| Oracle® Collaboration Suite Upgrade Guide 10g Release 1 (10.1.2) for hp-ux PA-RISC (64-bit), Linux x86 and Solaris Operating Environment (SPARC) Part Number B25495-07 |
|
|
View PDF |
| Oracle® Collaboration Suite Upgrade Guide 10g Release 1 (10.1.2) for hp-ux PA-RISC (64-bit), Linux x86 and Solaris Operating Environment (SPARC) Part Number B25495-07 |
|
|
View PDF |
This appendix describes common problems that you might encounter when upgrading to Oracle Collaboration Suite 10g Release 1 (10.1.2) and explains how to solve them. It contains the following topics:
This section contains the following topics:
Section A.1.1, "Upgrading the Database Version of the OracleAS Metadata Repository Database"
Section A.1.2, "Problems Using Middle Tier with Upgraded Infrastructure"
Section A.1.4, "Problems Encountered While Upgrading Oracle Internet Directory"
Section A.1.6, "Configuration Assistant Failure During Oracle9iAS Infrastructure Upgrade"
Section A.1.8, "Database Upgrade Assistant Failure During Oracle9iAS Infrastructure Upgrade"
Section A.1.9, "Unable to Create Realms in 10.1.2 Post Upgrade Because of Insufficient Rights"
Section A.1.10, "Database Listener Fails to Start while Upgrading the Infrastructure Tier"
The Oracle Application Server Metadata Repository requires an Oracle database. Before you upgrade to Oracle Collaboration Suite 10g Release 1 (10.1.2), your OracleAS Metadata Repository database must be the database versions described in Section 8.1.
Problem
When a newer version of Oracle Database is announced, should I upgrade the OracleAS Metadata Repository database to the new database version?
Solution
In general, use caution when upgrading your OracleAS Metadata Repository database to a new database version. Check Oracle Metalink (http://metalink.oracle.com) for posted articles or announcements that confirm that the database version and upgrade has been tested and is supported for an existing OracleAS Metadata Repository database.
Make sure you have followed the documented guidelines for upgrading your installations in the proper order.
Problem
After upgrading an Oracle9iAS Infrastructure, all the middle tiers generate errors when attempting to connect to the upgraded Infrastructure.
Solution
This problem occurs when you upgrade the Oracle9iAS Metadata Repository before you upgrade the middle tiers that depend on the repository. To fix the problem, you must do one of the following:
Upgrade the middle tiers so they are the same version as the OracleAS Metadata Repository.
Revert the OracleAS Metadata Repository to its original version by restoring a backup of the Oracle9iAS Metadata Repository; then, upgrade the middle tiers before upgrading the Oracle9iAS Metadata Repository.
You upgrade Oracle9iAS Infrastructure using the Oracle Collaboration Suite 10g Release 1 (10.1.2) installation procedure, which is performed using the Oracle Universal Installer.
Problem
When you attempt to upgrade Oracle9iAS Infrastructure on a UNIX system, the upgrade fails. An "insufficient privileges" error appears in the following log file:
SOURCE_ORACLE_HOME/assistants/dbma/logs/trace.log
Specifically, the error appears as follows:
oracle.sysman.assistants.util.sqlEngine.SQLFatalErrorException: ORA-01031: insufficient privileges
Solution
Before you start Oracle Universal Installer to begin the installation procedure, be sure to log in as a user that is a member of the DBA group for the database.
The Oracle Internet Directory upgrade assistant is one of the assistants that run near the end of the 10g Release 1 (10.1.2) installation procedure when you are upgrading an Oracle9iAS Infrastructure installation.
You can get information about the cause of Oracle Internet Directory upgrade assistant errors by looking at the following log file:
ORACLE_HOME/ldap/log/oidca.log
Problem 1
The upgrade assistant log file (oidca.log) reports the following:
OID processes are currently running
Solution
This is a result of some Oracle Internet Directory or Oracle Directory Integration and Provisioning processes not being shut down properly in the source Oracle Home.
Shut down the processes in the source Oracle Home before retrying the Oracle Internet Directory configuration assistant from the Oracle Universal Installer configuration assistants page.
|
See Also: Section 7.2 for instructions on stopping an Oracle9iAS Infrastructure or the Oracle Internet Directory documentation to stop the Oracle Internet Directory and Oracle Directory Integration and Provisioning processes in the source Oracle Home. |
Problem 2
The Oracle Internet Directory configuration assistant fails during the Configuration Assistants phase of the Oracle9iAS Infrastructure upgrade with Oracle Universal Installer.
Solution
Check the contents of the following configuration file in the destination Oracle home and verify that the file contains the correct SERVICE_NAME entry for your Oracle9iAS Metadata Repository. If the value assigned to this entry is incorrect, enter the correct name, save the file, and retry the assistant.
DESTINATION_ORACLE_HOME/network/admin/tnsnames.ora
Problem
When you run the Oracle Universal Installer to upgrade Oracle9iAS Infrastructure, a popup dialog notifies you that the installer will shut down some processes in the source Oracle home.
After the installer performs the shutdown, it checks that Oracle Internet Directory is stopped. If Oracle Internet Directory is not stopped for some reason, the installer displays another dialog notifying you of the problem.
Solution
Examine the following log file to determine the cause of the problem:
DESTINATION_ORACLE_HOME/cfgtoollogs/shutdownprocesses.log
Resolve the problem and then manually stop Oracle Internet Directory in the source Oracle home. Once Oracle Internet Directory is stopped, continue with the Oracle9iAS Infrastructure upgrade by clicking Continue in Oracle Universal Installer.
See Section 7.2 for instructions on stopping Oracle Internet Directory.
Problem
Oracle Universal Installer invokes configuration assistants at the end of the Oracle9iAS Infrastructure upgrade. Some of the configuration assistants require an Oracle Database 10g database listener to connect to the database. If an Oracle Database 10g database listener is not available, those configuration assistants fail.
Solution
The installer normally starts an Oracle Database 10g database listener in the destination Oracle home. However, if an Oracle9i (9.0.1.3) database listener is already running, then the installer fails to start the Oracle Database 10g (10.1.0.2) database listener.
The most common cause of this problem is that you missed the instruction in a pop-up dialog during the installation. This message indicates during the interview phase of the installation that there is a running database listener running and that you should stop the listener manually before proceeding.
To correct the problem, stop the existing Oracle9i (9.0.1.3) listener, and then start the database listener in the destination Oracle home, as follows:
Set the ORACLE_HOME environment variable to point to the destination Oracle home of upgrade.
Change directory to bin directory of the destination Oracle home.
After the Oracle Database 10g database listener is running, continue with the OracleAS Identity Management upgrade by clicking Retry on the Configuration Assistants page in Oracle Universal Installer.
Problem
Oracle Universal Installer starts Database Upgrade Assistant at the end of Oracle9iAS Infrastructure upgrade. Database Upgrade Assistant may take a long time depending on the size and contents of the database. The installer shows progress of the Database Upgrade Assistant by displaying percentage numbers, but no details about the progress are shown on the Configuration Assistants screen in Oracle Universal Installer.
Solution
If you would like to obtain more detailed information about the progress of the Database Upgrade Assistant, examine the log files generated by the Database Upgrade Assistant. The log files reside in:
DESTINATION_ORACLE_HOME/admin/SID/upgrade
In this example, replace SID with the system identifier of the database in the source Oracle home.
