Known issues for the combined cumulative fix | HCL Digital Experience 8.5

Learn what are the known issues in the HCL Digital Experience combined cumulative fixes.

This page will contain the known issues for the HCL Portal Version 8.5 Combined Cumulative Fix 08 or later releases, and also for Version 9.0 beginning with CF14, separated by CF level.

Known Issues for HCL Portal Version 8.5 and Version 9.0 Combined CF16 (PI91150)

  • Problem: As of HCL Portal Version 8.5 and 9.0 CF16, some features are not visible to users assigned to the Anonymous user role and the All authenticated users role, such as the tree view in Site Manager.
  • Solution: Add the necessary roles for user Anonymous and user group All Authenticated Users explicitly. For example, to use site manager features within the toolbar, you must assign the User role to the All Authenticated Users group for the Site Manager portlet.

  • Problem: The applyCF task may fail on iSeries when used in combination with certain WAS levels, potentially with WAS 8.5.5.13 or greater. An error similar to the following may be seen in the failure*.log in the ConfigEngine 'log' directory:
    /QIBM/ProdData/WebSphere/PortalServer/V85/ConfigEngine/config/includes/apply_
    PTF.xml:172: /QIBM/ProdData/WebSphere/PortalServer/V85/Server/installer/wp.ie
    hs.ear/config/includes/wp.iehs.ear_cfg.xml:62
  • Solution: This may be due to an OSGi issue stemming from a WAS level upgrade. Run the following on the iSeries system:
    cd /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/bin
    osgiCfgInit -all
    Then rerun the applyCF task.

  • Problem: The HCL Portal Version 8.5 and 9.0 CF15 applyCF task may fail if using the specific combination of WAS 8.5.5.12 on a Windows OS. An error similar to the following may be seen in the failure*.log in the (profile_root)/ConfigEngine/log directory:
    wp.theme.modules.webapp_cfg.xml:296: java.lang.NoClassDefFoundError: com.ibm.ffdc.config.Formattable
  • Solution: This is due to an OSGi issue in WAS 8.5.5.12 that will be fixed in their next release. Simply rerunning the applyCF task should allow continuation and success of the CF15 update. The following command also remediates the issue:
    (profile_root)/bin/osgiCfgInit.sh|bat

  • Problem: A failure, similar to the following, may be seen when running a Rollback on a system configured with WAS 8.5.5.x with JDK8:
    /opt/IBM/WebSphere/PortalServer/toolbar/wp.toolbar.themes/toolbar85/config/includes/wp.theme.themes.toolbar85_cfg.xml:217: The 
    following error occurred while executing this line:
    /opt/IBM/WebSphere/ConfigEngine/config/includes/utils.xml:2093: EJPXB0019E: Server response indicates an error. For status and 
    details of the XMLAccess error look at file /opt/IBM/WebSphere/wp_profile/ConfigEngine/config/work/xmlaccess.temp.file.657596238.xml.
  • Solution: This is due to an EJBDeploy incompatibility with JDK8 on WAS 8.5.5.x. The following steps can be used to bypass the issue: After rollback failure, use the following command to switch the JDK used by the profile to 1.7.

    For example:

