Installing 11g Grid Control on Linux
I recently installed Oracle Grid Control 11g on my linux workstation. Below are the steps that I went through to get it working. Prior to installing Control, I had already installed Grid Infrastructure 188.8.131.52 and Oracle Database 184.108.40.206 and created a database to be used as the Grid Control repository.
Step 1) Download the Grid Control software:
- go to technet.oracle.com, click Download and select Enterprise Manager
- Select Enterprise Manager 11g Grid Control Release 1 (220.127.116.11.0)
- select Linux x86-64 and download all 3 files:
- unzip all 3 files into a staging location
Step 2) Download and Install Oracle Web Logic Server
- go to technet.oracle.com, click Download and select Fusion Middleware (incl. Weblogic Server)
- Under “Oracle Weblogic Server 10.3.6 and Previous Releases”, click “see all releases”
- select Oracle Weblogic Server 11gR1 (10.3.2) - Package Installer and download the file under Linux (32-bit JVM).
The filename is wls1032_linux32.bin and is a self-extracting executable so to launch the installation, just run ./wls1032_linux32.bin. This will bring up a GUI installation program. I accepted the defaults and installed into /opt/oracle/product/Middleware. I was prompted to specify 2 passwords during the install: weblogic and nodemanager.
Step 3) Install Grid Control
From the staging location where you unzipped the files, execute ./runInstaller to bring up the Grid Control Installation GUI. The first screen prompts you to enter your Oracle Support credentials so you can download and install the updates. If you have a valid Oracle Support contract, I suggest you take this option as it saves you the trouble of separately installing any recommended/mandatory patches. After selecting this option, the Installer connected to Oracle Support, downloaded several patches and proceeded with the main installation. During the installation, I ran into a few problems that required attention before it would complete:
- compat-db-4.1.25 was not installed so I installed it from my Linux software distribution CD
- Database Control had been configured in my repository database and had to be removed:
emca -deconfig dbcontrol db -repos drop -SYS_PWD <sys password> -SYSMAN_PWD <sysman password>
- Incorrect database settings:
UNDO tablespace >=200M
After fixing each of these issues, I proceeded with the installation and Grid Control 11g was successfully installed. I was prompted to specify 2 passwords during the installation: sysman and the agent registration password
Step 4) Install Grid Control Agent on each node that you want to manage
Before Grid Control can manage a database, you must first install the Grid Control agent on the database server. The agent installation software installs with Grid Control so you can copy it from there to each database server and then install it. The installers are platform specific, so in my case, I copied $GC_HOME/sysman/agent_download/18.104.22.168.0/linux_x64/agentDownload.linux_x64 to a staging location on each of my RAC nodes. I then installed the agent with the following command, specifying the installation directory as well as the names of my cluster and nodes. The “-y” at the end will bypass the software updates portion of the installation.
./agentDownload.linux_x64 -b /u01/app/11.2.0/agent11g -n mdlinux-cluster -c "mdlinux-vm1,mdlinux-vm2" -y
Because I had created a registration password during the installation of Grid Control, I also need to secure the agent:
emctl secure agent emctl start agent emctl upload agent
Step 5) Discover the nodes in Grid Control
After you configure the Grid Control agent and it uploads its data, the hosts should become visible in Grid Control and you can then add the databases.