To obtain the timestamps of the different stages of the database upgrade, search for the string "COMP_TIME" in the log files. For example,
grep ^COMP_TIME *.log
The output of the command identifies each stage of the database upgrade, as well as a timestamp for each stage. For example:
Oracle_Server.log:COMP_TIMESTAMP DBUPG__BGN 2004-12-16 10:11:00 2453356 36660 Oracle_Server.log:COMP_TIMESTAMP UTLIP__END 2004-12-16 10:12:58 2453356 36778 Oracle_Server.log:COMP_TIMESTAMP CATALG_BGN 2004-12-16 10:27:44 2453356 37664 Oracle_Server.log:COMP_TIMESTAMP CATPROC 2004-12-16 11:18:45 Oracle_Server.log:COMP_TIMESTAMP RDBMS 2004-12-16 11:21:50 Oracle_Server.log:COMP_TIMESTAMP JAVAVM 2004-12-16 12:27:24 Oracle_Server.log:COMP_TIMESTAMP XML 2004-12-16 12:41:17 Oracle_Server.log:COMP_TIMESTAMP CATJAVA 2004-12-16 12:45:03 Oracle_Server.log:COMP_TIMESTAMP CONTEXT 2004-12-16 12:49:17 Oracle_Server.log:COMP_TIMESTAMP XDB 2004-12-16 12:56:32 Oracle_Server.log:COMP_TIMESTAMP OWM 2004-12-16 13:01:14 Oracle_Server.log:COMP_TIMESTAMP AMD 2004-12-16 13:11:04 Oracle_Server.log:COMP_TIMESTAMP ORDIM 2004-12-16 13:43:34 Oracle_Server.log:COMP_TIMESTAMP SDO 2004-12-16 13:52:30 Oracle_Server.log:COMP_TIMESTAMP WK 2004-12-16 13:56:24 Oracle_Server.log:COMP_TIMESTAMP DBUPG_END 2004-12-16 14:10:39 PostUpgrade.log:COMP_TIMESTAMP UTLRP_BGN 2004-12-16 14:12:32 PostUpgrade.log:COMP_TIMESTAMP UTLRP_END 2004-12-16 15:29:47
Oracle Universal Installer invokes Database Upgrade Assistant at the end of the Oracle9iAS Infrastructure upgrade. If the Database Upgrade Assistant fails, you can examine the log files generated by the Database Upgrade Assistant. The log files reside in:
DESTINATION_ORACLE_HOME/admin/SID/upgrade/
In this example, replace SID with the system identifier of the database in the source Oracle home.
Examine the log files and determine the cause of the failure. In most cases, it is not possible to retry the Database Upgrade Assistant. Instead, you will need to restore the source Oracle home and the database files to their state before the Oracle9iAS Infrastructure upgrade. After the restoration, make sure that the problems which caused the Database Upgrade Assistant to fail are resolved. Then start the Oracle9iAS Infrastructure upgrade again.
Following upgrade, attempting to create realms in Oracle Internet Directory using the Self Service Console results in an error.
Problem
This problem occurs if Oracle Internet Directory was upgraded from Release 2 (9.0.4.2), based on Oracle Application Server Release 2 (9.0.2) to 10g (10.1.2), based on Oracle Application Server 10g Release 2 (10.1.2) and multi-realms were not set up prior to upgrade in Oracle Application Server Release 2 (9.0.2). To resolve this issue, perform the following:
After upgrading, trying to create realms in Oracle Internet Directory using the Self Service Console results in the following error:
Cannot Create Identity Management Realm : Invalid Root Oracle Context. [LDAP: error code 50 - Insufficient Access Rights ]
Solution
To avoid this problem, do the following workaround after the Infrastructure upgrade or after all the tiers have been upgraded.
Start Oracle Directory Manager:
$ORACLE_HOME/bin/oidadmin
Log in to Oracle Directory Manager.
In the System Objects frame:
Expand Entry Management.
Navigate to the appropriate realm parent, such as dc=com | dc=visioncorp.
Select the Subtree Access tab.
Grant access rights to cn=ASPAdmin on the realm parent entry as follows:
Under Entry Level Options click Create.
Select the By Whom tab.
Select A Specific Group.
Enter the following:
cn=ASPAdmins,cn=groups,cn=visioncorpContext,dc=us,dc=visioncorp,dc=com
Select the Access Rights tab.
Grant Browse, Add and Delete access.
Grant access rights to cn=ASPAdmin on the realm parent attribute as follows:
Under Attribute Level Options click Create.
Select the By Whom tab.
Select A Specific Group.
Enter the following:
cn=ASPAdmins,cn=groups,cn=visioncorpContext,dc=us,dc=visioncorp,dc=com
Select the Attribute tab.
Make sure EQ and * are entered in the 2 fields.
Select the Access Rights tab.
Grant Read, Search, Write and Compare access rights.
Click Apply.
Expand Entry Management and navigate to cn=visioncorpContext > cn=Groups > cn=ASPAdmins.
In the Properties tab, make sure that cn=orcladmin exists in the uniquemember field.
Click Apply (if necessary).
Problem
While upgrading the Infrastructure tier, the database listener or CSS service fails to start and one of the following error messages displays:
Failed to start service, error 997. TNS-12560: TNS:protocol adapter error TNS-00530: Protocol adapter error
Or:
OS error in starting service OracleCSService
Solution
Change the user associated with the listener or CSS service from local system to the specific user name with which Oracle Collaboration Suite was installed.
Perform the following steps:
Run services.msc by selecting Start > Run and typing services.msc.
Right-click the Listener or CSS service (depending on which error message you received) and select Properties.
Click the Log On tab and select This account instead of Local system account (for example, domain\username).
Browse for the username under which Oracle Collaboration Suite was installed and enter the correct password in the spaces provided.
Click OK to save your changes and close the window.
This section contains the following topics:
Section A.2.2, "Short Global Database Name Prevents Upgrade of Information Storage Database"
Section A.2.3, "CommunicationException Thrown during Information Storage Database Upgrade"
Section A.2.6, "Information Storage Database Patch Throws Error ORA-01017"
Section A.2.8, "Configuring Oracle Collaboration Suite Search with an Enabled Database"
Problem
The recommended characterset for the Oracle Collaboration Suite 10g (10.1.2) database is AL32UTF8. If the characterset of your upgraded Information Storage Database is UTF8, foreign language e-mail headers may be broken after upgrading Oracle Collaboration Suite.
Solution
Perform the following steps:
In a text editor, open the Oracle Mail configuration file:
$ORACLE_HOME/opmn/conf/opmn.xml
Scroll down to the following line:
<variable id="NLS_LANG" value="AMERICAN_AMERICA.AL32UTF8" append="false"/>
Replace AL32UTF8 with UTF8.
Execute the following commands:
$ORACLE_HOME/opmn/bin/opmnctl validate $ORACLE_HOME/dcm/bin/dcmctl updateconfig $ORACLE_HOME/bin/oesctl sync_opmnn_conf
Restart the Oracle Mail server.
Problem
When running the Oracle Universal Installer to upgrade Oracle Collaboration Suite information storage database that was installed with a short global database name (empty database domain name), a dialog box displays the following error message:
"The Database selected for upgrade is not registered in the Oracle Internet Directory specified."
Solution
The error is a result of the global_name parameter having been appended with a default database domain name during the installation of the Oracle Collaboration Suite 9.0.4 Information Storage, when a short database name is given. To solve this problem, perform the following steps:
Verify that the database domain value is empty on your database by executing the following SQL command:
> show parameter db_domain;
Find the short global database name for your database by running the following SQL command:
> show parameter db_name;
Change the global_name parameter so that its value matches that of the short global database name for your database. For example, if your short global database name is STORE, change global_name to match this value. To change global_name, execute the following SQL command:
> update props$ set value$='STORE' where name like 'GLOBAL_DB_NAME';
After acknowledging the dialog box, the Oracle Universal Installer continues the upgrade process.
While the Oracle Collaboration Suite Schema Creation Assistant is running, between the messages "Modifying Schema for UM" and "Loading the mailstore object," the following error message appears:
javax.naming.CommunicationException: Request: 150 cancelled
Problem
This error occurs if the Oracle Internet Directory server shuts down while the Oracle Voicemail & Fax LDAP schema is being loaded from the Oracle Collaboration Suite information storage database.
Solution
Perform the following steps:
Copy the oidSchemaUpgradeFrom90400UM.sbs file from the upgraded Oracle Collaboration Suite 10g Database $ORACLE_HOME\um\install directory to the upgraded Oracle Collaboration Suite 10g Infrastructure $ORACLE_HOME\bin directory using the ftp utility.
Use the ldapmodify tool to upload the schema file using the "ignore errors" option. Enter the following command:
ORACLE_HOME/bin/ldapmodify -h host -p port -D userdn -w password -f oidSchemaUpgradeFrom90400UM.sbs -c
where
ORACLE_HOME is the Infrastructure destination Oracle home.
host is the name of the host where Oracle Internet Directory is running.
port is the port where Oracle Internet Directory is running.
userdn is the DN of the Oracle Internet Directory administrator user.
password is the password of the Oracle Internet Directory administrator user.
Start the Oracle Internet Directory Manager (oidadmin). Locate the Entry Management tree item and edit the attribute orclproductversion at cn=UM,cn=oracleschemaversion. Change the attribute's value to 10.1.1.0.2. Retry the upgrade configuration assistant.
