Oracle® Fusion Middleware Enterprise Deployment Guide for Oracle Enterprise Content Management Suite 11g Release 1 (11.1.1) E15483-02 |
|
![]() Previous |
![]() Next |
This chapter describes how to extend a domain with Oracle Universal Content Management (Oracle UCM) using the Oracle Fusion Middleware Configuration Wizard. It contains the following sections:
Section 7.4, "Starting Node Manager on ECMHOST1 and ECMHOST2"
Section 7.6, "Starting and Configuring the WLS_UCM1 Managed Server"
Section 7.7, "Updating the cwallet File in the Administration Server"
Section 7.8, "Starting and Configuring the WLS_UCM2 Managed Server"
Section 7.10, "Configuring Oracle HTTP Server for the WLS_UCM Managed Servers"
Section 7.11, "Validating Access Through Oracle HTTP Server"
Section 7.12, "Configuring Node Manager for the WLS_UCM and WLS_IPM Managed Servers"
Important: Oracle strongly recommends that you read the release notes for any additional installation and deployment considerations prior to starting the setup process. |
The Oracle Enterprise Content Management Suite system is installed using the WL_HOME and ORACLE_HOME locations created in Chapter 3, "Installing the Software" on a shared storage. ECMHOST1 and ECMHOST2 mount MW_HOME and reuse the existing WLS, SOA, and ECM binary installations. The pack
and unpack
utilities are used to bootstrap the domain configuration for the WLS_UCM1 and WLS_UCM2 servers in these two new nodes. As a result, you do not need to install any software in these two nodes. For the Oracle ECM system to work properly, ECMHOST1 and ECMHOST2 must maintain the same system requirements and configuration that was required for installing Oracle Fusion Middleware in SOAHOST1 and SOAHOST2. Otherwise, unpredictable behavior in the execution of binaries may occur.
Note: You will have already added SOA components to the domain as described in Section 6, "Extending the Domain with SOA Components." |
You must extend the domain created in Section 5, "Creating a Domain with Administration Server" to include UCM. The instructions in this section assume that the ECM deployment uses the same database service as the SOA deployment (ecmedg.mycompany.com).
Note: Before performing these steps, back up the domain as described in the Oracle Fusion Middleware Administrator's Guide. |
Perform these steps to extend the domain to include Oracle UCM:
Ensure that the database where you installed the repository is running. For Oracle RAC databases, it is recommended that all instances are running, so that the validation check later on becomes more reliable.
Change the directory to the location of the Oracle Fusion Middleware Configuration Wizard. This is within the common home directory (notice that domain extensions are run from the node where the Administration Server resides).
SOAHOST1> cd ORACLE_COMMON_HOME/common/bin
Start the Configuration Wizard:
SOAHOST1> ./config.sh
In the Welcome screen, select Extend an existing WebLogic domain, and click Next.
In the WebLogic Domain Directory screen, select the WebLogic domain directory (ORACLE_BASE/admin/domain_name/aserver/domain_name), and click Next.
The Select Extension Source screen opens. In this screen, do the following (as shown in Figure 7-1):
Select Extend my domain automatically to support the following added products.
Select the following product:
Oracle Universal Content Management - Content Server - 11.1.1.0 [ecm]
Click Next.
Figure 7-1 Select Extension Source screen for Oracle UCM
The Configure JDBC Component Schema screen opens. In this screen, do the following (as shown in Figure 7-2):
Select UCM Schema.
Select Configure selected component schemas as RAC multi data source schemas in the next panel.
Click Next.
Figure 7-2 Configure JDBC Component Schema Screen for Oracle UCM
The Configure RAC Multi Data Sources Component Schema screen opens. In this screen, do the following (as shown in Figure 7-3):
Select UCM Schema. Leave the other data sources as they are.
Enter values for the following fields, specifying the connect information for the Oracle RAC database that was seeded with RCU:
Driver: Select Oracle driver (Thin) for RAC Service-Instance connections, Versions: 10 and later.
Service Name: Enter the service name of the database (ecmedg.mycompany.com).
Username: Enter the complete user name (including the prefix) for the schemas. The user names shown in Figure 7-3 assume that DEV
was used as the prefix for schema creation from RCU.
