Add the 2 remaining node to the system
What we have got now?
- Grid infrastructure
- RAC database software
- 2 node Oracle EBS system
It's time to add the 2 remaining node using the shared APPL_TOP method.
I will use the 384248.1 note during the implementation.
The shared APPL_TOP software already there, all application tier node could reach through the shared partition, through the /u01 mount point.
Implementation steps
1. Stop the application tier on con01 server. Wait for the normal shutdown, wait until any application tier process is running on the server.
2. Run the preclone script on conc01 server.
3. Now log in to the first web tier node (app01) with applmgr user and initialize the session
# su - applmgr
# export PERL5LIB=/u01/apps/PROD/apps/tech_st/10.1.3/perl/lib/5.8.3
4. Just to be sure check the applmgr and the oracle unix users are exactly same as on conc01 server. (use id unix command)
5. Check that the /u01 and /u02 mount points are working and contain the same files as on conc01 server
6. Check local oraInst.loc file.
I am a little bit confused with this step. The shared appl top note is not uniform in this case.
Where oraInventory is part of the shared mount point where it is not.
So choose one of the following:
Where oraInventory is part of the shared mount point where it is not.
So choose one of the following:
- Make a same oraInst.loc file with the same content as on conc01 server.
- Make an oraInst.loc file which is point to a different oraInventory directory as on conc01 server.
After autoconfig step please check the oraInventory that is it contain the new EBS Oracle homes or not.
7. Create the new context file (Section 4, step 3 from the note)
# cd /u01/apps/PROD/apps/apps_st/comn/clone/bin
# perl adclonectx.pl addnode contextfile=/u02/apps/PROD/inst/apps/PROD_conc01/appl/admin/PROD_conc01.xml
parameters:
Target System Hostname: app01
Do you want the input to be validated: y
Target System Root Service: enabled
Target System Web Entry Point Services: enabled
Target System Application Services: enabled
Target System Batch Processing Services: disabled
Target System Other Services: disabled
Do you want to preserve the Display: n
Target System Display: app01:1.0
Do you want the target system to have the same port values as the source system: y
Choose the values which will be set as APPLTMP values on the target node: /usr/tmp
Be aware, you have to set to disabled the Batch Processing Services! This node will be only a web tier, it will not run concurrent programs.
After successful run check the log file, and the newly created PROD_app01.xml context file.
8. Run the adconfig
# cd /u01/apps/PROD/apps/apps_st/appl/ad/12.0.0
# perl $PWD/bin/adconfig.pl contextfile=/u02/apps/PROD/inst/apps/PROD_app01/appl/admin/PROD_app01.xml
9. run autoconfig on conc01
ssh applmgr@conc01
# . bin/PROD.env
# cd $ADMIN_SCRIPTS_HOME
# ./adautocfg.sh
Now I have a 1+2 node system. But before I continue I had to disable web services on the conc01 server.
It isn't a big problem if you don't disable them, by the way in this configuration the web services will not used. The load balancer will reach only the two application tier node's web services.
10. set to disabled value for s_web_entry_status and the s_web_applications_status parameters in the conc01's context file. Use the same previous ssh session.
11. run autoconfig again on the conc01 server
# cd $ADMIN_SCRIPTS_HOME
# ./adautocfg.sh
As a post implement step it is recommended to create start/stop scripts and an enviroment file (PROD.env) in applmgr's $HOME/bin directory on app01 server. I have created them.
Add the third application tier node
The shared APPL_TOP software already there, all application tier node could reach through the shared partition, through the /u01 mount point.
Implementation steps
1. Stop the application tier on con01 and on app01 server - only if you have started. Wait for the normal shutdown, wait until any application tier process is running on the servers.
2. Run the pre clone script on app01 server.
3. Now login into the second web tier node (app02) with applmgr user and initialize the session
# su - applmgr
# export PERL5LIB=/u01/apps/PROD/apps/tech_st/10.1.3/perl/lib/5.8.3
4. Just to be sure check the applmgr and the oracle unix users are exactly same as on conc01 and on app01 server.
5. Check that the /u01 and /u02 mount points are working and contain the same files as on conc01 and on app01 server
6. Check local oraInst.loc file
As I wrote in the second node I am confused with this step. Choose again:
- Make a same oraInst.loc file with the same content as on conc01 and on erapp01 server. In this case there will be a common oraInventory for the 3 application tier server
- Make an oraInst.loc file which is point to a different oraInventory directory than on conc01 and on erapp01 server.
After autoconfig step please check the oraInventory that is it contain the new EBS Oracle homes or not.
7. Create the new context file (Section 4, step 3 from the note)
# cd /u01/apps/PROD/apps/apps_st/comn/clone/bin
# perl adclonectx.pl addnode contextfile=/u02/apps/PROD/inst/apps/PROD_app01/appl/admin/PROD_app01.xml
parameters:
Target System Hostname: app02
Do you want the input to be validated: y
Target System Root Service: enabled
Target System Web Entry Point Services: enabled
Target System Application Services: enabled
Target System Batch Processing Services: disabled
Target System Other Services: disabled
Do you want to preserve the Display: n
Target System Display: app02:1.0
Do you want the target system to have the same port values as the source system: y
Choose the values which will be set as APPLTMP values on the target node: /usr/tmp
Be aware, you have to set to disabled the Batch Processing Services! This node will be only a web tier, it will not run concurrent programs.
After successful run check the log file, and the newly created PROD_app02.xml context file.
8. Run the adconfig
# cd /u01/apps/PROD/apps/apps_st/appl/ad/12.0.0
# perl $PWD/bin/adconfig.pl contextfile=/u02/apps/PROD/inst/apps/PROD_app02/appl/admin/PROD_app02.xml
9. run autoconfig on conc01, app01 and app02 in this order.
Now I have a 1+3 node system.
If you start the full system now then you could login into the system used by that's server URL where the autoconfig was latest ran.
It is hardly recommend to stop the full EBS system now - if you have started it - and take a safety backup from all Oracle EBS software on all nodes (1 db software and 1 shared app_top and all inst_top directory)
No comments:
Post a Comment