The Oracle Universal Installer displays the following error message during the process of enabling a customer Oracle Database 10g as an Oracle Collaboration Suite Database:
The Database has the following schema name conflicts. Ensure that this isfixed or select another database to be enabled as Oracle Colloboration SuiteDatabase before continuing. -ESMAIL
Problem
The enabling process is attempting to create a schema named es_mail, but an empty schema with the same name exists in the Oracle Collaboration Suite Database.
Solution
Verify that the existing es_mail schema is empty and then drop it from the database.
While upgrading the Information Storage Database, an error displays after choosing to use the same password for all accounts.
Problem
This problem occurs when upgrading the Information Storage from Release 2 (9.0.4.2) to 10g (10.1.2). If email is not configured in Release 2 (9.0.4.2), the following error is thrown on choosing to use the same password for all user accounts on the "Specify Database Schema Passwords" screen:
Error Validating ES_MAIL password
The ES_MAIL schema exists on the Release 2 (9.0.4.2) Information Storage Database but was not used by Oracle Mail.
Solution
To work around this problem:
Access the Release 2 (9.0.4.2) Information Storage Database and manually change the password for the default user ES_MAIL. The default password is es, so you can change the password to anything else.
After changing the password, upgrade the Release 2 (9.0.4.2) Information Storage Database to Information Storage Database 10.1.2.
When upgrading the Information Storage Database, an error is thrown.
Problem
This problem occurs when upgrading the Information Storage from Release 2 (9.0.4.2) to 10g (10.1.2). The error is as follows:
Fatal error:
ORA-01017: invalid username/password; logon denied
Infostore patch failed:
oracle.sysman.assistants.util.sqlEngine.SQLFatalErrorException: ORA-01017:
invalid username/password; logon denied.
at oracle.sysman.assistants.util.sqlEngine.SQLEngine.executeImpl(SQLEngine.java:1436)
at oracle.sysman.assistants.util.sqlEngine.SQLEngine.executeSql(SQLEngine.java:1677)
at oracle.mail.install.infostore.InfoStorePatch.doSQLPatch(InfoStorePatch.java:114)
at oracle.mail.install.infostore.InfoStorePatch.main(InfoStorePatch.java:47)
Error detected while applying OCS mail infostore patch
Solution
Use the following workaround before running the backend Configuration Assistant, or when you encounter the error. Change the password for the es_mail user to the default value of "es" both in the database and in the Oracle Internet Directory.
To change the password in the database:
Login as a user with sysdba privileges.
Issue the following command:
alter user es_mail identified by es /
To change the password in Oracle Internet Directory:
In the Oracle Internet Directory Manager, navigate to:
cn=OracleContext,cn=Products,cn=EmailServerContainer,cn=UM_SYSTEM,cn=Mailstores,cn=<service name>
Change the orclpassword attribute to "es".
When upgrading a Release 2 (9.0.4.2) environment with multiple databases and listeners on the same computer, the second listener fails to connect to the databases.
Problem
When upgrading a typical Release 2 (9.0.4.2) environment running two or three Oracle homes, the default listener uses port 1521 to connect to the first Oracle home only. Upgrading the remaining Oracle homes results in failure of the default listeners to connect to their respective Oracle homes as each listener attempts to use the same port (port 1521).
The following errors are thrown:
java.sql.SQLException: Listener refused the connection with the followingerror:ORA-12514, TNS:listener does not currently know of service requested inconnect descriptor
Solution
To work around the problem, stop the second listener and use NetCA/NetManager tool to register the listener with the additional databases.
If your Release 2 (9.0.4.2) infrastructure and Information Storage Database are on the same host, you must shut down your 10.1.2.2 upgraded infrastructure listener prior to running the Information Storage Database upgrade to 10.1.2.2.
Search does not return any results after updating Oracle Collaboration Suite with a customer database enabled as an Oracle Collaboration Suite database.
Problem
After upgrading Oracle Collaboration Suite with a customer database enabled as an Oracle Collaboration Suite Database, Oracle Collaboration Suite 10g Search does not return any results.
Solution
To correct this problem perform the following steps:
Log in to the Oracle Collaboration Suite Database as wk_test.
Run the following commands:
exec wk_adm.use_instance('wk_inst');
exec wk_util.register_dft_federated_sources;
Restart the OC4J_OCSClient process using OPMN by entering the following command on the Applications tier:
$ORACLE_HOME/opmn/bin/opmnctl restartproc process-type=OC4J_OCSClient
Upgrading the Mailstore may fail, causing an error.
Problem
Upgrading the Mailstore may fail, resulting in error ORA-29855 and ORA-20000.
Solution
While following the instructions in "Enabling Collaboration Suite Schema Creation in the Database Upgrade Assistant" in chapter 5 of Oracle Collaboration Suite Upgrade Guide, ORA-29855 with an Oracle Text error ORA-20000 may occur. If this happens, work around the problem as follows.
At the SQL prompt to which ORA-29855 was returned, enter the following commands.
drop index es_ot_ix_doc;
drop index es_ot_ix_search;
Enter the following command.
@oes/install/sql/recreate_text_index.sql
If recreate_text_index.sql printed a message that indicates successful completion of index creation similar to the following, skip to Step 5.
Creating text index... PL/SQL procedure successfully completed.
If the error persists, run:
BEGIN ctx_ddl.drop_preference('es_search_filter'); END;
BEGIN
ctx_ddl.create_preference('es_search_filter', 'MAIL_FILTER');
ctx_ddl.set_attribute('es_search_filter','index_fields','From:To:CC:Subject:Sender:Content-Disposition');
ctx_ddl.set_attribute('es_search_filter','PART_FIELD_STYLE','text');
END;
Enter the following command to execute recreate_text_index again:
oes/install/sql/recreate_text_index.sql
This sections describes common problems and solutions. It contains the following topics:
Section A.3.2, "No Instances Displayed in "Migrating Ultra Search Configuration Data" Screen"
Section A.3.3, "Oracle Ultra Search Upgrade Configuration Assistant Failure"
Section A.3.4, "Error Logging into the OracleAS Portal Page After the Upgrade"
Section A.3.5, "Problem Configuring the OracleAS Portal Page After the Upgrade"
Section A.3.7, "Unable to Log in to the Applications Tier through OracleAS Single Sign-On"
Section A.3.8, "Upgrade Assistant Fails to Migrate OracleAS Single Sign-On"
Section A.3.9, "Upgrade Assistant Fails to Migrate OracleAS Single Sign-On with HTTPS"
Section A.3.12, "Upgraded Oracle Collaboration Suite 10g Search Does Not Return Search Results"
Section A.3.14, "Oracle Ultra Search Configuration Assistant May Fail"
Section A.3.16, "Users Cannot Log in to Oracle Collaboration Suite Portal Page"
Section A.3.18, "Migrated orclguest Password Overwrites New orclguest Password"
Section A.3.19, "Oracle Web Conferencing Fails During Upgrade or Deconfiguration"
Section A.3.22, "Modifying Upgraded Calendar Accounts Using Delegated Administration Services"
Section A.3.24, "Calendar Resource Browsing Right is not Granted During Upgrade"
Section A.3.25, "Oracle Calendar Fails to Connect to the Database"
Section A.3.27, "Cannot Login to Oracle Discussions after 10g Upgrade"
Section A.3.29, "OCS Upgrade Assistant Hangs while Upgrading OC4J"
Section A.3.31, "OC4J Configuration Assistant Fails to Deploy ORAUDDI Application"
Problem
Upgrades of SSL 9.0.4 installations may not succeed until midtierSSLConfigTool.bat is run following the upgrade.
Solution
Edit C:\APPTIER1012_UPG\webcache>webcache.xml to change the HOST ID port number from 4444 to an unoccupied port (such as 4445).
Perform the steps in section Section 6.6.3.1, "Verifying the Secure Sockets Layer (SSL) Configuration After Upgrade".
Run midtierSSLConfigTool.bat as follows, and make sure to include the ocs_config option.
%ORACLE_HOME%\bin\midtierSSLConfigTool.bat oid_hostname oid_port oid_admin_dn oid_admin_password http_server_SSL_port https hostname_of_the_computer True | False ocs_config
Where True | False is the boolean flag that determines the communication to OID.
True: Communication to the Oracle Internet Directory using SSL only
False: Normal communication to Oracle Internet Directory
While running the Oracle Collaboration Suite Upgrade Assistants, the Instances table in the Migrating Ultra Search Configuration Data screen does not display any of the existing instances.
Problem
The WKSYS password in the Oracle9iAS Metadata Repository changed during the Oracle9iAS Infrastructure upgrade and no longer matches the password stored in Oracle Internet Directory.
