5.9
Linux=redhat-2.1AS
HPUX=B.11.00,B.11.11
Decunix=V5.1A,V5.1
AIX=5.1.0.0

[Solaris-5.6-required]
#Minimum temp space required by OUI, Increase it if your product installation so requires
#Temp space is required for bootstrap and during installation
#The TEMP_SPACE should be the first pre-req to be checked
TEMP_SPACE=250
SWAP_SPACE=500
#Use TEMP_SPACE_FAILURE_MESSAGE for custom error message for Swap space check failure
#PACKAGES required by JRE and Product, /usr/bin/pkginfo gets list of packages on system
PACKAGES=
#Use PACKAGES_FAILURE_MESSAGE for custom error message for PACKAGES check failure

#Use CPU_FAILURE_MESSAGE to display custom error message for CPU check failure
#MIN_DISPLAY_COLORS required by OUI, use /usr/openwin/bin/xdpyinfo to get the colors
MIN_DISPLAY_COLORS=256
#Use MIN_DISPLAY_COLORS_FAILURE_MESSAGE for custom error message for DISPLAY colors check failure

[Solaris-5.6-optional]
#PATCHES required by JRE and Product, /usr/bin/showrev -p gets list of patches on system
PATCHES=
#Use PATCHES_FAILURE_MESSAGE for custom error message for PATCHES check failure

[Solaris-5.7-required]
#Minimum temp space required by OUI, Increase it if your product installation so requires
#Temp space is required for bootstrap and during installation
#The TEMP_SPACE should be the first pre-req to be checked
TEMP_SPACE=250
SWAP_SPACE=500
#PACKAGES required by JRE and Product, /usr/bin/pkginfo gets list of packages on system
PACKAGES=

#MIN_DISPLAY_COLORS required by OUI, use /usr/openwin/bin/xdpyinfo to get the colors
MIN_DISPLAY_COLORS=256

[Solaris-5.7-optional]
#PATCHES required by JRE and Product, /usr/bin/showrev gets list of patches on system
PATCHES=

[Solaris-5.8-required]
#Minimum temp space required by OUI, Increase it if your product installation so requires
#Temp space is required for bootstrap and during installation
#The TEMP_SPACE should be the first pre-req to be checked
TEMP_SPACE=250
SWAP_SPACE=500
#PACKAGES required by JRE and Product, /usr/bin/pkginfo gets list of packages on system
PACKAGES=

#MIN_DISPLAY_COLORS required by OUI, use /usr/openwin/bin/xdpyinfo to get the colors
MIN_DISPLAY_COLORS=256

[Solaris-5.8-optional]
#PATCHES required by JRE and Product, /usr/bin/showrev gets list of patches on system
PATCHES=108652-66,108921-16,108940-53,108773-18,111310-01,109147-24,111308-03,111111-03,112396-02,110386-03,111023-02,108987-13,108528-21,108989-02,108993-18

[Solaris-5.9-required]
#Minimum temp space required by OUI, Increase it if your product installation so requires
#Temp space is required for bootstrap and during installation
#The TEMP_SPACE should be the first pre-req to be checked
TEMP_SPACE=250
SWAP_SPACE=500
#PACKAGES required by JRE and Product, /usr/bin/pkginfo gets list of packages on system
PACKAGES=

#MIN_DISPLAY_COLORS required by OUI, use /usr/openwin/bin/xdpyinfo to get the colors
MIN_DISPLAY_COLORS=256

[Solaris-5.9-optional]
#PATCHES required by JRE and Product, /usr/bin/showrev gets list of patches on system
PATCHES=

#[HPUX-B.11.11-required]
#PATCHES required by JRE and Product, /usr/sbin/swlist -l patch gets list of patches on system
#PATCHES=
#PACKAGES required by JRE and Product, /usr/sbin/swlist -l product gets list of packages on system
#PACKAGES=
#Patch bundles required for Oracle.
#/usr/sbin/swlist -l bundle gives list of bundles.
#Each item should follow the format "bundle:version"
#ex.,GOLDBASE11i:B.11.11.0106.9,GOLDAPPS11i:B.11.11.0106.9
#BUNDLE=
#Minimum swap space required by OUI, Increase it if your product installation so requires
#SWAP_SPACE=150
#BUNDLE_FAILURE_MESSAGE=Some patch bundles were not found on the system. Please verify if the system is at a higher bundle level.

