Background

The easiest and most reliable means of extending your OEM management server cluster is through the Deployment Procedure Library under the Enterprise menu.  You can tailor it to your needs and it runs reliably every time.

What if you want to connect to the same repository from somewhere else, like your DR site?  When you use the Deployment Procedure Library any new OMS is added as a slave to the existing OMS server.  All Weblogic (WLS) tasks for members of the cluster have to be run through the WLS Admin Console on your Master Server.  Your DR OMS is slave to the primary site so — if the primary site goes down you have no convenient means for managing your DR weblogic server.  Problem.

Solution

You can manually add an OMS server connected to the same repository as your primary sites with a little elbow grease.  That’s what this posting is for.

Procedure

  1. Manually install the OMS binaries to match your primary site.  Directory structures on the server must match exactly.
  2. Backup the configuration of your primary OMS using omsca commands
  3. Copy the configuration file from the primary OMS to the new server
  4. Recover the new server using that backup

That seems simple enough.

What Could Possibly Go Wrong

Your existing OMS has been busy since you first installed it.  You’ve discovered hosts and targets like databases.  With each discovery you’ve added plugins to your OMS configuration and they’ve put related sysman objects into your repository.

That create a gap between the software in your installation media and your servers that you’ll need to fill.  We’ll discuss that when it bites us.

Procedure

  • Install the new OMS server from the same installation media as your current OMS server.  Use identical settings for everything (MIDDLEWARE_HOME, administrator accounts, ports etc.).  The only entry that should differ is the host name.
  • Apply the same patches and patchsets as your primary OMS.  PSU’s one-off patches, etc. must match.
  • Backup the configuration file from your existing OMS using omsca, then copy the backup file to your new server.
  • Attempt to recover the new OMS using that same omsca backup file.  It will throw errors.  Use the information about the names of the plugins it requires for this environment to update your plugins.
  • Download and unzip the plugin set matching your primary OMS from MOS. For Linux the download page is  http://www.oracle.com/technetwork/oem/grid-control/downloads/linuxx8664soft-085949.html
  • After unzipping it on your new OMS find the fully-qualified path to that new plugins directory
  • From $OMS_HOME/sysman/install, execute PluginInstall.sh -pluginLocation <that unzipped plugin directory>
  • Try to run the omsca recovery again.  Better?
  • Start the OMS using emctl start oms

Examples

Problems During First Run

> omsca recover -ms -backup_file /oraoms/exportconfig/blahblahblah/opf_ADMIN_20130729_073137.bka

Oracle Enterprise Manager Cloud Control 12c Release 12.1.0.2.0
Copyright (c) 1996, 2012, Oracle. All rights reserved.
OS check passed.
OMS version check passed.
Populated install params from backup...
Enter Admin Server user password:
Enter location for OMS config files[/orabase/gc_inst]:
Enter Repository database user password:
Enter Agent Registration password:
Confirm Password:
Doing pre requisite checks ......
Pre requisite checks completed successfully
Checking Plugin software bits
The following plug-ins are missing from the software installation.Please install these plug-ins before attempting recovery. Missing plug-ins will cause a failure in recovery process.
Missing Plugin bits for Plugin id: oracle.sysman.mos with version: 12.1.0.4.0 and revision: 0
Missing Plugin bits for Plugin id: oracle.sysman.db with version: 12.1.0.3.0 and revision: 0

Success After Plugins are Installed

> omsca recover -ms -backup_file /oraoms/exportconfig/blahblahblah/opf_ADMIN_20130729_073137.bka
Oracle Enterprise Manager Cloud Control 12c Release 12.1.0.2.0
Copyright (c) 1996, 2012, Oracle. All rights reserved.
OS check passed.
OMS version check passed.
Populated install params from backup...
Enter Admin Server user password:
Enter location for OMS config files[/orabase/gc_inst]:
Enter Repository database user password:
Enter Agent Registration password:
Confirm Password:
Doing pre requisite checks ......
Pre requisite checks completed successfully
Checking Plugin software bits
Proceed to recovery
A new managed server EMGC_OMS2 will be created.
Now launching AddOMS...
Infrastructure setup of EM completed successfully.
Doing pre deployment operations ......
Pre deployment of EM completed successfully.
Deploying EM ......
Deployment of EM completed successfully.
Configuring webtier ......
Configuring webTier completed successfully.
Importing OMS configuration from recovery file...
If you have software library configured 
please make sure it is functional and accessible 
from this OMS by visiting:
 Setup->Provisioning and Patching->Software Library
Importing configuration for pluggable modules...
Securing OMS ......
Post "Deploy and Repos Setup" operations completed successfully.
Performing Post deploy operations ....
Total 0 errors, 77 warnings. 0 entities imported.
OMSCA-WRN:Few post deploy operation failed. Check the file:/orabase/Middleware/oms/cfgtoollogs/omsca/logs20130729142445/omsca_err.txt
Post deploy operations completed successfully.
EM configuration completed successfully.
EM URL is:https://blahblahblah:7802/em
Add OMS finished successfully
Creation and setup of new server EMGC_OMS2 completed successfully
OMSCA Recover completed successfully

> emctl start oms
Oracle Enterprise Manager Cloud Control 12c Release 2  
Copyright (c) 1996, 2012 Oracle Corporation.  All rights reserved.
Starting Oracle Management Server...
Starting WebTier...
WebTier Successfully Started
Oracle Management Server Already Started
Oracle Management Server is Up

Reference: E24089-23 OEM Cloud Control Advanced Installation and Configuration Guide 12c Release 3; Section 5 Installing Additional OOracle Management Server in Silent Mode