Solution
Reset the WKSYS password in the Oracle9iAS Metadata Repository database to match the password in Oracle Internet Directory by performing the following steps:
Exit the Oracle Universal Installer.
Look up the WKSYS password in Oracle Internet Directory. From the system running Oracle Internet Directory enter the following command from the Oracle home bin directory:
ldapsearch -h oidhost -p oidport -D "cn=orcladmin" -w oiduser_password -b "cn=IAS Infrastructure Databases,cn=IAS,cn=Products,cn=OracleContext" -s sub "orclResourceName=wksys" orclpasswordattribute
where
oidhost is the name of the system running Oracle Internet Directory
oidport is the port where Oracle Internet Directory is running
oiduser_password is the password for the Oracle Internet Directory administrative user
The utility returns the password as the value of orclpasswordattribute. For example:
OrclResourceName=WKSYS,orclReferenceName=asdb.oracle.com,cn=IAS Infrastructure Databases,cn=IAS,cn=Products,cn=OracleContextorclpasswordattribute=T8l7Q155
On the system running the Oracle9iAS Metadata Repository, start SQL*Plus and log into the database as SYSDBA:
$ORACLE_HOME/bin/sqlplus sysdba/sysdba_password
Change the WKSYS password to the password obtained from Oracle Internet Directory by entering the following command at the SQL*Plus prompt:
ALTER USER WKSYS IDENTIFIED BY orclpasswordattribute;
where orclpasswordattribute is the password returned by the ldapsearch utility. For example:
ALTER USER WKSYS IDENTIFIED BY T8l7Q155;
Restart the Oracle Universal Installer. The existing Oracle Ultra Search instances should now display in the Instances table.
In the Configuration Assistants screen, the status of the Oracle Ultra Search Configuration Assistant is "Failed" and the log file contains the following errors:
CST ERROR:1 Framework ORA-01017: invalid username/password; logon deniedjava.sql.SQLException: ORA-01017: invalid username/password; logon denied
Problem
The WKSYS password in the Oracle9iAS Metadata Repository database was changed during the Oracle9iAS Infrastructure upgrade and no longer matches the password stored in Oracle Internet Directory. This problem is caused by a known issue.
Solution
Perform the following steps:
Leave the Oracle Universal Installer running.
Look up the WKSYS password in Oracle Internet Directory and reset it in the Oracle9iAS Metadata Repository database as explained in Section A.3.2.
Select the Oracle Ultra Search configuration assistant in the table in the Configuration Assistants screen.
Click Retry.
After the middle tier upgrade, you are not able to log into OracleAS Portal.
Problem
The following error message appears when you log into the OracleAS Portal Page:
The address from which this authentication request was made, does not match your IP address. Notify your administrator if you believe this message to be in error. (WWC-41452)
Solution
Reconfigure OracleAS Portal by entering the following command from the Applications tier Oracle home configured with OracleAS Portal:
$ORACLE_HOME/portal/conf/ptlconfig -dad DAD -sso
After upgrading the middle tier, you cannot configure the OracleAS Portal page.
Problem
If OracleAS Portal is not configured in the source oracle home, it remains unconfigured after upgrading to 10g Release 1 (10.1.2).
When logging into the OracleAS Portal Page, the Portal page does not display. Instead, a "page not found" error displays.
Solution
You can configure the OracleAS Portal after upgrading the middle tier or at the end of the entire upgrade process, after upgrading the OracleAS Metadata Repository.
To configure the Portal post upgrade, do following:
Modify the port entries in the ptlem.sh file (DESTINATION_ORACLE_HOME/assistants/opca/ptlem.sh) as follows:
http_port—Oracle HTTP Server port
wc_invalid—Web Cache Invalidation port
wc_admin—Web Cache Administration port
em_port—Application Server Control port
These port values, except http_port, are displayed in the Ports page of the Application Server Control Console.
Retrieve the http_port value from the following configuration file httpd.conf (DESTINATION_ORACLE_HOME/Apache/Apache/conf/httpd.conf).
Use the Application Server Control Console to configure the OracleAS Portal component. See " Configuring Additional Components" in the in the Application Server Control online help.
Run the ptlconfig command to associate the OracleAS Portal repository with the upgraded middle tier, as follows:
DESTINATION_ORACLE_HOME/portal/conf/ptlconfig -dad portal_dad
Refer to the Oracle Application Server Portal Configuration Guide for more information about the iasconfig.xml file and the ptlconfig tool.
If you have already configured OracleAS Portal using the Application Server Control Console, without updating the ports in the ptlem.sh file, then you must also perform the following task:
Modify the port entries in the ptlem.sh configuration file, as described previously in this section.
Backup the following Oracle Enterprise Manager configuration file:
DESTINATION_ORACLE_HOME/sysman/emd/targets.xml
Remove the entry for the OracleAS Portal target you are about to reconfigure.
You must review the entire entry that defines the OracleAS Portal target, starting with the <Target TYPE="oracle_portal"> tag and ending with the </Target> tag.
For example:
<Target TYPE="oracle_portal"
.
.
.
</Target>
Before removing the target entry for this portal, make sure that this target contains property <Property NAME="portal_DAD" VALUE="portal"/> and that its OracleHome property matches the destination middle tier Oracle home.
Save the targets.xml file.
Set the ORACLE_HOME environment variable to the destination Oracle home.
Run the ptlem script as follows:
DESTINATION_ORACLE_HOME/assistants/opca/ptlem.sh
Check the log file for any errors:
DESTINATION_ORACLE_HOME/assistants/opca/install.log
Update the following OracleAS Portal configuration file with new port values for the WebCacheComponent and EMComponent elements:
DESTINATION_ORACLE_HOME/portal/conf/iasconfig.xml
You can obtain the required port numbers from the Ports page in the Application Server Control Console.
Run the ptlconfig command to associate the OOracleAS Portal repository with the upgraded middle tier, as follows:
DESTINATION_ORACLE_HOME/portal/conf/ptlconfig -dad portal_dad [-pw <portal schema password>]
Bounce HTTP_Server and OC4J_Portal from the Application Server Control Console.
Refer to the Oracle Application Server Portal Configuration Guide for more information about the iasconfig.xml configuration file and the ptlconfig tool.
Oracle Calendar is integrated with Oracle Real-Time Collaboration. After the upgrade of Oracle Real-Time Collaboration, the status of meeting IDs is set to PENDING in Oracle Calendar clients.
Problem
As explained in Section 6.6.7, if Oracle Calendar is integrated with Oracle Real-Time Collaboration, then following the upgrade of Oracle Real-Time Collaboration, you need to reconfigure Oracle Calendar to get the new value of the key used to authenticate with Oracle Real-Time Collaboration. Until you perform the steps in this section, Oracle Calendar cannot communicate with Oracle Real-Time Collaboration. The status of meeting IDs is set to PENDING in Oracle Calendar clients.
Solution
Setting Oracle Calendar's initialization parameter siteauthkey with the new value allows Web conferences created in Oracle Calendar to propagate to Oracle Real-Time Collaboration.
Perform the steps in Section 6.6.7.
If you do not perform these steps, then 72 hours after a meeting ID is set to PENDING, the status will remain PENDING. The only way to update the meeting ID is to change the meeting time or title.
After upgrading to Oracle Collaboration Suite 10g Applications, users cannot log in using OracleAS Single Sign-On.
Problem
When trying to log in to the upgraded Applications tier, users get an error message "Oracle SSO Failure - Unable to process request."
Solution
Change the port where the users are accessing OracleAS Single Sign-On from the HTTP listening port to the Web cache port using the OracleAS Single Sign-On administration page. For example, if the HTTP listening port is 7778 and the Web cache port is 7777, change the port for accessing OracleAS Single Sign-On to 7777.
While upgrading the Oracle Collaboration Suite middle tier from Release 2 (9.0.4.1) to 10.1.2, the Upgrade Assistant fails to migrate the OracleAS Single Sign-On component.