[Linux-redhat-2.1AS-required]
#Minimum kernel version supported on this flavour of Linux
KERNEL=2.4.9
#Minimum glibc version supported on this flavour of Linux
GLIBC=glibc-2.2.4-25
#PACKAGES required by JRE and Product, "rpm -qa" gets list of packages on system
PACKAGES=
#Minimum swap space required by OUI, Increase it if your product installation so requires
SWAP_SPACE=150

[Decunix-V5.1A-required]
#PATCHES required by JRE and Product, /usr/sbin/dupatch -track -type kit
#gets list of patches on system
PATCHES=
#PACKAGES required by JRE and Product, /usr/sbin/setld -i gets
#list of packages on system
PACKAGES=
#Minimum swap space required by OUI, Increase it if your product installation so requires
SWAP_SPACE=150

[Decunix-V5.1-required]
#PATCHES required by JRE and Product, /usr/sbin/dupatch -track -type kit
#gets list of patches on system
PATCHES=
#PACKAGES required by JRE and Product, /usr/sbin/setld -i gets
#list of packages on system
PACKAGES=
#Minimum swap space required by OUI, Increase it if your product installation so requires
SWAP_SPACE=150

[AIX-5.1.0.0-required]
#PATCHES required by JRE and Product, /usr/sbin/instfix -a -ivk
#gets list of patches on system
PATCHES=
#PACKAGES required by JRE and Product, /usr/bin/lslpp -l
#gets list of packages on system
PACKAGES=
#Minimum swap space required by OUI, Increase it if your product installation so requires
SWAP_SPACE=150

$

在这里我们看到hp pa9206的patch要求

HPUX=B.11.00,B.11.11

我们目前hp pa 11.23没在这个里面，因此我当时尝试了在这个文件中把11.23也添加进去：

……
[Certified Versions]
#You can customise error message shown for failure through CERTIFIED_VERSION_FAILURE_MESSAGE
Solaris=5.6,5.7,5.8,5.9
Linux=redhat-2.1AS
HPUX=B.11.00,B.11.11,B.11.23
Decunix=V5.1A,V5.1
AIX=5.1.0.0
……

再次执行runInstaller，搞定。

当时身处某省由于通信管制没法上网，回来之后，查了metalink，确定是个unpublished Bug:4054600.详见文档ID 294682.1，处理方法，也和我的一样。
我猜想是oracle出hp pa9206的patch的时候，hp pa还没出11.23的os版本，而等hp pa出11.23的os版本的时候，oracle已经出9208的patch了，因此在9208的oraparam.ini文件中，我们看到了B.11.23这个配置，而在9206这个patch，没配置。

修改vmware rac的IP网段

小荷 — Sun, 17 Jan 2010 16:25:57 +0800

之前在vmware上搭了一个rac，在配置rac的时候，由于当时的电脑是接在路由器上的，因此将public和vip配置在了需要连接路由器的IP上，用的是bridge的连接方式。但是也造成了一个不便，就是在没有连接路由的情况下，public和vip无法连接，因此我希望用host only而不是bridge的连接方式，将rac的vip改为vmware的虚拟网卡的IP。

改造之前：

192.168.190.131 rac1
192.168.1.31 rac1-vip
10.10.10.31 rac1-priv
192.168.190.132 rac2
192.168.1.32 rac2-vip
10.10.10.32 rac2-priv
192.168.190.133 rac3
192.168.1.33 rac3-vip
10.10.10.33 rac3-priv

其中rac1的2块网卡，rac2和rac3的类似：
eth0 192.168.1.131 用bridge连接
eth1 10.10.10.31 用host only连接

另外，vmware的虚拟网卡，在这次改造中不做变化：

VMware Network Adapter VMnet1：192.168.190.1
VMware Network Adapter VMnet8：192.168.174.1