Password: Enter the password to use to access the schemas.
Click Add and enter the details for the first Oracle RAC instance.
Repeat step c for each Oracle RAC instance.
Note: Leave the SOA Infrastructure, IPM Schema, User Messaging Service, OWSM MDS Schema, and SOA MDS Schema information as they are. |
Click Next.
Figure 7-3 Configure RAC Multi Data Source Component Schema Screen for Oracle UCM
In the Test JDBC Data Sources screen, the connections should be tested automatically. The Status column displays the results. Ensure that all connections were successful. If not, click Previous to return to the previous screen and correct your entries.
Click Next when all the connections are successful.
In the Optional Configuration screen, select the following:
Managed Servers, Clusters and Machines
Deployment and Services
Click Next.
In the Configure Managed Servers screen, click Add to add the required managed servers as shown in Table 7-1. Do not modify the other servers that appear in this screen; leave them as they are.
Table 7-1 Managed Servers
Name | Listen Address | Listen Port | SSL Listen Port | SSL Enabled |
---|---|---|---|---|
WLS_UCM1 |
ECMHOST1 |
16200 |
n/a |
No |
WLS_UCM2 |
ECMHOST2 |
16200 |
n/a |
No |
Click Next.
In the Configure Clusters screen, click Add to add the clusters as shown in Table 7-2. Do not modify the other clusters that appear in this screen; leave them as they are.
Table 7-2 Clusters
Name | Cluster Messaging Mode | Multicast Address | Multicast Port | Cluster Address |
---|---|---|---|---|
UCM_Cluster |
unicast |
n/a |
n/a |
Leave empty |
Click Next.
In the Assign Servers to Clusters screen, add the following. Do not modify the other assignments that appear in this screen; leave them as they are.
UCM_Cluster:
WLS_UCM1
WLS_UCM2
Click Next.
In the Configure Machines screen, click the Unix Machine tab and add the following two new machines:
Leave all other fields to their default values. Click Next.
In the Assign Servers to Machines screen, assign servers to machines as follows:
Assign WLS_UCM1 to ECMHOST1.
Assign WLS_UCM2 to ECMHOST2.
Click Next.
In the Target Deployments to Clusters or Servers screen, make sure that the DMS Application is targeted to the SOA_Cluster, UCM_Cluster and Admin Server. Click Next.
In the Target Services to Clusters or Servers screen, make sure the JOC Startup and JOC Shutdown classes are targeted to the SOA_Cluster only. Click Next.
In the Configuration Summary screen, click Extend.
If you see a warning dialog about port conflicts for the domain, click OK.
In the Creating Domain screen, click Done.
Restart the Administration Server to make these changes to take effect. See Section 7.5, "Restarting the Administration Server."
Perform these steps to propagate the domain configuration:
Run the pack
command on SOAHOST1 to create a template pack using the following commands:
SOAHOST1> cd ORACLE_COMMON_HOME/common/bin SOAHOST1> ./pack.sh -managed=true -domain=ORACLE_BASE/admin/domain_name/aserver/domain_name -template=edgdomaintemplate2.jar -template_name=edgdomain_template
Copy the template to ECMHOST2:
Note: Assuming that ECMHOST1 shares the ORACLE_HOME with SOAHOST1, the template will be present in the same directory in ECMHOST1; otherwise, copy it also to ECMHOST1. |
SOAHOST1> scp edgdomaintemplate2.jar oracle@ECMHOST2:ORACLE_BASE/product/fmw/oracle_common/common/bin
Run the unpack
command on ECMHOST1 to unpack the propagated template.
Note: Make sure to rununpack from the ORACLE_COMMON_HOME/common/bin directory, not from WL_HOME/common/bin. |
ECMHOST1> cd ORACLE_COMMON_HOME/common/bin ECMHOST1> ./unpack.sh -domain=ORACLE_BASE/admin/domain_name/mserver/domain_name -template=edgdomaintemplate2.jar -app_dir=ORACLE_BASE/admin/domain_name/mserver/applications
Note: The ORACLE_BASE/admin/domain_name/mserver directory must exist before runningunpack . In addition, the ORACLE_BASE/admin/domain_name/mserver/applications must be empty. |
Repeat step 3 for ECMHOST2.