    /opt/IBM/WebSphere/wp_profile/bin/managesdk.sh -enableProfileAll -sdkName 1.7_64
    Then rerun the failing rollbackCF command, from the start by removing the rollbackCFprogress directory in, example, (profile_root)/ConfigEngine/log. After rollback success, the JDK can then be switched back to the original JDK8 via the same command and substituting -sdkName 1.8_64. This is planned to be fixed in the next Portal CF.
  • Problem: The following error may be seen when working with or installing Portal Web Modules, when configured with WAS 8.5.5.12 / 9.0.0.5 generates this error:
    com.ibm.portal.tasks.wtf.XmlAccessException: Xmlaccess request failed with an error. Result [failed], 
    Messages: 
    [[EJPXA0043E: EJPXA0043E: An error occurred while creating or updating the resource.], 
    [EJPPD0015E: EJPPD0015E: Portlet application manager failed when user xmlaccess scripting user executed command InstallWebApplication.], 
    [EJPPE0013E: EJPPE0013E: The installation of Web Module from WAR file /opt/IBM/WebSphere/wp_profile/PortalServer/deployed/ilwwcm-wcmlibutil.war did not complete successfully when invoking the WebSphere Application Server administration interface.], 
    [EJPPE0001E: EJPPE0001E: Instantiation of the WebSphere Application Server administration manager failed.], 
    [EJPPH0007E: EJPPH0007E: Instantiation of the administration module for Portal context failed.], 
    [EJPPH0010E: EJPPH0010E: Failed to retrieve and initialize administration runtime variables from Application Server.],
  • Solution: This is expected to be fixed in an upcoming WAS level. The temporary workaround to alleviate this problem is to set the following JVM Property:
    com.ibm.ws.management.connector.soap.disableSOAPAuthCheck = true
  • Problem: Errors may be seen in DCS (Document Conversion Service) based on Oracle OIT (Outside In Technology) levels. The new OIT 8.5.3 level has additional prerequisites outside of Portal's scope.
  • Solution: Go and log in to HCL Software Support page to find related information about the PI84847 introduced in Portal Version 8.5 and Version 9.0 CF15 section.
  • Problem: On a Linux or Unix system, it is possible that the applyCF command may fail, with this error message recorded in ConfigTrace.log:
    Couldn't flush user prefs: java.util.prefs.BackingStoreException: Couldn't get file lock.
  • Solution: The problem concerns Java system and/or user preferences and can potentially be addressed in one of the following ways:
    1. Make sure that Java's preferences directory is accessible. Ensure that the /etc/.java/.systemPrefs directory exists and is read/writable by the user performing the upgrade as well as any alternate user ID used to do WAS administration.
    2. Create a new unique Java preferences path for use by WAS.
      • a. Create a new sub-directory of your choice (for example, under /home/wasadmin/prefs/).
      • b. Assign permissions on this directory so that it is read/writable by the user performing the upgrade as well as any alternate user ID used to do WAS administration.
      • c. Set the following argument on the affected WebSphere JVM: -Djava.util.prefs.userRoot=/home/wasadmin/prefs. To set this argument, go to the administrative console, click Servers > Server Types > WebSphere application servers, and select the server that you want to add the generic argument to. Then, in the Server Infrastructure section, click Java and Process management > Process definition > Java virtual machine.
      • d. Add -Djava.util.prefs.userRoot=/home/wasadmin/prefs to any existing properties in the Generic JVM Arguments field and click OK. e. Then click Save to save your changes and restart the application server.
  • Problem: An error may occur during applyCF on the primary cluster node, similar to the following:
    [wplc-remove-ear] AppName EphoxTbioServices 
    [wplc-remove-ear] ADMA5017I: Uninstallation of EphoxTbioServices started. 
    [wplc-remove-ear] ADMA5104I: The server index entry for WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin11+WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin09+WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin10Node01 is updated successfully. 
    [wplc-remove-ear] ADMA5102I: The configuration data for EphoxTbioServices from the configuration repository is deleted successfully. 
    [wplc-remove-ear] ADMA5011I: The cleanup of the temp directory for application EphoxTbioServices is complete. 
    [wplc-remove-ear] ADMA5106I: Application EphoxTbioServices uninstalled successfully. 
    [wplc-remove-ear] WASX7015E: Exception running command: "source "C:/ibm/wp85/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; exception information: 
    [wplc-remove-ear] com.ibm.websphere.management.exception.ConfigServiceException 
    [wplc-remove-ear] com.ibm.ws.sm.workspace.WorkSpaceException 
    [wplc-remove-ear] com.ibm.websphere.management.exception.DocumentIOException: cells/psvtmigwin08Cell01/applications/EphoxTbioServices.ear/EphoxTbioServices.ear
  • Solution: Go and log in to HCL Software Support page to find related information about disabling auto plugin generation. Once applyCF completes, re-enable the two plugin settings.
  • Problem: The post-was-configuration-update task may fail due to an error in the Portal server startup. SystemOut.log shows an exception related to the wps_scheduler, and the wps_scheduler does not start when the Portal server starts. Error logs may appear similar to the following:
    [6/21/17 23:05:06:112 EDT] 00000080 InternalGetSc E com.ibm.wps.command.scheduler.internal.InternalGetSchedulerTaskCommand AbstractCommand.throwCommandException EJPDD0009E: JNDI naming lookup failed for name = [ejb/wpsSchedulerManager].
    javax.naming.NameNotFoundException: Context: w7944aCell/nodes/w7944aNode/servers/HCL Portal and HCL Web Content Manager, 
    name: ejb/wpsSchedulerManager: First component in name wpsSchedulerManager not found. 
    [Root exception is org.omg.CosNaming.NamingContextPackage.NotFound: IDL:omg.org/CosNaming/NamingContext/NotFound:1.0]
  • Solution: Edit the WAS 9 properties file, for example AppServer/properties/migration.properties.
    #-------------------------------------------------------------------------
    # Specify the number of apps which can be installed in parallel.  default=3
    #-------------------------------------------------------------------------
    #com.ibm.ws.migration.application.install.threadPoolSize=3
    Then uncomment the property and set to a value of 1. For example:
    com.ibm.ws.migration.application.install.threadPoolSize=1
    Then, save the change and continue with the profile migration. Stop the Portal server, and run the following command again:
    post-was-configuration-update