路由器的网关地址：192.168.1.200

改造后：

192.168.190.131 rac1
192.168.190.31 rac1-vip
10.10.10.31 rac1-priv
192.168.190.132 rac2
192.168.190.32 rac2-vip
10.10.10.32 rac2-priv
192.168.190.133 rac3
192.168.190.33 rac3-vip
10.10.10.33 rac3-priv

其中rac1的2块网卡，rac2和rac3的类似：
eth0 192.168.190.131 用host only连接
eth1 10.10.10.31 用host only连接

修改步骤：
1.停数据库和crs：

只在一个节点上运行：
crs_stop -all
检查：
rac1-> crs_stat -t
Name Type Target State Host
------------------------------------------------------------
ora.devdb.db application OFFLINE OFFLINE
ora....b1.inst application OFFLINE OFFLINE
ora....b2.inst application OFFLINE OFFLINE
ora....b3.inst application OFFLINE OFFLINE
ora....SM1.asm application OFFLINE OFFLINE
ora....C1.lsnr application OFFLINE OFFLINE
ora.rac1.gsd application OFFLINE OFFLINE
ora.rac1.ons application OFFLINE OFFLINE
ora.rac1.vip application OFFLINE OFFLINE
ora....SM2.asm application OFFLINE OFFLINE
ora....C2.lsnr application OFFLINE OFFLINE
ora.rac2.gsd application OFFLINE OFFLINE
ora.rac2.ons application OFFLINE OFFLINE
ora.rac2.vip application OFFLINE OFFLINE
ora....SM3.asm application OFFLINE OFFLINE
ora....C3.lsnr application OFFLINE OFFLINE
ora.rac3.gsd application OFFLINE OFFLINE
ora.rac3.ons application OFFLINE OFFLINE
ora.rac3.vip application OFFLINE OFFLINE
rac1->

在所有节点上执行：
[root@rac1 bin]# /etc/init.d/init.crs stop
Shutting down Oracle Cluster Ready Services (CRS):
Stopping resources.
Successfully stopped CRS resources
Stopping CSSD.
Shutting down CSS daemon.
Shutdown request successfully issued.
Shutdown has begun. The daemons should exit soon.
[root@rac1 bin]#

2.修改hosts的文件：

[root@rac1 bin]# cat /etc/hosts
# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1 localhost
192.168.190.131 rac1.mycorpdomain.com rac1
192.168.190.31 rac1-vip.mycorpdomain.com rac1-vip
10.10.10.31 rac1-priv.mycorpdomain.com rac1-priv
192.168.190.132 rac2.mycorpdomain.com rac2
192.168.190.32 rac2-vip.mycorpdomain.com rac2-vip
10.10.10.32 rac2-priv.mycorpdomain.com rac2-priv
192.168.190.133 rac3.mycorpdomain.com rac3
192.168.190.33 rac3-vip.mycorpdomain.com rac3-vip
10.10.10.33 rac3-priv.mycorpdomain.com rac3-priv

## ntp server
192.168.190.1 HEJIANMIN
[root@rac1 bin]#

3.修改主机的IP。

4.由于采用了ocfs2，将crs和votingdisk放在ocfs2上，修改主机IP之后，重启主机之前，也需要修改ocfs2的配置：

原来：
[root@rac1 bin]# cat /etc/ocfs2/cluster.conf
node:
   ip_port = 7777
   ip_address = 192.168.1.131
   number = 0
   name = rac1
   cluster = ocfs2

node:
   ip_port = 7777
   ip_address = 192.168.1.132
   number = 1
   name = rac2
   cluster = ocfs2

node:
   ip_port = 7777
   ip_address = 192.168.1.133
   number = 2
   name = rac3
   cluster = ocfs2

cluster:
   node_count =3
   name = ocfs2
[root@rac1 bin]#

现改成：
[root@rac1 bin]# cat /etc/ocfs2/cluster.conf
node:
   ip_port = 7777
   ip_address = 192.168.190.131
   number = 0
   name = rac1
   cluster = ocfs2

node:
   ip_port = 7777
   ip_address = 192.168.190.132
   number = 1
   name = rac2
   cluster = ocfs2

node:
   ip_port = 7777
   ip_address = 192.168.190.133
   number = 2
   name = rac3
   cluster = ocfs2

