So here we are finally to install the database software over our recently installed 188.8.131.52 Grid Software.
Go to the database directory where we have unzipped the software and run the installer. First screen will come up to start up the process:
As we are planning to install RAC Database immediately with the software installation so we have to select the 2nd option.
For now, we are just going ahead with “Admin Managed”. I am yet to cover Policy Managed Database in near future.
Both the nodes will be checked automatically, however sometimes i have seen we may have to do it manually.
So we will go by Advanced Install, so that we can choose the options from the installer.
Make sure these directories are existing and select them, if ORACLE_BASE and ORACLE_HOME is configured before calling runInstaller, these values will be picked up automatically.
Here we go – we have to use Container Database Name along with the Pluggable Database Name, we can have this installation with or without Pluggabale, however its good to use features provided by Oracle 12c.
Allocate the memory based on your Shared Memory (/dev/shm), else it may throw an error.
Change the Character Sets as AL32UTF8.
Select the option Oracle ASM to have our Database Files along with other important Files.
No cloud control is configured yet so just unchecked it.
Flash Recovery Area can be configured any time later, so for now i just unchecked it.
Below one is the only Warning I got, which was like easy to be configured.
Configure the values in /etc/security/limits.conf and “Check Again”. This time it will through.
So all set, check the configuration once again before we hit the button “Install”.
Execute the Scripts on each node, make sure you execute the script on master node 1st (Node 1 in this case) followed by the other nodes.
On Node 1:
on Node 2:
Let the process move on…
So finally we got something on the screen which I did not like though, however we can fix this small problem later on, it just happened because of Memory Target would have not satisfied the Oracle Installer. Press “OK” and move on.
Here we go, RAC Database is installed with the configuration details below:.
Close it and we are done with the installation.
Check the process and we found out that, Database is not up on node 2 because of the error we got above.
Fixing the Error which we got on Node 2 due to Memory Target issue:
So for now 3G is configured, which is ideally same as node 1, however still its not enough on node 2, so i planned to increase to have the complete SGA in it.
Simple command to do that though -> mount -o remount,size=4G /dev/shm. I made the same configuration on Node 1 too to keep the configuration same.
After the configuration, stop and start the cluster database and it will be up without any problem this time.
After we stopped it, check the status:
Now finally start the database and it will be through this time without any problem:
Cool, we are good with the Oracle RAC Database Installation.
Last thing is to configure this /dev/shm increased value in /etc/fstab file and restart the system to make it permanent. This time all will go well.
Note: In OEL 6.3 there is bug that even though you change the below value in /etc/fstab, after restart the value wil be changed to half of physical RAM.
So to fix this issue, make sure you add – mount -o remount /dev/shm in /etc.rc.local file.
Now even after restart the value will remain 4g, it will not be overwritten due to Bug.
Any questions folks, let me know.