  • Problem: When creating a portal page, there is an option to Create Content. When clicking on that option, it presents available templates to create content from, Content Hub Authoring Template is one of them. If selecting this template, it may show an error:
    "The new content item could not be created. The selected template is not supported. A workflow is required, but the workflow is not 
    set for the item"
  • Solution: This authoring template is not designed for this feature. It has to be exempted from creating content items off it from toolbar. To do that:
    1. Go to the Content Hub library.
    2. Open Authoring templates, edit the Content Hub authoring template.
    3. Go to the tab Properties.
    4. Open section Profile.
    5. Add this keyword hcl.portal.toolbar.hiddenTemplate to the keywords text box.
    6. Save and close.

  • Problem: XMLAccess in the applyCF command may fail when updating a virtual portal.
  • Solution: The CF install attempts to update the Portal owned pages in the base portal and all virtual portals. Prior to CF11 there were some cases where error conditions were suppressed for both the base portal and all virtual portals. This has changed and error conditions are no longer suppressed. If the CF install (applyCF command) reports an error while running the XMLAccess tool on a virtual portal, there may be a problem with the virtual portal. If virtual portal(s) have been modified and removed some of the Portal owned pages on purpose then these errors can be safely suppressed and allow the CF install to continue. This can be done by adding the ignore-errors-all-vps property to the applyCF command.

    For example:

    applyCF.(sh|bat) ..... -Dignore-errors-all-vps='true'
    If Portal owned pages have not been purposely removed from the virtual portal, then the error should be investigated and attempted to be fixed.

  • Problem: After HCL Portal Version 8.5 CF12 or later is applied, warning messages may be seen in the SystemOut.log similar to the following, and custom themes or custom portlets may stop working:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] does not specify a blackwhite list when accessing resource [...], 
    falling back to the default [[whitelist(null), blacklist(.*)]]...
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF12 (and later) contains a security fix (PI65954). Similar to APAR PI47714, this APAR warns and then blocks access to files in a web application that does not have a blacklist or whitelist. The difference between PI47714 and PI65954 is that the default value for the blacklist is now ".*" instead of "WEB-INF/.*" which is much more restrictive than before. Applications can define a custom blacklist or whitelist list by adding the keys com.ibm.portal.resource.whitelist and com.ibm.portal.resource.blacklist to their web.xml deployment descriptor. Custom themes must be redeployed for these changes to take effect. Alternatively, as a temporary solution, you can define a different default value for the blacklist. Details on the action required for the custom code mentioned in these messages and the setting of a default value for the blacklist can be found on the HCL Software Support page.