Perform these steps to start Node Manager on ECMHOST1 and ECMHOST2 if Node Manager has not started already:
On each server, run the setNMProps.sh script, which is located in the ORACLE_COMMON_HOME/common/bin directory, to set the StartScriptEnabled
property to 'true' before starting Node Manager:
ECMHOSTn> cd ORACLE_COMMON_HOME/common/bin ECMHOSTn> ./setNMProps.sh
Note: You must use theStartScriptEnabled property to avoid class loading failures and other problems. See also Section 12.8.3, "Incomplete Policy Migration After Failed Restart of SOA Server." |
Note: If the UCM server is sharing the MW_HOME in a local or shared storage with SOA, as suggested in the shared storage configuration described in Chapter 2, "Database and Environment Preconfiguration," it is not required to run setNMProps.sh again. In this case, Node Manager has already been configured to use a start script. |
Run the following commands on both ECMHOST1 and ECMHOST2 to start Node Manager:
ECMHOSTn> cd WL_HOME/server/bin ECMHOSTn> ./startNodeManager.sh
Restart the Administration Server to make these changes take effect. To restart the Administration Server, stop it first using the Administration Console and then start it again as described in Section 5.5, "Starting the Administration Server on SOAHOST1."
Starting the WLS_UCM1 Managed Server
Perform these steps to start the WLS_UCM1 managed server:
Start the WLS_UCM1 managed server using the Oracle WebLogic Server Administration Console as follows:
Expand the Environment node in the Domain Structure window.
Choose Servers. The Summary of Servers page opens.
Click the Control tab.
Select WLS_UCM1 and then click Start.
Verify that the server status is reported as "Running" in the Administration Console. If the server is shown as "Starting" or "Resuming," wait for the server status to change to "Started." If another status is reported (such as "Admin" or "Failed"), check the server output log files for errors. See Section 12.8, "Troubleshooting" for possible causes.
Configuring the WLS_UCM1 Managed Server
Perform these steps to configure the WLS_UCM1 managed server:
Log in to WLS_UCM1 at http://ECMHOST1:16200/cs using your Oracle WebLogic administration user name and password to display a configuration page.
Note: The UCM configuration files are on a shared disk so that all members of the cluster can access them. The shared disk location for the Oracle ECM enterprise deployment is at ORACLE_BASE/admin/ecm_domain/aserver/ucm_cluster. |
Change the following values on the server configuration page (make sure to select the "Is New Content Server Instance" check box to see all options):
Content Server Instance Folder: Set this to ORACLE_BASE/admin/ecm_domain/ucm_cluster/cs.
Native File Repository Location: Set this to ORACLE_BASE/admin/ecm_domain/ucm_cluster/cs/vault.
WebLayout Folder: Set this to ORACLE_BASE/admin/ecm_domain/ucm_cluster/cs/weblayout.
Server Socket Port: Set this to 4444.
Socket Connection Address Security Filter: Set this to a pipe-delimited list of localhost and the server IPs:
127.0.0.1|ECMHOST1|ECMHOST2|WEBHOST1|WEBHOST2
Note: This will be changed later to a host name-based list (see Section 8.13, "Adding the Oracle I/PM Server Listen Addresses to the List of Allowed Hosts in Oracle UCM"). At this point, we need the connections to be allowed for operations that will be done before the security filter is changed to host names. |
WebServer HTTP/HTTPS Address: Set this to 'ecm.mycompany.com:443'.
Web Address is HTTPS: Select this check box.
Server Instance Label: Set this to 'UCM_Cluster1'.
Server Instance Description: Set this to 'Cluster ucm_cluster1'.
Auto_Number Prefix: Set this to 'ucm_cluster1-'.
Click Submit when finished and restart the managed server using the Oracle WebLogic Server Administration Console.