cluster:
   node_count =3
   name = ocfs2
[root@rac1 bin]#

注意如果不改ocfs配置，会在启动crs的时候报错，共享存储只会在第一个启动的机器上挂载。

5.重启主机，注意将网卡模式改为host only。

6.在3个节点上启动crs服务：/etc/init.d/init.crs start，重启后记得再次crs_stop。

7.用oifconfig来修改网卡被oracle使用的方式：

#检查：
[root@rac1 bin]# ./oifcfg getif
eth0 192.168.1.0 global public
eth1 10.10.10.0 global cluster_interconnect
[root@rac1 bin]#
[root@rac1 bin]#
#修改：
[root@rac1 bin]# ./oifcfg delif -global eth0/192.168.1.0
[root@rac1 bin]# ./oifcfg setif -global eth0/192.168.190.0:public
[root@rac1 bin]#
[root@rac1 bin]#
#再次检查：
[root@rac1 bin]# ./oifcfg getif
eth0 192.168.190.0 global public
eth1 10.10.10.0 global cluster_interconnect
[root@rac1 bin]#

8.修改vip，注意这里是vip，不是public ip：

[root@rac1 bin]# ./srvctl modify nodeapps -n rac1 -A 192.168.190.31/255.255.255.0/eth0
[root@rac1 bin]# ./srvctl modify nodeapps -n rac2 -A 192.168.190.32/255.255.255.0/eth0
[root@rac1 bin]# ./srvctl modify nodeapps -n rac3 -A 192.168.190.33/255.255.255.0/eth0
[root@rac1 bin]#

9.修改侦听中的vip为新地址。

10.crs_start 启动rac。

写正确了tnsnames文件却tnsping无法ping通

小荷 — Wed, 06 Jan 2010 18:09:00 +0800

今天遇到个问题比较奇怪，tnsnames.ora明明配置正确，在其中写的IP是10.1.2.40，但是在tnsping中却解析到10.1.2.45：

[oracle@si_adm02 admin]$ cat tnsnames.ora
# TNSNAMES.ORA.SI_WWW01 Network Configuration File: /ora9i/app/oracle/product/9.2.0/network/admin/tnsnames.ora.si_www01
# Generated by Oracle configuration tools.
EXTPROC_CONNECTION_DATA =
(DESCRIPTION =
   (ADDRESS_LIST =
   (ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC))
   )
   (CONNECT_DATA =
   (SID = PLSExtProc)
   (PRESENTATION = RO)
   )
)

dev2 =
(DESCRIPTION =
   (ADDRESS_LIST =
   (ADDRESS = (PROTOCOL = TCP)(HOST = 10.1.2.40)(PORT = 1521))
   )
   (CONNECT_DATA =
   (SERVICE_NAME =misc1)
   )
)
[oracle@si_adm02 admin]$

[oracle@si_adm02 admin]$ tnsping dev2

TNS Ping Utility for Linux: Version 9.2.0.4.0 - Production on 06-1Ղ -2010 15:03:15

Copyright (c) 1997 Oracle Corporation. All rights reserved.

Used parameter files:
/oracle/product/9.2.0/network/admin/sqlnet.ora

Used TNSNAMES adapter to resolve the alias
Attempting to contact (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = 10.1.3.54)(PORT = 1521))) (CONNECT_DATA = (SERVICE_NAME = ora9i)))
OK (0 msec)
[oracle@si_adm02 admin]$

而且我把tnsnames.ora改名成别的名称，这个service name依然可以被解析到。

既然都改名了，难道还有别的tnsnames.ora文件在被使用？用find名称查了一下：

[oracle@si_adm02 admin]$ find / -name tnsnames.ora >hjm.txt
[oracle@si_adm02 admin]$ cat hjm.txt
/opt/aspire/product/imallcpp/product/9.2.0/network/admin/samples/tnsnames.ora
/opt/aspire/product/imallcpp/product/9.2.0/network/admin/tnsnames.ora
/opt/instantclient_10_2/tnsnames.ora
/opt/instantclient_10_2_4/tnsnames.ora
/oracle/product/9.2.0/network/admin/samples/tnsnames.ora