  • Problem: If you drag and drop digital assets into:
    1. The Dropzone in the Sharing Tab of Rich Media Edition console, or;
    2. Shared files directory (MBAssets folder and its subfolders) that the Rich Media Edition uses
    then default renditions may not be created automatically.
  • Solution: One of the two workaround approaches can be followed to place assets into Rich Media Edition and generate default renditions.
    1. You can either use WCM interface to place assets in Rich Media Edition instead of Dropzone or shared files directory of the Rich Media Edition server.
    2. Launch the link from Shared files section in the Sharing tab of Rich Media Edition console in a browser to open the user's workspace. Login and double-click the assets for which you want to generate renditions and open Renditions tab. Click on Regenerate Renditions button.
  • Problem: If HCL Portal is configured with an empty value for the wps context root or an empty value for the portal mapping for public pages then the following message may be seen in the browser when a user accesses a page that has a whitespace character or any other special character in the friendly page name:
    Error 400: EJBCD0006E: The resolution of a URI failed. Refer to the SystemOut.log for more detailed information.
    A similar warning in the SystemOut.log may also be seen:
    ExceptionLogg W com.ibm.wps.logging.ExceptionLogger logThrowable An exception occurred: [Illegal character in path at index 43: com.ibm.portal.friendly.name:/Search/Search Center]. Enable traces for [com.ibm.wps.logging.ExceptionLogger=all] to see the exception stack trace.
  • Solution: This is a known issue and is fixed by applying the WAS interim fix for PI67942 for WAS 8.5.5.10. This issue will also be fixed or included in WAS 8.5.5.11. HCL Software page in the maintenance package.

  • Problem: A decrease in performance may occur when Textbox.io is used with large amounts of HTML (~4000 lines/20,000 characters) that can cause the Firefox browser to stop working when switching between code and design views.
  • Solution: When switching between code and design views, save and close the content and then reopen it before returning to the design view. For other browsers, waiting and allowing processes to continue running eventually returns to the design view and no further action is required.

  • Problem: If WAS 8.5.5.10 is installed with JDK8 for the Portal server profile and CF12 or later is installed, the following errors may be found in the ConfigTrace.log file: \
    [exec] SEVERE:
    [exec] java.lang.UnsupportedClassVersionError: JVMCFRE003 bad major version; class=sun/rmi/rmic/Main, offset=6
    and
    [exec] SEVERE: Error generating RMI code: Cannot run RMIC because it is not installed. Expected location of RMIC is the 
    following: /opt/IBM/WebSphere/AppServer/java_1.8_64/lib.
    [exec] Exception in thread "main" java/lang/NullPointerException
  • Solution: There is a problem communicating with RMIC while preparing an Enterprise Application for installation. This is fixed by installing WAS APAR PI67166 and retrying the applyCF.sh command. It can also be avoided by installing WAS APAR PI67166 before the Portal CF update.

  • Problem: On iSeries environments configured with JDK8 and WAS 8.5.5.10, you may get the error "Java runtime version not supported" in ConfigTrace.log during the installation of Portal CF12 or later. In more detail, an error similar to the following may be seen in the ConfigTrace.log file:
    [exec] Java runtime version not supported
    [exec] 1.8.0
    [exec] Java runtime version not supported
    [exec] com.ibm.etools.ejbdeploy.EJBDeploymentException: 1.8.0
    [exec] Java runtime version not supported
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.ValidateJavaRTenv(EJBDeploy.java:475)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.execute(EJBDeploy.java:74)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.deploy(EJBDeploy.java:349)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.main(EJBDeploy.java:310)
    [exec] EJBDeploy level:
    --- Exception Thrown ---
    /QIBM/ProdData/WebSphere/PortalServer/V85/Server/wcm/prereq.wcm/config/includes/prereq.wcm.fp_cfg.xml:121: exec returned: 1
  • Solution: To use JDK8 with WAS 8.5.5.10 on IBM i, it will be necessary to first run a script to patch the ejbdeploy file. This script should be run after the IM install step and before the applyCF manual step. If this error has already been encountered during the applyCF command, then the script can be run, and then re-run the applyCF command. To run the script, open a command prompt in the (profile_root)/PortalServer/bin/ directory then run the command:
    i5ejbscript update
    To revert to the old state, running the following command will restore the link to the original file:
    i5ejbscript restore
    Note: WAS 8.5.5.11 should contain the fix for the ejbdeploy file, and this action will no longer be needed.