The Oracle UCM server updates the cwallet.sso file located in ORACLE_BASE/admin/domain_name/mserver/domain_name/config/fmwconfig when it starts. This change needs to be propagated back to the Administration Server. To do this, copy the file to ORACLE_BASE/admin/domain_name/aserver/domain_name/config/fmwconfig in SOAHOST1 using the following command (all on a single line):
ECMHOST1> scp ORACLE_BASE/admin/domain_name/mserver/domain_name/config/fmwconfig/cwallet.sso oracle@SOAHOST1:ORACLE_BASE/admin/domain_name/aserver/domain_name/config/fmwconfig/
Note: If any operation is performed in the WLS_UCMn servers that modifies the cwallet.sso file in the ORACLE_BASE/admin/domain_name/mserver/domain_name/config/fmwconfig directory, the file will have to be immediately copied to the Administration Server domain directory on SOAHOST1 at ORACLE_BASE/admin/domain_name/aserver/domain_name/config/fmwconfig. |
Starting the WLS_UCM2 Managed Server
Perform these steps to start the WLS_UCM2 managed server:
Start the WLS_UCM2 managed server using the Oracle WebLogic Server Administration Console as follows:
Expand the Environment node in the Domain Structure window.
Choose Servers. The Summary of Servers page opens.
Click the Control tab.
Select WLS_UCM2 and then click Start.
Verify that the server status is reported as "Running" in the Administration Console. If the server is shown as "Starting" or "Resuming," wait for the server status to change to "Started." If another status is reported (such as "Admin" or "Failed"), check the server output log files for errors. See Section 12.8, "Troubleshooting" for possible causes.
Configuring the WLS_UCM2 Managed Server
Perform these steps to configure the WLS_UCM2 managed server:
Log in to WLS_UCM2 at http://ECMHOST2:16200/cs using your Oracle WebLogic administration user name and password to display a configuration page.
Note: The UCM configuration files are on a shared disk so that all members of the cluster can access them. The shared disk location for the Oracle ECM enterprise deployment is at ORACLE_BASE/admin/ecm_domain/aserver/ucm_cluster. |
Change the following values on the server configuration page:
Content Server Instance Folder: Set this to ORACLE_BASE/admin/ecm_domain/ucm_cluster/cs
Native File Repository Location: Set this to ORACLE_BASE/admin/ecm_domain/ucm_cluster/cs/vault
WebLayout Folder: Set this to ORACLE_BASE/admin/ecm_domain/ucm_cluster/cs/weblayout
Make sure that the 'Is new Content Server Instance?' check box is not selected.
Click Submit when finished and restart the managed server using the Oracle WebLogic Server Administration Console.
The following parameter should be set in Oracle Content Server's config.cfg file in order to enable login retries during an Oracle RAC failover:
ServiceAllowRetry=true
If this value is not set, users will need to manually retry any operation that was in progress when the failover began.
Perform these steps to add the configuration parameter for Oracle UCM:
Go to the WebLogic Server Administration Console for Oracle UCM at http://ECMHOST1:16200/cs, and log in using your Oracle WebLogic administration user name and password.
Open the Administration page, and then choose Admin Server. The Content Admin Server page opens.
Click General Configuration on the left. The General Configuration page opens.
In the Additional Configuration Variables box, add the parameter:
ServiceAllowRetry=true
Click Save and restart all UCM managed servers.
Note: The new parameter is included in the config.cfg file, which is at the following location:
ORACLE_BASE/admin/ecm_domain/ucm_cluster/cs/config/config.cfg
(You can also edit this file directly in a text editor. Do not forget to restart all UCM managed servers.) |
To enable Oracle HTTP Server to route to UCM_Cluster, which contain the WLS_UCM1 and WLS_UCM2 managed servers, you must set the WebLogicCluster
parameter to the list of nodes in the cluster:
On WEBHOST1 and WEBHOST2, add the following lines to the ORACLE_BASE/admin/instance_name/config/OHS/component_name/mod_wl_ohs.conf file:
# UCM <Location /cs> WebLogicCluster ECMHOST1:16200,ECMHOST2:16200 SetHandler weblogic-handler WLCookieName JSESSIONID WLProxySSL ON WLProxySSLPassThrough ON </Location> <Location /adfAuthentication> WebLogicCluster ECMHOST1:16200,ECMHOST2:16200 SetHandler weblogic-handler WLCookieName JSESSIONID WLProxySSL ON WLProxySSLPassThrough ON </Location> <Location /_ocsh> WebLogicCluster ECMHOST1:16200,ECMHOST2:16200 SetHandler weblogic-handler WLCookieName JSESSIONID WLProxySSL ON WLProxySSLPassThrough ON </Location>
Restart Oracle HTTP Server on both WEBHOST1 and WEBHOST2.
