好贴,谢谢

Q7: Out of Memory Errors Under Heavy Load  (Grid Control)
A7:
Under heavy load, especially when connecting to multiple target databases, The Oracle Management Service may encounter
out-of-memory errors and restart. When an Oracle Management Service restarts, there is no loss of state in the repository, but users logged in to that Management Service will be asked to login
again. If the restarts are frequent, do the following for each
installed Management Service:

a)  Stop the oms;
    $ORACLE_HOME/opmn/bin/opmnctl stopall

   (Note: it is important to use opmnctl stopall here rather than
    "emctl stop oms" since we need to completely stop the associated
    Application Server instance as well)

b)  Edit $ORACLE_HOME/opmn/conf/opmn.xml

Search for the OC4J_EM section, which is a line that looks like this:

<process-type id="OC4J_EM" module-id="OC4J">

Edit the "java-options" data tag under this process type. Before edit,
this line should look similar to this:

<data id="java-options" value="-server
-Djava.security.policy=/private/smptest/oms_rel9/j2ee/OC4J_EM/config/java2.policy
-Djava.awt.headless=true -Xmx512M -XX:MaxPermSize=64m
-DORACLE_HOME=/private/smptest/oms_rel9 "/>

Change the -Xmx option from 512M to 1024M, and the -XX:MaxPermSize option from
64M to 256M.Also add the -Xnoclassgc option.

After edits, the line should look like this:

<data id="java-options" value="-server
-Djava.security.policy=/private/smptest/oms_rel9/j2ee/OC4J_EM/config/java2.policy
-Djava.awt.headless=true -Xmx1024M -XX:MaxPermSize=256M -Xnoclassgc
-DORACLE_HOME=/private/smptest/oms_rel9 "/>

c) Restart the Oracle Management Service using "emctl start oms".

Q8: How to diagnose emd upload error

Note:  This is  a  frequently asked question in some forums.

Case:
I builded up a database with Dbcontrol inside itsself,
Everything worked fine.
then I revoked
utl_FILE, UTL_SMTP, UTL_TCP, UTP_HTTP, DBMS_RANDOM
from public.

Then I got the error when i did a emctl upload
EMD upload error: Failed to upload file B0000002.xml, ret = -2

I granted
utl_FILE
UTL_SMTP
UTL_TCP
UTP_HTTP
DBMS_RANDOM

to public but the error remains.

Any soutions?

Answer 1:

This is a bug (in grid control,db control)

The bug 4573965 based on bug 3638973 is not fixed in 10.1.0.4; it is
fixed in Grid Control - Release 2 (10.2 version).

Bug indicates as a workaround: removed some targets from targets.xml
(mainly database targets), agent successfully uploaded and then added
targets back via GC Console.

Upgrade to 10.2 when the release will be available

You can remove some targets (mainly database targets) by running the
following command:
emctl config agent deletetarget oracle_database <db_name>


Answer 2:
1. Log into the repository database as sys:
grant execute on sys.utl_smtp to sysman
grant execute on sys.utl_tcp to sysman

2. alter package sysman.emd_notification compile;

And some other schemas need grants as well -

grant execute on utl_file to xdb;
grant execute on sys.utl_file to DMSYS;
etc......

Then recompile invalid objects in db using utlrp.sql found here =

$ORACLE_HOME/bla/bla/RDBMS/ADMIN/utlrp.sql

Then refresh the agent

emctl clearstate agent
wait around 30 minutes for the state to clear

BTW: I'll create a  PDF file to contain some frequently asked questions.

很好，谢谢

现在好像很少看到楼主了，希望楼主能有更多的总结帖上来和大家分享

学习了！

今天更新几个Tip,基本都是我在测试过程中发现的.

Q9:. How to solve agent upload timeout issue ?
A9:
Agent upload timeout may related to many factors include:

1) The OMS is very busy on deal with specified agent's http request;

2) The OMS box is over workloaded and the Apache time out occurs, so it refuses agent http request from specified agent side;

3) The connection between OMS and repository db is in poor performance, e.g. using remote repository db sometimes will
trigger this issue.

4) The meta data xml files are corrupted or bad, so OMS refueses
to accept it.

From performance tuning point, there is one possible approach
can be considered to try:

    . Increase the Apache time out value;
      
      File 1 : $OMS_HOME/Apache/Apache/conf/httpd.conf
      parameter: Timeout (Default 300 seconds)
      
      File 2: $OMS_HOME/sysman/config/httpd_em.conf
      parameter: Timout   
     
      Then restart oms, clean agent and restart agent on target box.
      
   
    . Tune the connection issue between OMS box and repository box. If possible,
      using local repository db or make the repository db locates on the same lan
      as oms box.
      
   
    . Increase the timeout parameter in agent side, then restart agent.
     
      UploadTimeout=xxx (default is 1800)
      UploadMaxTime=xxx (default is 300)

Q10. How to Configure EM Grid Control to use the Single Sign-On Logon Page

A10: Note, This tip comes from Metalink.

To configure the Grid Control Console for use with Oracle Application Server Single Sign-On:

Steps:

1. Set the ORACLE_HOME environment variables to the Management Service home directory.

For example:

$PROMPT> setenv ORACLE_HOME /dev01/oracle/em10g_GridControl

2. Change directory to the bin directory of the Management Service Oracle home:

$PROMPT> cd $ORACLE_HOME/opmn/bin

3. Stop the Management Service, the Oracle HTTP Server, and the other components of the application server:

$PROMPT> ./opmnctl stopall

4. Change directory to the bin directory of the Management Service Oracle home:

$PROMPT> cd $ORACLE_HOME/bin

5. Enter the following command at the operating system prompt:

$PROMPT> ./emctl config oms sso -host ssoHost -port ssoPort -sid ssoSid  -pass ssoPassword -das [url]http://ssohostort/[/url]

For example:

$PROMPT> ./emctl config oms sso -host sshost1.acme.com -port 1521 -sid asdb -pass Ch22x5xt -das http://ssohost1.acme.com:7777


6. Restart the Management Service, Oracle HTTP Server, and the other application server components:

$PROMPT> cd $ORACLE_HOME/opmn/bin
$PROMPT> ./opmnctl startall

Go the Grid Control Console URL.

For example:
http://mgmthost1.acme.com:7777/em

The browser is redirected to the standard Single Sign-On Logon page.