  • Problem: ISAM v7 does not support JDK8. Any customer choosing to move to JDK8 will no longer have compatibility for ISAM v7. ISAM v9 will support JDK8, but Portal does not yet support this level.
  • Solution: This is currently a Portal limitation. Future plans may include ISAM v9 support in a later CF.

  • Problem: No content indexed in default search collection and system was configured with a custom Portal context root
  • Solution: Prior to CF06 there was a manual step required after changing the context root. The context in the search content source had to be manually adjusted. If this was not performed, then the content source will be using the wrong context root and will not find any content to index. Beginning with CF06 the update of the search content source was included in the task modify-servlet-path. To solve this problem, the context in the search content source can be adjusted manually or just run the task modify-servlet-path again.

  • Problem: In the WCM Authoring environment, when the DAM popup is opened to select an image, it may show all the assets available in DAM, but the thumbnails will not be generated properly. This may be due to a particular library which is installed with Linux, but is at a different version to that which MediaBeacon utilizes.
  • Solution: The typical solution is to create a symbolic link, as shown below.
    1. Open Terminal.
    2. Run the following command:
      cd /lib64
    3. If Linux is installed with libbz2.so.1.0.4 for example, then do:
      ln -s libbz2.so.1.0.4 libbz2.so.1.0
  • Problem: In installations with a large number of resources, it is possible that the applyCF script could run out of memory and fail. Such a failure would record a "java.lang.OutOfMemoryError: Java heap space" message and a Java stack trace in the ConfigTrace.log file.
  • Solution: If this problem is encountered, the amount of memory available to the JVM can be increased, and the configuration can then be re-run.
    1. For Unix/Linux:
      1. Backup the (AppServer_home)/bin/wsadmin.sh file.
      2. Open this file in a text editor.
      3. Locate the PERF_JVM_OPTIONS= line specific to your operating system.
      4. Modify this line to have an Xmx parameter of 2048, e.g. from
        PERF_JVM_OPTIONS="-Xms256m -Xmx256m -Xquickstart" ;;
        to
        PERF_JVM_OPTIONS="-Xms256m -Xmx2048m -Xquickstart" ;;
      5. Save changes.
      6. Re-run applyCF.sh. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
    2. For Windows:
      1. Backup the AppServer_home)\bin\wsadmin.bat file.
      2. Locate the set PERFJAVAOPTION statement.
      3. Modify this line to have an Xmx parameter of 2048, e.g.
        set PERFJAVAOPTION=-Xms256m -Xmx256m -Xquickstart
        to
        set PERFJAVAOPTION=-Xms256m -Xmx2048m -Xquickstart
      4. Save changes.
      5. Re-run applyCF.bat. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.

  • Problem: Where Textbox.io is the configured rich text editor in WCM, rapidly opening and re-saving an existing item containing a rich text element can cause the loss of existing content if the item is saved before the rich text content is fully loaded. This is only likely to be an issue in environments with high network latency.
  • Solution: Ensure when editing an existing content, site area, template or component containing a rich text element, the rich text content is fully loaded in the editor before re-saving the item.

  • Problem: During the migration process the node may not start after upgrading to WAS 8.5.5.9 and before executing upgrade-profile. The node does not start with error message:
    ADMU3200I: Server launched. Waiting for initialization status.
    ADMU3011E: Server launched but failed initialization. Server logs, startServer.log, and other log files
    The log contains the following error message:
    HMGR0031E: A conflicting IP address and port has been detected for the DCS_UNICAST_ADDRESS end point.
    The node does not start due to conflicting port configuration for the DCS_UNICAST_ADDRESS end point of server1 and nodeagent in the node 
    serverindex.xml.
  • Solution: It is recommended to execute upgrade-profile before upgrading to WAS 8.5.5.9. Alternatively the conflicting port configuration for the DCS_UNICAST_ADDRESS needs to be solved by setting unique port numbers for server1 and nodeagent.