Problem
When upgrading the Oracle Collaboration Suite middle tier, the following errors are thrown:
2006/01/18 14:35:06.137 CST NOTIFICATION:1 OHS: The osso.conf file will be upgraded by making required directories in the target OH and copying the file over. Then, ssomigrate and ssoPropagate tools will be run to generate osso.conf in the DCM repository. 2006/01/18 14:35:06.140 CST NOTIFICATION:1 OHS: cp /cfc/oracle/calfdc/Apache/Apache/conf/osso/osso.conf /cfc/oracle/cal_1012/Apache/Apache/conf/osso/osso.conf 2006/01/18 14:35:06.140 CST NOTIFICATION:1 OHS: The above target file will not be overwritten if it exists 2006/01/18 14:35:06.141 CST WARNING:1 OHS: /cfc/oracle/cal_1012/Apache/Apache/conf/osso/osso.conf exists. 2006/01/18 14:35:06.141 CST TRACE:17 OHS: The execution string for the OSSO utility is: /cfc/oracle/cal_1012/Apache/Apache/bin/ssomigrate /cfc/oracle/calfdc/Apache/Apache/conf/osso/osso.conf /cfc/oracle/cal_1012/Apache/Apache/conf/osso/osso.conf clrtxt /cfc/oracle/cal_1012/Apache/Apache/conf/osso/osso.conf.clr 2006/01/18 14:35:06.447 CST ERROR:1 OHS: Could not deobfuscate /cfc/oracle/calfdc/Apache/Apache/conf/osso/osso.conf, error = -1013! 2006/01/18 14:35:06.447 CST ERROR:1 OHS: (Bad padding pattern detected in the last block) 2006/01/18 14:35:06.455 CST ERROR:1 OHS: The OSSO utility encountered errors!! 2006/01/18 14:35:06.456 CST ERROR:1 OHS: The OSSO utility encountered errors!! oracle.ias.upgrade.exception.UpgradeException: The OSSO utility encountered errors!! at oracle.ias.upgrade.config.ohs.OHSConfigUpgrade.runSSOMigrate(OHSConfigUpgrade. java:693)
Solution
To work around the problem:
In a text editor, edit the ias.properties file.
Clear the Oracle Internet Directory password.
Run the resetiASPassword script.
Execute the following command:
dcmctl shutdown force -v -o <ORACLE_HOME>
where <ORACLE_HOME> is the exact path of the Oracle home.
Run the Configuration Assistant again.
While upgrading the Oracle Collaboration Suite middle tier from Release 2 (9.0.4.1) to 10.1.2, the Upgrade Assistant fails to migrate OracleAS Single Sign-On with HTTPS.
Problem
When upgrading the Oracle Collaboration Suite middle tier, the following errors are thrown:
2006/01/18 16:55:16.698 CST ERROR:1 OHS: Could not deobfuscate /cfc/oracle/calfdc/Apache/Apache/conf/osso/<hostname>.https-osso.conf, error = -1013! 2006/01/18 16:55:16.698 CST ERROR:1 OHS: (Bad padding pattern detected in the last block) 2006/01/18 16:55:16.709 CST ERROR:1 OHS: The OSSO utility encountered errors!! 2006/01/18 16:55:16.711 CST ERROR:1 OHS: The OSSO utility encountered errors!! oracle.ias.upgrade.exception.UpgradeException: The OSSO utility encountered errors!! at oracle.ias.upgrade.config.ohs.OHSConfigUpgrade.runSSOMigrate(OHSConfigUpgrade.java:693) at oracle.ias.upgrade.config.ohs.OHSConfigUpgrade.doUpgrade(OHSConfigUpgrade.java:512) at oracle.ocs.upgrade.UpgradeSupport.doUpgrade(UpgradeSupport.java:964) at oracle.ocs.upgrade.OutOfPlace.execute(OutOfPlace.java:805) at oracle.ocs.upgrade.UpgradeDriver.outOfPlace(UpgradeDriver.java:330) at oracle.ocs.upgrade.UpgradeDriver.main(UpgradeDriver.java:267)
Solution
To work around the problem:
In a web browser, navigate to the URL of your OracleAS Single Sign-On server as follows: https://<ssoserver_host>:<port>/pls/orasso, where ssoserver_host is the server host name and port is the port number.
Login as orasso and do the following:
Edit Partner Application
Add Partner Application
Click OK to display the partner information
Paste the following text into a file, such as clearcontent.txt:
sso_server_version=v1.2 cipher_key= site_id= site_token= login_url=http://<host>:7777/pls/orasso/orasso.wwsso_app_admin.ls_login logout_url=http://<host>:7777/pls/orasso/orasso.wwsso_app_admin.ls_logout cancel_url=http://<host>:8000 sso_timeout_cookie_name=SSO_ID_TIMEOUT sso_timeout_cookie_key=9E231B3C1A3A808A
where <host> is the name of the OracleAS Single Sign-On server.
Using the partner information you displayed in step 2, fill in values for cipher_key, site_id, and site_token.
Execute the following:
ORACLE_HOME/Apache/Apache/bin/iasobf clearcontent.txt target_osso.conf_path root
Restart the source Oracle home. If the source Oracle home no longer reports deobfuscation errors while reading the target_osso.conf_path file, resume upgrading the Oracle Collaboration Suite by running the Configuration Assistant again.
Following the upgrade, remove the clearcontent.txt file from the system as it contains sensitive key information.
When you click on a link to Oracle Calendar server from the Oracle Enterprise Manager or from Oracle Calendar application system, you cannot access it through the URL specified.
Problem
The upgrade process registers the Oracle Calendar service using the https port. Oracle Enterprise Manager and Oracle Calendar application system are trying to access Oracle Calendar server using this port. However, the URL will not work because the https port has not been registered as an OracleAS Single Sign-On Partner Application.
This problem will occur if you are upgrading from Release 1 (9.0.3.1) and did not configure the middle tier https port as a Partner Application or if you are upgrading from Release 2 (9.0.4.2) and you removed the https port as a Partner Application but left the https port in the webcache.xml file.
Solution
Check the OracleAS Single Sign-On administration page to make sure that the URL specified by Oracle Enterprise Manager 10g or another application is registered as a OracleAS Single Sign-On Partner Application.
During the middle tier upgrade, after the Specify Username and Password for Oracle Internet Directory screen, the following error appears and the upgrade fails:
There are no Oracle Collaboraion Suite databases registered with thespecified OID. You must specify a different OID install and register an existing database Oracle Collaboration Suite database before you cancontinue Database information is unavailable for the following componentsOracle collaboration suite search- oracle calendar server"
Problem
This error occurs when the middle tier being upgraded does not use an Oracle Collaboration Suite information storage database. This configuration was possible if you installed a middle tier and configured only Oracle Calendar. This component used its own database in previous releases.
Solution
Load the Oracle Collaboration Suite schema information into the Infrastructure database by perform the following steps:
Exit the Oracle Universal Installer.
In the upgraded Oracle Collaboration Suite 10g Infrastructure, copy the install/OCSdbSchemaReg.ini.sample file to OCSdbSchemaReg.ini. Open OCSdbSchemaReg.ini and fill in values for all the variables listed in the file.
Run the following script from ORACLE_HOME/install where ORACLE_HOME is the Oracle home of the upgraded Infrastructure:
ORACLE_HOME/install/OCSdbSchemaReg.sh -f ORACLE_HOME/install/OCSdbSchemaReg.ini
This configures the Oracle Collaboration Suite 10g Infrastructure database to work with components from Oracle Collaboration Suite 10g Applications and performs the necessary registration with Oracle Internet Directory.
Run the Oracle Universal Installer again.
After the upgrade, queries performed using Oracle Collaboration Suite 10g Search do not return any results.
Problem
This problem occurs when there are strings consisting of multi-byte characters in the file DESTINATION_ORACLE_HOME/j2ee/OC4J_OCSClient/connectors/UltraSearch/UltraSearch/META-INF/oc4j-ra.xml.
For example, you create an instance with a name that uses multi-byte characters and you configure this instance as the source of Oracle Collaboration Suite 10g Search. When you copy and paste this string into oc4j-ra.xml, the Oracle Collaboration Suite 10g Search application does not recognize the instance name and does not perform the search.
Solution
Convert the oc4j-ra.xml file to UTF-8 encoding by performing the following steps:
Edit the DESTINATION_ORACLE_HOME/j2ee/OC4J_OCSClient/connectors/UltraSearch/UltraSearch/META-INF/oc4j-ra.xml file by assigning the username, password and instancename properties. For example:
<config-property name="userName" value="ultra_tw"/>
<config-property name="password" value="welcome"/>
<config-property name="instanceName" value="multibytename"/>
where multibytename is a multi-byte character string, for example a traditional Chinese string copied and pasted from a Big-5 encoded document.
Download the oc4j-ra.xml file to your local system.
Convert oc4j-ra.xml to UTF-8 using a character encoding conversion tool.