WEBHOST1> ORACLE_BASE/admin/instance_name/bin/opmnctl restartproc ias-component=ohs1 WEBHOST2> ORACLE_BASE/admin/instance_name/bin/opmnctl restartproc ias-component=ohs2
You should verify URLs to ensure that appropriate routing and failover is working from Oracle HTTP Server to UCM_Cluster. Perform these steps to verify the URLs:
While WLS_UCM2 is running, stop WLS_UCM1 using the Oracle WebLogic Server Administration Console.
Access http://WEBHOST1:7777/cs to verify it is functioning properly.
Start WLS_UCM1 from the Oracle WebLogic Server Administration Console.
Stop WLS_UCM2 from the Oracle WebLogic Server Administration Console.
Access http://WEBHOST1:7777/cs to verify it is functioning properly.
Oracle recommends using host name verification for the communication between Node Manager and the servers in the domain. This requires the use of certificates for the different addresses communicating with the Administration Server and other servers. See Chapter 9, "Setting Up Node Manager" for further details. The procedures in that chapter must be performed twice using the following information:
Run | Host Name (HOST) | Virtual IP (VIP) | Server Name (WLS_SERVER) |
---|---|---|---|
Run 1: | ECMHOST1 | ECMHOST1VHN1* | WLS_UCM1 |
Run 2: | ECMHOST2 | ECMHOST2VHN1* | WLS_UCM2 |
* Optional; required only for WLS-managed Oracle I/PM servers (WLS_IPM).
Please note the following:
Even though the WLS_IPM managed servers are not yet configured at this point (see Chapter 8, "Extending the Domain with Oracle I/PM") and are not mandatory for a UCM-only Oracle ECM installation, the virtual host names used by Oracle I/PM are configured here in order to provide a one-step configuration process that includes both types of servers.
Perform all steps in Chapter 9, "Setting Up Node Manager" except for Section 9.3.5, "Configuring Managed WLS Servers to Use the Custom Keystores" for the WLS_IPM servers. This step can be done after the WLS_IPM servers are added to the domain.
After you have verified that the extended domain is working, back up the installation. This is a quick backup for the express purpose of immediate restore in case of problems in the further steps. The backup destination is the local disk. This backup can be discarded once the enterprise deployment setup is complete. At that point, the regular deployment-specific backup and recovery process can be initiated. The Oracle Fusion Middleware Administrator's Guide provides further details. For information on describing the Oracle HTTP Server data that must be backed up and restored, refer to the "Backup and Recovery Recommendations for Oracle HTTP Server" section in that guide. For information on how to recover components, see the "Recovery of Components" and "Recovery After Loss of Component" sections in the guide. For recommendations specific to recovering from the loss of a host, see the "Recovering Oracle HTTP Server to a Different Host" section in the guide. Also refer to the Oracle Database Backup and Recovery Guide for information on database backup.
Perform these steps to back up the installation at this point:
Back up the web tier:
Shut down the instance using opmnctl
.
WEBHOST1> ORACLE_BASE/admin/instance_name/bin/opmnctl stopall
Back up the Middleware Home on the web tier using the following command (as root):
WEBHOST1> tar -cvpf BACKUP_LOCATION/web.tar MW_HOME
Back up the Oracle Instance Home on the web tier using the following command:
WEBHOST1> tar -cvpf BACKUP_LOCATION/web_instance_name.tar ORACLE_INSTANCE
Start the instance using opmnctl
:
WEBHOST1> cd ORACLE_BASE/admin/instance_name/bin WEBHOST1> opmnctl startall
Back up the database. This is a full database backup (either hot or cold) using Oracle Recovery Manager (recommended) or operating system tools such as tar for cold backups if possible.
Back up the Administration Server and managed server domain directory to save your domain configuration. The configuration files all exist in the ORACLE_BASE/admin/domain_name directory. Run the following command to create the backup:
SOAHOST1> tar -cvpf edgdomainback.tar ORACLE_BASE/admin/domain_name