发现在其他路径下还是有tnsnames.ora文件的，而且用于解析的，就是其他路径的这个文件。

在此，基本的问题就搞明白了，由于用了其他路径的tnsnames.ora，所以无法正确解析。为解决这个问题，我们指定$ORACLE_HOME下的network/admin路径，我们export一下TNS的路径，或者在环境变量中加上TNS_ADMIN，就能解决这个问题了：

[oracle@si_adm02 admin]$ export TNS_ADMIN=/oracle/product/9.2.0/network/admin
[oracle@si_adm02 admin]$
[oracle@si_adm02 admin]$
[oracle@si_adm02 admin]$ tnsping dev2

TNS Ping Utility for Linux: Version 9.2.0.4.0 - Production on 06-1Ղ -2010 15:12:03

Copyright (c) 1997 Oracle Corporation. All rights reserved.

Used parameter files:
/oracle/product/9.2.0/network/admin/sqlnet.ora

Used TNSNAMES adapter to resolve the alias
Attempting to contact (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = 10.1.2.40)(PORT = 1521))) (CONNECT_DATA = (SERVICE_NAME =misc1)))
OK (10 msec)
[oracle@si_adm02 admin]$

最后，查到原来在root用户下，已经指定了这个TNS_ADMIN：

[root@si_adm02 root]# id
uid=0(root) gid=0(root) groups=0(root),1(bin),2(daemon),3(sys),4(adm),6(disk),10(wheel)
[root@si_adm02 root]# env |grep -i tns
TNS_ADMIN=/opt/instantclient_10_2
[root@si_adm02 root]#

应该是oracle用户没有配TNS_ADMIN，且oracle用户又继承了root的环境变量，所以在tnsping的时候，没有去正确的路径找tnsnames.ora文件，所以，解析不到了。

legato备份时不执行备份直接写index

小荷 — Tue, 05 Jan 2010 17:55:53 +0800

今天某省的同事来找，说备份有问题。从log上看备份似乎成功了，但是在legato的界面中看还是一个红叉。于是登录上legato的界面再次测试了备份，发现monitor中日志如下：

从日志上看，legato的备份在没有进行rman的备份前，就直接去写index了。在正常情况下，legato应该是先用nmo调用rman的备份，备份完之后，再写index。
问了当地的同事，备份也确实没备份成功，归档日志还留在原地。

进一步观察backup server上c:\program files\legato\logs\daemon.log的信息：

01/05/10 16:46:23 nsrd: savegroup info: starting OracleArch (with 1 client(s))
01/05/10 16:46:42 savegrp: jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch will retry 1 more time(s)
01/05/10 16:46:56 savegrp: jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch will retry 0 more time(s)
01/05/10 16:46:56 nsrd: media info: suggest mounting 000700L2 on jx_bak01 for writing to pool 'index'
01/05/10 16:46:56 nsrd: media waiting event: Waiting for 1 writable volumes to backup pool 'index' tape(s) on jx_bak01
01/05/10 16:46:56 nsrd: media info: loading volume 000700L2 into \\.\Tape0
01/05/10 16:47:06 nsrmmd #13: Start nsrmmd #13, with PID 8020, at HOST jx_bak01
01/05/10 16:47:13 nsrd: \\.\Tape0 1:Verify label operation in progress
01/05/10 16:47:34 nsrd: \\.\Tape0 1:Mount operation in progress
01/05/10 16:48:00 nsrd: media event cleared: Waiting for 1 writable volumes to backup pool 'index' tape(s) on jx_bak01
01/05/10 16:48:00 nsrd: jx_bak01:index:jx_db saving to pool 'index' (000700L2)
01/05/10 16:48:01 nsrd: jx_bak01:index:jx_db done saving to pool 'index' (000700L2) 93 KB
01/05/10 16:48:06 nsrd: jx_bak01:bootstrap saving to pool 'index' (000700L2)
01/05/10 16:48:07 nsrmmdbd: media db is saving its data. This may take a while.
01/05/10 16:48:07 nsrmmdbd: media db is open for business.
01/05/10 16:48:08 nsrd: jx_bak01:bootstrap done saving to pool 'index' (000700L2) 2645 KB
01/05/10 16:48:15 nsrd: savegroup info: Added 'jx_bak01' to the group 'OracleArch' for bootstrap backup.
01/05/10 16:48:15 nsrd: savegroup alert: OracleArch completed, total 2 client(s), 0 Hostname(s) Unresolved, 1 Failed, 1 Succeeded. (jx_db Failed)
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch nsrnmostart returned status of 255
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch /opt/networker/bin/nsrnmo1 exiting.
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch RMAN script execution is not successful. RMAN exited with return code '1'.
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch 1 retry attempted
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch nsrnmostart returned status of 255
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch /opt/networker/bin/nsrnmo1 exiting.
* jx_db:/oracle/app/oracle/product/9.2.0/bin/oraclearch RMAN script execution is not successful. RMAN exited with return code '1'.
01/05/10 16:48:16 nsrd: runq: NSR group OracleArch exited with return code 1.
01/05/10 16:48:16 nsrmmdbd: Starting compression of media database
01/05/10 16:48:16 nsrmmdbd: Finished compression of media database
01/05/10 16:48:16 nsrd: index notice: nsrim has finished cross checking the media db
Blat v2.4 w/GSS encryption (build : Jan 15 2005 08:32:11)

