- Posted by Gavin Soorma
- On October 21, 2010
- 1 Comments
Here are some notes I put together after performing a couple of installations of 11g Grid Control on a Solaris SPARC 64 bit platform. I also encountered some issues with the 11g agent and 9i and 10g databases. There is not too much user experiences available online, so I thought that I would share a few of my observations and feedback of the issues encountered and how I managed to overcome those issues.
Kindly provide feedback of your experience with the 11g Grid Control product for the good of all the user community.
For example: java -d64 -jar wls1032_generic.jar
$ /u01/app/oracle/product/10.2.0/bin $ which java
Check the java version:
$ cd /u01/app/oracle/product/10.2.0/bin
$ java -version
java version “1.6.0_18”
Java(TM) SE Runtime Environment (build 1.6.0_18-b07)
Since we did not have enough space on /tmp – we had to use the tmpdir option
java -d64 -Djava.io.tmpdir=/u01/oracle/stage -jar wls1032_generic.jar
If connection to the internet from the host is not possible, then read this Metalink note which describes how to do an offline download of the WLS patches
How to Apply WebLogic Server (WLS) Patches to an Offline System [ID 876004.1]
If the SYSMAN schema exists, it can be dropped using the following command
emca -deconfig dbcontrol db -repos drop -SYS_PWD [SYS password] -SYSMAN_PWD [SYSMAN password]
Undo tablespace must be at least 200M in size.
log_buffer needs to be minimum 10485760
session_cached_cursors should be 200 or higher
processes should be 500 or higher
open_cursors should be 300 or higher
Agent Home on OMS server
Web Tier home
Grid Control Instance Home
For example: /u01/app/oracle/product/gc_inst/em/EMGC_OMS1/sysman/log
Can also use this command to find out log file locations:
opmnctl is no longer used in 11g Grid Control
To stop and start we now use the following commands:
$OMS_HOME/bin>./emctl start oms
Which will start the OMS and also the WebLogic processes required for Grid Control
opmnctl commands in 11g can be used only to check the status or start the HTTP_Server component
$ export ORACLE_INSTANCE=/u01/app/oracle/product/gc_inst/WebTierIH1
$ cd /u01/app/oracle/product/Middleware/oms11g/opmn/bin
$ ./opmnctl status
Processes in Instance: instance1
ias-component | process-type | pid | status
ohs1 | OHS | 9939 | Alive
Stopping Grid Control
Issues Encountered and Fixed
10.2.0.4 listener is showing status Down in EM Grid Control even though it is actually running and status confirmed via lsnrctl status command.
The error message seen in Grid Control console is:
“The listener is down: Failed to tnsping mhost.com.au”
This is documented on Metalink as :
Bug 10070623: STATUS DOWN FOR LISTENER 10.2 MANAGED BY AGENT 11G
The fix for this was to edit the emctl executable which is located in the AGENT_HOME/bin directory.
Stop the agent via the emctl stop agent command
Take a backup of the emctl file and remove the following lines
if [ “$LD_LIBRARY_PATH_64” = “” ] ; then
Start the agent.
Check the status of the Oracle 10.2.0.4 listener. It should now show as up!
ORA-7445 errors messages in 184.108.40.206 database alert log and core dumps after starting the 11g Agent on a machine which has an Oracle 220.127.116.11 database Oracle Home
The alert log will show an error message like this :
Error : ORA-7445 [00000001001F5B04] [SIGBUS] [Invalid address alignment]
On the Grid Console when trying to configure the 18.104.22.168 database, we will see an error message like this :
“Exception in computing dynamic properties”
Apparently this was an issue with the 10.2.0.5 agent and a patch was issued for it.
Bug.8501576 9I DATABASES COREDUMPS IF 10.2.05. AGENT MONITORING
This is what is mentioned on Metalink….
If you cannot upgrade your target database to 22.214.171.124, there is also a fix on the Agent side:
• Apply Agent PSU2 Patch 9162498 (Unix) / Patch 9223116 (Windows) on top of the 10.2.0.5 Agent. Please check Note 986539.1 for more details about PSU2.
• To activate the fix, add the parameter AvoidOCIPing=1 in $AGENT_HOME/sysman/config/emd.properties and restart the Agent. This parameter is not there by default: it is a toggle to activate the bug fix for this particular issue. The default value is 0, which means that the fix is not activated
The workaround for this is to go to the AGENT_HOME/sysman/config directory and edit the emd.properties file.
Change these lines :
# Flag to Determine if OCIPing can be used. Older versions of databases such
# as 9206 do not support this.
#@description=Flag to Determine if OCIPing can be used.