  • Problem: After Rollback from CF09 or above to CF08 or below, the portal toolbar may show extra items that have titles $Banner$, $Teaser$ and $Video$
  • Solution: These items can be ignored and should not be used. During an update to CF09 or above again, these items will disappear. To have these items no longer visible in the toolbar, go to Applications > Content > Web Content Authoring. Search for each of these pieces of content. Edit each of the pieces of content and go to properties. In the profile keyword field, remove the keyword hcl.portal.toolbar.NewContent. Save this change.

  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too 
    long on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path. Go and log in to HCL Software Support page to find instructions in changing the location of the wstemp temporary workspace directory.

  • Problem: During the upgrade-profile task in migration, there may be an intermittent issue which can cause the uninstall of an application to fail because files from the application are locked and cannot be deleted. This is an example of an error message that may be seen:
    BUILD FAILED
    C:\ibm\wp85\ConfigEngine\config\includes\apply_PTF.xml:172: 
    C:\ibm\wp85\PortalServer\jcr\wp.content.repository.install\config\includes\jcr.mig_cfg.xml:382: Exception found when executing wsadmin: <hostname>
    c:/ibm/wp85/ConfigEngine/lib/wkplc.misc.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batchboot.jar;C:/ibm/wp85/AppServer/deploytool/itp/batch2.jar;
    C:/ibm/wp85/PortalServer/base/wp.base/shared/app/wp.base.jar;C:/ibm/wp85/PortalServer/shared/app/wp.base.jar
  • Solution: If this issue is encountered, wait 5 minutes and re-run the upgrade-profile task with the same options to resume the migration. The lock on the file should be released and the uninstall can continue as expected.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (IN STANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page. This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of the CF the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page. This is a permanent change in behavior effective in CF08 or later. Go to Creating page templates topic in the HCL Digital Experience Version 8.5 documentation for more information about this behavior.

  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site or Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: Exception found when executing wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'
    Ear File C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear
    WASX7015E: Exception running command: "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; 
    exception information: com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file.

    For Linux, AIX, Solaris, the hosts file is:

    /etc/hosts
    For Windows, the hosts file is:
    %SystemRoot%\System32\drivers\etc\hosts
    Add a line similar to the following, substituting your actual machine name where indicated like
    127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
    If a line for localhost already exists, you can add your hostname to this existing line. To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.

  • Problem: After the CF is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) 
    Servlet context [/...] does not specify a blackwhite list when accessing resource [...], falling back to the default 
    [[(null), (WEB-INF/.*)]]. Applications can define a custom list by adding the keys [com.ibm.portal.resource.whitelist] and 
    [com.ibm.portal.resource.blacklist] to their web.xml deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 (or later) contained a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the 'res' datasource. You must redeploy your custom theme for these changes to take effect. Details on the action required for the custom code mentioned in these messages can be found in the HCL Software Support page.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the JSP components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting the following link on their support site: https://ephox.zendesk.com/entries/96347398-Java-1-8-u60-Internet-Explorer-Issue.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.io. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it

  • Problem: Error messages can be seen in the SystemOut.log after rollback regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190)
  • Solution: These logging errors are expected if content was created against a Portal feature at an updated CF level and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables

  • Problem: Starting with HCL Portal Version 8.5 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title) 
    (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal Version 8.5 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal Version 8.5 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal Version 8.5 Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: For IBM i only: May have the following error when running applyCF.sh when Portal databases have been transferred to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database directory.
  • Solution: Run the following steps then rerun applyCF.sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed.
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Relational database alias: The hostname. Use the short TCP/IP hostname of the database system.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches:
         [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh
    --- Exception Thrown ---
    /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named EJBDeploy tool for pre-EJB 3.0 modules.

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing HCL Portal Version 8.5 GA and before updating to the current CF, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply HCL Portal Version 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall HCL Portal Version 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall HCL Portal Version 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 or later and retry the rollback. For more information, go to the HCL Software Support page.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback.
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.