Sending d:\1.txt to weidc@aspire-tech.com, mon724@aspire-tech.com, jxsupport@aspire-tech.com
Subject: LEGATO_warnning!!
Login name is ss_mon@aspire-tech.com
Try number 1 of 3.
01/05/10 16:48:54 nsrd: media warning: \\.\Tape0 reading: 有更多数据可用。
01/05/10 16:48:54 nsrd: media warning: \\.\Tape0 reading: 有更多数据可用。
01/05/10 16:48:54 nsrd: media warning: \\.\Tape0 reading: 有更多数据可用。
01/05/10 16:48:54 nsrd: media warning: \\.\Tape0 reading: 有更多数据可用。
01/05/10 16:49:09 nsrd: media info: verification of volume "000700L2", volid 3987777691 succeeded.
01/05/10 16:49:09 nsrd: write completion notice: Writing to volume 000700L2 complete

在这里面我们看到“RMAN script execution is not successful. RMAN exited with return code ‘1′.”

于是我们到backup client，即db主机上去看备份的log：/nsr/applogs/msglog.log：

connected to recovery catalog database

connected to target database: JXMISC (DBID=2525447544)

allocated channel: t1
channel t1: sid=486 devtype=SBT_TAPE
channel t1: NMO v4.1.0.0
sent command to channel: t1

allocated channel: t2
channel t2: sid=434 devtype=SBT_TAPE
channel t2: NMO v4.1.0.0
sent command to channel: t2
"msglog.log" [Incomplete last line] 658103 lines, 35000290 characters
13> release channel t1;
14> release channel t2;
15> }
16>
connected to recovery catalog database

connected to target database: JXMISC (DBID=2525447544)

RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of allocate command at 01/05/2010 16:46:45
RMAN-03014: implicit resync of recovery catalog failed
RMAN-03009: failure of partial resync command on default channel at 01/05/2010 16:46:45
ORA-01536: space quota exceeded for tablespace 'RMAN_DATA'

发现原来是在同步catalog的时候，catalog库的rman用户的quato到了。我们进行以下处理：

SQL> grant unlimited tablespace to rman;

授权成功。

SQL>
Status Name Type Extent Man Total Size (M Used (M) Free (M) Used %
--------- ------------------------------ --------- ---------- ------------- ------------- ------------- -----
-
ONLINE SYSTEM PERMANENT LOCAL 500.000 192.875 307.125 38.
8
ONLINE RMAN_DATA PERMANENT LOCAL 500.000 83.000 417.000 16.
0
ONLINE TEMP TEMPORARY LOCAL 200.000 6.000 194.000 3.
0
ONLINE UNDOTBS1 UNDO LOCAL 200.000 1.500 198.500 0.
5
ONLINE TOOLS PERMANENT LOCAL 10.000 .063 9.938 0.
3
ONLINE INDX PERMANENT LOCAL 25.000 .063 24.938 0.
5
ONLINE USERS PERMANENT LOCAL 25.000 .063 24.938 0.
5

已选择7行。

SQL>