Upload the new oc4j-ra.xml in UTF-8 to DESTINATION_ORACLE_HOME/j2ee/OC4J_OCSClient/connectors/UltraSearch/UltraSearch/META-INF/
Removed the cached version of oc4j-ra.xml from DESTINATION_ORACLE_HOME/j2ee/OC4J_OCSClient/application-deployments/defaults/UltraSearch/oc4j-ra.xml
Restart the OC4J_OCSClient instance.
Manually migrating the Oracle Ultra Search index and configuration data and creating a new schema with a new password results in an outdated password in the data-sources.xml file.
Problem
If you performed the procedures to manually migrate the Oracle Ultra Search index and configuration data and you created a new schema with a different password than the original schema, then the password in the data-sources.xml file is out-of-date. This issue and the action required to correct it pertain only to Oracle Ultra Search and not to the Oracle Collaboration Suite 10g Release 1 (10.1.2) Federated Search application.
Solution
Open the j2ee/OC4J_OCSADMIN/config/data-sources.xml file in the upgraded Applications tier Oracle home and change the password value to the new password.
The Ultra Search Configuration Assistant may fail during an upgrade.
Problem
In certain cases, the Ultra Search Configuration Assistant may fail during an upgrade. The error given looks similar to the following:
Failed to create service entries for existing Ultra Search instances java.sql.SQLException: ORA-20000: Oracle Ultra Search error ORA-06512: at "WKSYS.WK_ERR", line 179 ORA-06512: at line 1
This may happen if the following catalogs were created in Oracle Internet Directory using an Oracle Collaboration Suite tool or a manual workaround:
targetdn
orclownerguid
pager
personalmobile
personalpager
middlename
orclSuiteType
orclServiceType
orclServiceSubType
orclServiceMember
orclstatus
In this case, during the upgrade, the command to create these catalogs fails, and subsequent catalogs for other components are also not created.
To check if a catalog exists, log in to the Metadata Repository Database as the ODS user and then perform the following command:
desc ct_catalog_name;
In the preceding command, catalog_name should be replaced by the name of the catalog you need to check.
For example, the command to check whether the targetdn catalog exists will be:
desc ct_targetdn;
If the message Òct_catalog_name does not existÓ is displayed, then you can proceed with the upgrade normally, without the need of any additional steps. However, if this message is not displayed, then you must perform the following workaround.
Workaround
Perform the following steps:
Before upgrading Oracle Internet Directory, remove the catalogs using catalog.sh.
$ORACLE_HOME/ldap/bin/catalog.sh -connect connect_descriptor -delete -attr\
catalog_name
In the preceding command, replace catalog_name with the name of the catalog you need to remove.
For example, to remove the targetdn catalog:
$ORACLE_HOME/ldap/bin/catalog.sh -connect connect_descriptor -delete -attr\ targetdn
In the preceding command, connect_string is the database connect string for the Infrastructure database, such as asdb.
|
Note: After this step you must upgrade Oracle Internet Directory. |
If you are upgrading an Applications tier configured with Oracle Ultra Search, then download patch 4493920 from Oracle MetaLink and apply it to the upgraded Oracle Collaboration Suite Database 10g (10.1.2), according to the instructions contained in the patch ZIP file.
The patch fixes the following issues:
The configtool.log file contains the following Oracle Collaboration Suite Upgrade Assistant error message:
oracle.ocs.upgrade.exception.UpgradeException: Can not migrate data frommetadata repository database to Collaboration Suite database.
The wk0create_user.sql script does not grant all necessary privileges.
The Oracle Ultra Search Configuration Assistant fails when Oracle Internet Directory is in SSL-Only mode.
After the Applications tier upgrade, the Web search does not return results.
If you encounter any of these issues during or after the Applications tier upgrade, then apply the patch to the upgraded Oracle Collaboration Suite Database 10g (10.1.2).
After upgrading from Oracle Collaboration Suite Release 1 (9.0.3.1), when some users log into the Oracle Collaboration Suite 10g Release 1 (10.1.2) portal page at:
http://hostname:port/pls/portal
all the portlets display errors.
Problem
This error occurs for the orclguest user or for any user whose default portal page changed. When logging in to the portal, the user is redirected to the old portal page rather than the new Oracle Collaboration Suite 10g Release 1 (10.1.2) Collaborative Portlets page.
Solution
To set the user's default home page to the new portal, perform the following steps:
Go to the Oracle Collaboration Suite 10g Release 1 (10.1.2) portal page and log in as orcladmin.
Select the Builder link.
Click the Administer tab.
In the Portal User Profile portlet, enter orclguest, then click Edit.
On the Default Home Page field, click Reset.
Click OK.
When logging into the portal page, the user is directed to the new page.
Users cannot sign on to a new Oracle Content Services instance that is using an upgraded Oracle Internet Directory.
Problem
This problem occurs because of insufficient access rights set in the upgradedOracle Internet Directory instance.
Solution
Grant Browse access to Everyone on the user search base and group search base containers for each realm in Oracle Internet Directory. This needs to be done only for orclentrylevelaci.Follow these steps to grant Browse access:
As the orcladmin user, connect to Oracle Internet Directory using Oracle Directory Manager.
For each realm, browse to the user search base and group search base containers. The following examples show user search base and group search base containers for a realm named "us" and a realm search base
"dc=oracle,dc=com": cn=Users,dc=us,dc=oracle,dc=com cn=Groups,dc=us,dc=oracle,dc=com
For each container, click the Local Access tab, then grant Browse access to Everyone in the Structural Access Items box by selecting Everyone and clicking Edit for the existing ACI.
When creating a new orclguest user account during an upgrade with an existing, migrated orclguest user account, the new password you enter is ignored.
Problem
During an upgrade, you are informed that an orclguest user account will be created and you are asked to enter a password for it. However, if an orclguest user account exists in the version being upgraded, the password you enter is ignored and the existing account is simply migrated.
Solution
If an orclguest user account exists in the version being upgraded, just enter a false password when prompted. The password you enter will be ignored, and the original account information (including its password) will be migrated.
When upgrading or deconfiguring Oracle Web Conferencing in Oracle Collaboration Suite Release 2 (9.0.4.2), Oracle Universal Installer reports that the password for RTC_APP is incorrect.
Problem
Different passwords were used for RTC, RTC_APP, and RTC_REP in Oracle Collaboration Suite Release 2 (9.0.4.2), or the old password for RTC_APP was not specified during Information Storage upgrade. Oracle Collaboration Suite Release 2 (9.0.4.2) retains the old RTC_APP password, while Oracle Internet Directory saves the new password.
Solution
Change the new RTC_APP password so that it matches the old, pre-upgrade RTC schema password.
Upgrading Oracle Collaboration Suite Release 2 (9.0.4.2) without Oracle Email configured results in an error.
Problem
When upgrading to Oracle Collaboration Suite 10g (10.1.2) including Oracle Mail, the following error is thrown:
error validating es_mail password
Solution
Login to the Oracle Mail database and run the following query:
alter user es_mail identified by welcome1 account unlock
When configuring Oracle Real-Time Collaboration through Oracle Enterprise Manager, the HTTP Listen Port property is not filled in. As a result, Apache sets the listening port to the default value of 80.
Problem
An incorrect listening port is set after leaving blank the HTTP Listen Port property for Oracle Real-Time Collaboration in Oracle Enterprise Manager.
Solution
Change the ApacheWebPort value to 7777 using the rtcctl command.
After an upgrade, if you try to modify certain user accounts using Delegated Administration services, an error similar to the following can occur:
Error Post Plugin Errors. Calendar - oracle.idm.provisioning.plugin.PluginException: Calendar storage cannot be modified. Use the calendar administration tool to move the calendar account from one storage location to another
This generally happens with user accounts that did not exist in Oracle Calendar before the upgrade. (4513293)
There are two workarounds for this problem; both involve changing the status of the account from UPGRADE_IN_PROGRESS to PROVISIONING_NOT_REQUIRED.
Workaround 1: Using the Oracle Directory Manager
Start Oracle Directory Manager:
$ORACLE_HOME/bin/oidadmin Start > Programs > Oracle Application Server - OracleHome >Integrated Management Tools > Oracle Directory Manager
Log in to Oracle Directory Manager.
In the System Objects frame:
Expand Entry Management.
Navigate to the appropriate subscriber dn, such as dc=com | dc=acme.
Expand cn=users.
Select the affected user.
In the right pane, find the following attribute:
orclUserApplnProvStatus;CALENDAR_CALENDAR
Set its value to PROVISIONING_NOT_REQUIRED.
