10gR2Clusterware Installation Steps

10gR2Clusterware Installation Steps
1. Install the Operating System (omitted)Note: I use the OEL 5u5 version here. Because it comes with the environment variable installation package provided by Oracle, It is very convenient to install the Oracle software. Therefore, we recommend that you use this version.
2. Configure the local yum Source
-- Mount the installation disk to mediamount-t iso9660/dev/cdrom/media
-- Configure repos to add the following content vi/etc/yum. repos. d/oel5.repos [oel5] name = oel 5 DVDbasurl = file: // media/Servergpgcheck = 0 enabled = 1 -- 1 indicates that this repos is enabled
Iii. Configure the trust relationship (oracle user)
-- Run ssh-keygen-t rsassh-keygen-t dsa on node 1 and node 2 respectively.
-- Run Node 1: cat ~ /. Ssh/*. pub> ~ /. Ssh/authorized_keysssh rac2 cat ~ /. Ssh/*. pub> ~ /. Ssh/authorized_keypd ~ /. Ssh/authorized_keys rac2 :~ /. Ssh/authorized_keys
In short, the information of the two nodes is written to the public key file, and then the password is no longer required for mutual ssh, because the RAC installation requires copying files between nodes
-- Verify Node 1 execution: ssh rac1 datessh rac2 datessh rac1-priv datessh rac2-priv date
Node 2 execution: ssh rac1 datessh rac2 datessh rac1-priv datessh rac2-priv date
Verify it several times until you do not need to enter the password. Note that do not miss s in authorized_keys. I tried it for a long time and did not succeed. Finally, I found that the public key file name was wrong, caused by a loss of s
4. configure a shared Disk
-- Add several disks and partition fdisk-lfdisk/dev/sdbfdisk/dev/sdcfdisk/dev/sddfdisk/dev/sde

If you need to restart to identify the partition information, you can use the partporbe command in linux, run the command to enable the kernel to identify the new partition without restarting. You can also run partprobe-s to view the partition.
[Root @ rac1 ~] # Fdisk/dev/sdc
Command (m for help): nCommand action e extended p primary partition (1-4) pPartition number (1-4): 1 First cylinder (1-12, default 1 ): using default value 1 Last cylinder or + size or + sizeM or + sizeK (1-12, default 12): Using default value 12
Command (m for help): wThe partition table has been altered!
Calling ioctl () to re-read partition table.
WARNING: Re-reading the partition table failed with error 16: Device or resource busy. the kernel still uses the old table. the new table will be used at the next reboot. syncing disks. [root @ rac1 ~] # Partprobe-s/dev/sda: msdos partitions 1 2/dev/sdb: msdos partitions 1/dev/sdc: msdos partitions 1/dev/sdd: msdos partitions 1/dev/sde: msdos partitions 1 [root @ rac2 install] #