Known Issues for V8.5.0.0 / V9.0 Combined CF15 (PI83476)

  • Problem: The HCL Portal Versions 8.5 and 9.0 CF15 applyCF task may fail if using the specific combination of WAS 8.5.5.12 on a Windows OS. An error similar to the following may be seen in the failure*.log in the (profile_root)/ConfigEngine/log directory:
    wp.theme.modules.webapp_cfg.xml:296: java.lang.NoClassDefFoundError: com.ibm.ffdc.config.Formattable
  • Solution: This is due to an OSGi issue in WAS 8.5.5.12 that will be fixed in their next release. Simply rerunning the applyCF task should allow continuation and success of the CF15 update. The following command also remediates the issue:
    (profile_root)/bin/osgiCfgInit.sh|bat

  • Problem: A failure, similar to the following, may be seen when running a Rollback on a system configured with WAS 8.5.5.x with JDK8:
    /opt/IBM/WebSphere/PortalServer/toolbar/wp.toolbar.themes/toolbar85/config/includes/wp.theme.themes.toolbar85_cfg.xml:217: The 
    following error occurred while executing this line:
    /opt/IBM/WebSphere/ConfigEngine/config/includes/utils.xml:2093: EJPXB0019E: Server response indicates an error. For status and 
    details of the XMLAccess error look at file /opt/IBM/WebSphere/wp_profile/ConfigEngine/config/work/xmlaccess.temp.file.657596238.xml.
  • Solution: This is due to an EJBDeploy incompatibility with JDK8 on WAS 8.5.5.x The following steps can be used to bypass the issue:
    1. After rollback failure, use the following command to switch the JDK used by the profile to 1.7: Ex.
      /opt/IBM/WebSphere/wp_profile/bin/managesdk.sh -enableProfileAll -sdkName 1.7_64
    2. Then rerun the failing rollbackCF command from the start by removing the rollbackCFprogress directory in, for example (profile_root>)ConfigEngine/log.
    3. After rollback success, the JDK can then be switched back to the original JDK8 via the same command and substituting -sdkName 1.8_64 This is planned to be fixed in the next Portal CF.
  • Problem: The following error may be seen when working with / installing Portal Web Modules, when configured with WAS 8.5.5.12 / 9.0.0.5:
    com.ibm.portal.tasks.wtf.XmlAccessException: Xmlaccess request failed with an error. Result [failed], 
    Messages: 
    [[EJPXA0043E: EJPXA0043E: An error occurred while creating or updating the resource.], 
    [EJPPD0015E: EJPPD0015E: Portlet application manager failed when user xmlaccess scripting user executed command InstallWebApplication.], 
    [EJPPE0013E: EJPPE0013E: The installation of Web Module from WAR file /opt/IBM/WebSphere/wp_profile/PortalServer/deployed/ilwwcm-wcmlibutil.war did not complete successfully when invoking the WebSphere Application Server administration interface.], 
    [EJPPE0001E: EJPPE0001E: Instantiation of the WebSphere Application Server administration manager failed.], 
    [EJPPH0007E: EJPPH0007E: Instantiation of the administration module for Portal context failed.], 
    [EJPPH0010E: EJPPH0010E: Failed to retrieve and initialize administration runtime variables from Application Server.]
  • Solution: This is expected to be fixed in an upcoming WAS level. The temporary workaround to alleviate this problem is to set the following JVM Property:
    com.ibm.ws.management.connector.soap.disableSOAPAuthCheck = true
  • Problem: Errors may be seen in DCS (Document Conversion Service) based on Oracle OIT (Outside In Technology) levels. The new OIT 8.5.3 level has additional prerequisites outside of HCL Portal's scope.
  • Solution: For more information related to the PI84847, go to the HCL Software Support page.