Workaround 2: Using the command line
Run the following command:
% printf "dn: `$ORACLE_HOME/bin/ldapsearch -host host -port port -D managerdn -w password -b basedn (mail=mail_of_the_user) dontremoveme`\nchangetype: modify\nreplace: orclUserApplnProvStatus;CALENDAR_CALENDAR\norclUserApplnProvStatus;CALENDAR_CALENDAR: PROVISIONING_NOT_REQUIRED\n" | $ORACLE_HOME/bin/ldapmodify -host host -port port -D managerdn -w password
For example:
% printf "dn: `$ORACLE_HOME/bin/ldapsearch -host host1.acme.com -p 389 -D cn=orcladmin -w test1 -b dc=acme,dc=com (mail=john.doe@acme.com) dontremoveme`\nchangetype: modify\nreplace: orclUserApplnProvStatus;CALENDAR_CALENDAR\norclUserApplnProvStatus;CALENDAR_CALENDAR: PROVISIONING_NOT_REQUIRED\n" | $ORACLE_HOME/bin/ldapmodify -host host1.acme.com -p 389 -D cn=orcladmin -w test1
The value 10.1.2 is not accepted by the upgrade assistant for the destver parameter.
Problem
When upgrading standalone Oracle Calendar application system installations, the upgrade assistant does not accept 10.1.2 as the destver parameter.
Solution
Do not use the destver parameter—it is not needed in 10g Release 1 (10.1.2). For example:
/u01/ocs10/ocas/upgrade/ocasua.sh -srcver 9.0.4 -srcdir /u01/ocs904 -confdir /usr/local/apache/conf
The Calendar resource browsing right is not created in an upgrade scenario (4507691). In addition, in an upgrade scenario, Oracle Workspaces does not have sufficient permissions to list Calendar resources.
Problem
Permission to browse Calendar resources is not granted during an upgrade.
Solution
To correct the problem:
Create an LDIF file with the following contents:
dn: cn=Resources,cn=Calendar,cn=Products,cn=OracleContext,<subscriber dn> changetype: modify add: orclaci orclaci: access to entry filter=(objectclass=orclCalendarResource) by dn=".*,cn=Application Entities,cn=Calendar,cn=Products,cn=OracleContext" (browse,add,delete) by * BindMode="Proxy" (browse,noadd,nodelete) - add: orclaci @ orclaci: access to attr!=(userpassword, @ authpassword,orclpassword,orclPasswordAttribute) by dn=".*,cn=Application Entities,cn=Calendar,cn=Products,cn=OracleContext" (read,search,compare,write,selfwrite) by * BindMode="Proxy" (read,search,compare)
Run the following command to apply the LDIF:
ldapmodify -h hostname -p port -D binddn -w passwd -f file
While upgrading Oracle Calendar, the JDBC connector fails to connect to the Oracle Calendar database.
Problem
Upgrading Oracle Calendar to 10.1.2 fails with the following error:
Error connecting to the database you have selected for Oracle Calendar Contacts and Provisioning. Please make sure that the database is up and running and the CALENDAR user account is unlocked or select another database.
Solution
To work around the problem:
Log into Oracle Internet Directory Manager (oidadmin).
Remove the SERVER=DEDICATED parameter from the database connect descriptor using ldapmodify or oidadmin.
After performing the 10g (10.1.2) upgrade, the Search link does not function.
Problem
If the Oracle Collaboration Suite middle tier (9.0.3.1, 9.0.4.1 or 9.0.4.2) has Files configured and portal not configured, after upgrading to 10g (10.1.2), the Search link does not function.
Solution
Upon completing the upgrade, you must manually change the port value of the Search URL in the Oracle Internet Directory service registry to the value of your HTTP Server Port. Afterwards, restart the OC4J_OCSClient process on the middle tier.
It may not be possible to log in to Oracle Discussions following an upgrade.
Problem
After upgrading, it may not be possible to log into Oracle Discussions as orclguest or as any other user.
Solution
Do the following on the infrastructure tier (requires ods password):
Stop Oracle Internet Directory.
Delete the targetdn catalog:
$INFRA_ORACLE_HOME/ldap/bin/catalog.sh -connect connect_descriptor_of_OiD_db -delete -attr targetdn
Recreate the targetdn catalog:
$INFRA_ORACLE_HOME/ldap/bin/catalog.sh -connect connect_descriptor_of_OiD_db -add -attr targetdn
Start Oracle Internet Directory.
Restart the Oracle Mail and Oracle Discussions Applications tiers.
When upgrading from Oracle Collaboration Suite 9.0.4 email to 10g (10.1.2), the upgrade_mailstore.pl fails with a duplicate value in the index.
Problem
This issue is caused by duplicate values of username/domain in the es_user table.
Solution
Update all but one row to a non-existant domain name.
When upgrading the Oracle Collaboration Suite Middle tier from Release 2 (9.0.4.2) to 10.1.2, the Oracle Collaboration Suite Upgrade Assistant hangs while attempting to upgrade the Oracle Application Server Containers for J2EE.
Problem
While upgrading the Oracle Application Server Containers for J2EE, the Oracle Collaboration Suite Upgrade Assistant hangs, displaying the following message:
Upgrading Oracle Application Server Containers for J2EE (OC4J).........
Solution
To work around the problem, kill the running Oracle Collaboration Suite Upgrade Assistant process and retry.
When upgrading the Oracle Collaboration Suite Middle Tier, the Oracle Collaboration Suite Upgrade Assistant log file (ocsua.log) shows an error.
Problem
The source file does not exist in the source directory, yet it exists in the target directory. The ocsua.log file lists the following error:
Could not copy $OH/install/ocs_location_directives.conf to $OH/Apache/Apache/conf at $OH/install/apache_s2s_directives.pl line 63
Solution
Comment out the copy command in lines 63 and 64 of the apache_s2s_directives.pl file.
The OC4J Configuration Assistant fails to deploy the ORAUDDI application while running the Oracle Collaboration Suite Oracle Universal Installer to upgrade the 9.0.4.2 middle tier Oracle home with only Oracle Calendar Server enabled.
Problem
The Oracle Universal Installer indicates a "Failed" status in the "OC4J Instance Configuration Assistant" pane.
The following information is recorded in the InstallActions log:
Launched configuration assistant 'OC4J Instance Configuration Assistant'
Tool type is: Recommended.
The command being spawned is: '/tbmwrc/collab/product/csmt_101/jdk/bin/java
-Djava.net.preferIPv4Stack=true -Djava.io.tmpdir=/tmp -mx512M -classpath
/tbmwrc/collab/product/csmt_101/dcm/lib/dcm.jar:/tbmwrc/collab/product/csmt_10
1/dcm/lib/oc4j_deploy_tools.jar:/tbmwrc/collab/product/csmt_101/opmn/lib/opmnp
lugin.jar
-Doracle.ias.sysmgmt.logging.logdir=/tbmwrc/collab/product/csmt_101/j2ee/home/
log oracle.j2ee.tools.deploy.Oc4jDeploy -oraclehome
/tbmwrc/collab/product/csmt_101 -verbose -inifile
/tbmwrc/collab/product/csmt_101/j2ee/deploy.ini '
.
Start output from spawned process:
----------------------------------
.
Reading ini file - '/tbmwrc/collab/product/csmt_101/j2ee/deploy.ini'
Adding web-app 'orauddi.war' for app 'orauddi'.
Adding web-app 'oraudrepl.war' for app 'oraudrepl'.
Initializing DCM...done.
OC4J instance 'OC4J_Portal' already exists.
Copying LDAP-configured jazn.xml from 'home' OC4J instance...done.
Adding dependent libraries for application 'default'...done.
Starting OC4J instance 'OC4J_Portal'...done.