-- Bind disks for systems 4.x and 5.x, the binding method is different.
4. x: # vi/etc/sysconfig/rawdevices
/Dev/raw/raw1/dev/sdc1/dev/raw/raw2/dev/sdd1/dev/raw/raw3/dev/sde1
# Vi/etc/udev/permissions. d/50-udev.permissions # raw devicesram *: root: disk: 0660 # raw/*: root: disk: 0660raw/*: oracle: dba: 0660
5. x: # vi/etc/udev/rules. d/60-raw.rulesACTION = "add", KERNEL = "/dev/sdb1 ", RUN + = "/bin/raw/dev/raw/raw1% N" ACTION = "add", ENV {MAJOR} = "8 ", ENV {MINOR} = "17", RUN + = "/bin/raw/dev/raw/raw1% M % m" ACTION = "add ", KERNEL = "/dev/sdc1", RUN + = "/bin/raw/dev/raw/raw2% N" ACTION = "add ", ENV {MAJOR} = "8", ENV {MINOR} = "33 ", RUN + = "/bin/raw/dev/raw/raw2% M % m" ACTION = "add", KERNEL = "/dev/sdd1 ", RUN + = "/bin/raw/dev/raw/raw3% N" ACTION = "add", ENV {MAJOR} = "8 ", ENV {MINOR} = "49", RUN + = "/bin/raw/dev/raw/raw3% M % m" ACTION = "add ", KERNEL = "/dev/sde1", RUN + = "/bin/raw/dev/raw/raw4% N" ACTION = "add ", ENV {MAJOR} = "8", ENV {MINOR} = "65 ", RUN + = "/bin/raw/dev/raw/raw4% M % m" ACTION = "add", KERNEL = "raw *", OWNER = "oracle ", GROUP = "oinstall", MODE = "660"
You only need to add the red part, and then restart udev. If the configuration is normal, the following content is displayed:
[Root @ rac1 ~] $ Ll/dev/raw
Total 0crw-rw ---- 1 oracle oinstall 162, 1 Jan 15 raw1crw-rw ---- 1 oracle oinstall 162, 2 Jan 15 raw2crw-rw ---- 1 oracle oinstall 162, 3 Jan 15 raw3crw-rw ---- 1 oracle oinstall 162, 4 Jan 15 raw4
5. Add Environment Variables
Node 1: [oracle @ rac1 ~] $ Vi. bash_profileexport ORACLE_SID = RAC1export ORACLE_HOME =/u01/oracle/10.2.0/db_1export ORA_CRS_HOME =/u01/oracle/10.2.0/crs_1export PATH = $ ORACLE_HOME/bin: $ ORA_CRS_HOME/bin: $ ORACLE_HOME/jdk/bin: $ PATH
Node 2: [oracle @ rac1 ~] $ Vi. bash_profileexport ORACLE_SID = RAC2export ORACLE_HOME =/u01/oracle/10.2.0/db_1export ORA_CRS_HOME =/u01/oracle/10.2.0/crs_1export PATH = $ ORACLE_HOME/bin: $ ORA_CRS_HOME/bin: $ ORACLE_HOME/jdk/bin: $ PATH
6. Install Clusterware
-- Check cd./clusterware/cluvfy before installation
./Cluvfy. sh-stage pre crsinst-n rac1, rac2-verbose
It will list a lot of content and check whether it meets the conditions for installing Clusterware. You need to install a compat_db package separately, the oracle-validated package does not install this package. In addition, it will report a few other compat packet detection failures, so ignore this because the version is incorrect. If there are other content without passed, you need to handle it, until all the above errors are passed.
-- Start to install cd./clusterware
./RunInstall-ignoreSysPreReqs (the parameter can be case-insensitive and the command is not allowed)
At the end of running the OUI, two scripts are required to be executed on two nodes, in the order of script 1: RAC1-> RAC2-> Script 2: RAC1-> there is no problem with the first three executions of RAC2. in step 2, run root on node 2. sh, the following error is reported:
[Root @ rac2 bin] #/u01/oracle/10.2.0/crs_1/root. sh
WARNING: directory '/u01' is not owned by rootChecking to see if Oracle CRS stack is already configured
Setting the permissions on OCR backup directorySetting up NS directoriesOracle Cluster Registry configuration upgraded successfullyWARNING: directory '/u01' is not owned by rootclscfg: EXISTING configuration version 3 detected. clscfg: version 3 is 10G Release 2. assigning default hostname rac1 for node 1. assigning default hostname rac2 for node 2. successfully accumulated necessary OCR keys. using ports: CSS = 49895 CRS = 49896 EVMC = 49898 and EVMR = 49897. node <nodenumber>: <nodename> <private interconnect name> No keys were written. supply-force parameter to override. -force is destructive and will destroy any previous clusterconfiguration. oracle Cluster Registry for cluster has already been initializedStartup will be queued to init within 90 seconds. adding daemons to inittabExpecting the CRS daemons to be up within 600 seconds. CSS is active on these nodes. rac1 rac2CSS is active on all nodes. waiting for the Oracle CRSD and EVMD to startOracle CRS stack installed and running under init (1 M) Running vipca (silent) for processing ing nodeapps/u01/crs_1/jdk/jre // bin/java: error while loading shared libraries: libpthread. so.0: cannot open shared object file: No such file or directory
Wait patiently for several minutes (90 s + 600 s) and an error will be reported, which is caused by the Oracle bug in 10.2.0.1, the solution is to modify the vipca and srvctl files full under $ ORA_CRS_HOME/bin, add unset LD_ASSUME_KERNEL at the end of the file, save and exit, and then run the root command again on node 2. sh
[Root @ rac2 bin] #/u01/oracle/10.2.0/crs_1/root. shWARNING: directory '/u01' is not owned by rootChecking to see if Oracle CRS stack is already configuredOracle CRS stack is already configured and will be running under init (1 M) [root @ rac2 bin] #. /crs_stat-tCRS-0202: No resources are registered.
At this time, because no vip is configured, no resources are registered. Run vipca on any node (provided that the vipca of this node has been modified). If the following error is reported:
[Oracle @ rac1 bin] $ vipcaError 0 (Native: listNetInterfaces: [3]) [Error 0 (Native: listNetInterfaces: [3])]
Configure the NIC: [oracle @ rac1 bin] $. /oifcfg iflisteth0 192.168.1.0eth1 10.0.0.0 [oracle @ rac1 bin] $. /oifcfg getif [oracle @ rac1 bin] $. /oifcfg setif-global eth0/192.168.1.0: public [oracle @ rac1 bin] $. /oifcfg setif-global eth1/10.10.10.1: cluster_interconnect [oracle @ rac1 bin] $. /oifcfg getifeth0 192.168.1.0 global publiceth1 10.10.10.1 global cluster_interconnect
Note that you must have the permission to open the graphic interface and run vipca with the root user, instead of the oracle user. Otherwise, the system reports that the permission is Insufficient [oracle @ rac1 bin] $ vipcaInsufficient privileges. Insufficient privileges.
Then, the OUI interface of the vip configuration assistant is displayed, and the vip is configured. After the vip node alias is entered, the vip IP address (process omitted) is automatically filled and then exited after running vipca, run crs_stat again and you will find that all resources have been registered to crs [root @ rac1 bin] #. /crs_stat-tName Type Target State Host ------------------------------------------------------------ ora. rac1.gsd application ONLINE rac1 ora. rac1.ons application ONLINE rac1 ora. rac1.vip application ONLINE rac1 ora. rac2.gsd application ONLINE rac2 ora. rac2.ons application ONLINE rac2 ora. rac2.vip application ONLINE rac2