Deploying application 'orauddi' to OC4J instance 'OC4J_Portal'.
java.lang.reflect.InvocationTargetException
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at
sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
at
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
at java.lang.reflect.Method.invoke(Method.java:324)
at oracle.j2ee.tools.deploy.Oc4jDeploy.doDeploy(Unknown Source)
at oracle.j2ee.tools.deploy.Oc4jDeploy.execute(Unknown Source)
at oracle.j2ee.tools.deploy.Oc4jDeploy.deploy(Unknown Source)
at oracle.j2ee.tools.deploy.Oc4jDeploy.main(Unknown Source)
Caused by: com.evermind.client.orion.AdminCommandException: Deploy error:
deploy failed!: ; nested exception is:
oracle.oc4j.admin.internal.DeployerException: Constructing
UserManager for orauddi:caught oracle.security.jazn.JAZNInitException
at
com.evermind.client.orion.DeployCommand.execute(DeployCommand.java:90)
at
com.evermind.client.orion.Oc4jAdminConsole.executeCommand(Oc4jAdminConsole.java:139)
... 8 more
Caused by: java.rmi.RemoteException: deploy failed!: ; nested exception is:
oracle.oc4j.admin.internal.DeployerException: Constructing
UserManager for orauddi:caught oracle.security.jazn.JAZNInitException
at
com.evermind.server.rmi.RMIConnection.EXCEPTION_ORIGINATES_FROM_THE_REMOTE_SER
VER(RMIConnection.java:1621)
at
com.evermind.server.rmi.RMIConnection.invokeMethod(RMIConnection.java:1572)
at
com.evermind.server.rmi.RemoteInvocationHandler.invoke(RemoteInvocationHandler.java:55)
at
com.evermind.server.rmi.RecoverableRemoteInvocationHandler.invoke(RecoverableR
emoteInvocationHandler.java:22)
at __Proxy0.deploy(Unknown Source)
at
com.evermind.client.orion.DeployCommand.execute(DeployCommand.java:87)
... 9 more
Caused by: oracle.oc4j.admin.internal.DeployerException: Constructing
UserManager for orauddi:caught oracle.security.jazn.JAZNInitException
at
oracle.oc4j.admin.internal.DeployerBase.execute(DeployerBase.java:91)
at
com.evermind.server.administration.DefaultApplicationServerAdministrator.internalDeploy(DefaultApplicationServerAdministrator.java:429)
at
com.evermind.server.administration.DefaultApplicationServerAdministrator.deploy(DefaultApplicationServerAdministrator.java:348)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at
sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
at
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
at java.lang.reflect.Method.invoke(Method.java:324)
at
com.evermind.server.rmi.RMICallHandler.run(RMICallHandler.java:124)
at com.evermind.server.rmi.RMICallHandler.run(RMICallHandler.java:48)
at
com.evermind.util.ReleasableResourcePooledExecutor$MyWorker.run(ReleasableResourcePooledExecutor.java:186)
at java.lang.Thread.run(Thread.java:534)
Caused by: java.lang.InstantiationException: Constructing UserManager for
orauddi:caught oracle.security.jazn.JAZNInitException
at
com.evermind.server.XMLServerConfig.mkUserManager(XMLServerConfig.java:183) atcom.evermind.server.XMLServerConfig.initUserManager(XMLServerConfig.java:333) atcom.evermind.server.XMLServerConfig.initUserManagers(XMLServerConfig.java:240)
at
com.evermind.server.XMLServerConfig.postInit(XMLServerConfig.java:643) at com.evermind.xml.XMLConfig.init(XMLConfig.java:181) at
com.evermind.server.ServerComponent.initDeployment(ServerComponent.java:317)
at
com.evermind.server.ServerComponent.copyAndDeploy(ServerComponent.java:386)
at
com.evermind.server.ServerComponent.initDeployment(ServerComponent.java:288)
at
oracle.oc4j.admin.internal.ApplicationDeployer.initArchive(ApplicationDeployer.java:300)
at
oracle.oc4j.admin.internal.ApplicationDeployer.doDeploy(ApplicationDeployer.java:130)
at
oracle.oc4j.admin.internal.DeployerBase.execute(DeployerBase.java:73)
... 10 more
Deploying application 'oraudrepl' to OC4J instance 'OC4J_Portal'.
FAILED!
Solution
To work around the problem:
Edit the deploy.ini file to remove all entries related to the orauddi application.
Click Retry to continue.
This section contains the following topics:
After applying the patch to upgrade the Oracle9iAS Portal schema to Oracle Application Server 10g (9.0.4), the OracleAS Metadata Repository Upgrade Assistant reports that the version is still at 9.0.2.3.
Problem
This error occurs because one of the Applications tiers is still running.
Solution
Check each Applications tier that uses the Oracle9iAS Metadata Repository being upgraded and make sure that there are no processes executing in the Applications Oracle home.
The message "No Suitable Driver" displays when logging into the Oracle Ultra Search administrative page.
Problem
This error is caused by a missing Oracle Internet Directory entry.
Solution
Add the entry to Oracle Internet Directory by performing the following steps:
Log into Oracle Internet Directory Manager (oidadmin).
Navigate to Entry Management -> cn=OracleContext -> cn=Portal -> cn=Ultrasearch -> cn=Database Instances -> orclApplicationCommonName=db_name -> cn=Associated Mid-tiers
In the uniquemember attribute field, add the following line:
orclapplicationcommonname=mid_instance_name,cn=midtier instances,cn=ultrasearch,cn=portal,cn=products,cn=oraclecontext
where mid_instance_name is the Oracle Application Server instance name of the Oracle Ultra Search Applications tier, for example, mid_upgrade.mysystem.oracle.com.
Click Apply.
The message "Calling upgrade plugin for WIRELESS" is logged, but MRUA is stuck in wireless upgrade.
Problem
Wireless upgrade hangs due to some job in the queue not committing its transaction.
Solution
Perform the following:
Check if job queue process is non zero:
SQL> show parameter job_queue_processes;NAME TYPE VALUE------------------------------------ ----------- ------------------------job_queue_processes integer 6
Stop the mrua process.
Set the job queue process to zero.
SQL> alter system set job_queue_processes = 0;
Restart the Metadata Repository database.
Run mrua again.
Change job_queue_processes to original non-zero value.
While running Oracle Application Server Metadata Repository Upgrade Assistant, the upgrade plugin for Wireless fails.
Problem
The following error is written to the command line:
Calling upgrade plugin for WIRELESS Error: Component upgrade failed WIRELESS Error: WIRELESS component version is: 9.0.2.8.0 UPGRADING
Solution
If Wireless was not configured in the 9.0.4.2 source environment, do the following:
Download Automated Release Update 8564583 from the following URL: ().
http://aru.us.oracle.com:8080/ARU/ViewPatchRequest/process_form?aru=8251016
Navigate to the mrua\wireless\lib directory and replace the old wireless.jar file with the new wireless.jar file.
Run Oracle Application Server Metadata Repository Upgrade Assistant.
This section lists additional known issues and solutions for Oracle Collaboration Suite Upgrade.
Table A-1 Known Issues in Oracle Collaboration Suite Upgrade
| Description | Action |
|---|---|
|
On most platforms, the path for sockets used by FastCGI is limited to 108 characters. If your path exceeds this length, you may see the following error during the Oracle HTTP Server upgrade:
|
Before starting the middle tier upgrade, add the following line to the FastCGI module configuration section in httpd.conf file located in the source Oracle home's Apache/Apache/conf directory: FastCgiIpcDir "/tmp/fcgi" For example:
|
|
Single Sign-On fails with Mozilla/Firefox-based browsers after performing an upgrade |
To work around this issue, either:
|
|
After upgrading the middle tier, Web searches using Oracle Collaboration Suite Search do not return any results |
If you did not select the option to use the existing index, then after the middle tier upgrade, modify the oc4j-ra.xml file so that Oracle Collaboration Suite Search points to the index in the upgraded Oracle Collaboration Suite Database.Perform the following steps on the system running the upgraded Applications tier.
|
|
When upgrading to Oracle Real Time Collaboration 10g Release 1 (10.1.2), the schema password supplied must be the RTC_APP schema password and all Oracle Real Time Collaboration schema passwords (RTC, RTC_REP, RTC_IM, and RTC_APP) get reset to the supplied password. |
The workaround is to change the RTC_APP schema password to be the same as the RTC schema password before upgrading. |
|
If the e-mail schema in Oracle Internet Directory is upgraded more than once (as part of an e-mail Applications tier installation), you may see errors starting with the message |
Ignore these errors as they do not affect email or any other related service. |
|
When upgrading the Oracle Collaboration Suite Middle tier, Oracle Ultra Search throws multiple errors but the upgrade succeeds. |
This is not an error. The upgrade process results in the database link dropping before it is created. When this occurs, if the database does not exist (which is likely) then an error will be logged. Turning on tracing will reveal the line 'Error dropping db link. db link might not exist.' before the stack trace. |
You can find more solutions on Oracle MetaLink, http://metalink.oracle.com. If you do not find a solution for your problem, log a service request.
|
See Also: Oracle Collaboration Suite Release Notes for Solaris Operating System, available on the Oracle Technology Network:http://www.oracle.com/technology/documentation/ |