  • Problem: On a Linux or Unix system, it is possible that the applyCF command may fail, with this error message recorded in ConfigTrace.log:
    Couldn't flush user prefs: java.util.prefs.BackingStoreException: Couldn't get file lock.
  • Solution: The problem concerns Java system and/or user preferences and can potentially be addressed in one of the following ways.
    1. Make sure that Java's preferences directory is accessible. Ensure that the /etc/.java/.systemPrefs directory exists and is read/writable by the user performing the upgrade as well as any alternate user ID used to do WAS administration.
    2. Create a new unique Java preferences path for use by WAS.
      • a. Create a new sub-directory of your choice (for example, under /home/wasadmin/prefs/).
      • b. Assign permissions on this directory so that it is read/writable by the user performing the upgrade as well as any alternate user ID used to do WAS administration.
      • c. Set the following argument on the affected WebSphere JVM: -Djava.util.prefs.userRoot=/home/wasadmin/prefs. To set this argument, go to the administrative console, click Servers > Server Types > WebSphere application servers, and select the server that you want to add the generic argument to. Then, in the Server Infrastructure section, click Java and process management > Process definition > Java virtual machine.
      • d. Add -Djava.util.prefs.userRoot=/home/wasadmin/prefs to any existing properties in the Generic JVM Arguments field and click OK.
      • e. Then click Save to save your changes and restart the application server
  • Problem: An error may occur during applyCF on the primary cluster node, similar to the following:
    [wplc-remove-ear] AppName EphoxTbioServices [wplc-remove-ear] ADMA5017I: Uninstallation of EphoxTbioServices started. 
    [wplc-remove-ear] ADMA5104I: The server index entry for WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin11+WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin09+WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin10Node01 is updated successfully. 
    [wplc-remove-ear] ADMA5102I: The configuration data for EphoxTbioServices from the configuration repository is deleted successfully. 
    [wplc-remove-ear] ADMA5011I: The cleanup of the temp directory for application EphoxTbioServices is complete. 
    [wplc-remove-ear] ADMA5106I: Application EphoxTbioServices uninstalled successfully. 
    [wplc-remove-ear] WASX7015E: Exception running command: "source "C:/ibm/wp85/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; exception information: 
    [wplc-remove-ear] com.ibm.websphere.management.exception.ConfigServiceException [wplc-remove-ear] com.ibm.ws.sm.workspace.WorkSpaceException 
    [wplc-remove-ear] com.ibm.websphere.management.exception.DocumentIOException: cells/psvtmigwin08Cell01/applications/EphoxTbioServices.ear/EphoxTbioServices.ear
  • Solution: Disable auto plugin generation. After applyCF completes, re-enable the two plugin settings. For instructions on disabling auto plugin generation, go to the HCL Software Support page.

  • Problem: The post-was-configuration-update task may fail due to an error in the Portal server startup. SystemOut.log shows an exception related to the wps_scheduler, and the wps_scheduler does not start when the Portal server starts. Error logs may appear similar to the following:
    [6/21/17 23:05:06:112 EDT] 00000080 InternalGetSc E com.ibm.wps.command.scheduler.internal.InternalGetSchedulerTaskCommand AbstractCommand.throwCommandException EJPDD0009E: JNDI naming lookup failed for name = [ejb/wpsSchedulerManager].
    javax.naming.NameNotFoundException: Context: w7944aCell/nodes/w7944aNode/servers/HCL Portal and HCL Web Content Manager, name: ejb/wpsSchedulerManager: First component in name wpsSchedulerManager not found. [Root exception is org.omg.CosNaming.NamingContextPackage.NotFound: IDL:omg.org/CosNaming/NamingContext/NotFound:1.0]
  • Solution: Edit the WAS 9 properties file, for example, AppServer/properties/migration.properties:
    #-------------------------------------------------------------------------
    # Specify the number of apps which can be installed in parallel.  default=3
    #-------------------------------------------------------------------------
    #com.ibm.ws.migration.application.install.threadPoolSize=3
    Uncomment the property and set to a value of 1. Ex.
    com.ibm.ws.migration.application.install.threadPoolSize=1
    Save the change and continue with the profile migration. Stop the Portal server, and run the post-was-configuration-update again.

  • Problem: When creating a portal page, there is an option to Create Content. When clicking on that option, it presents available templates to create content from Content Hub Authoring Template is one of them. If selecting this template, it may show an error:
    The new content item could not be created. The selected template is not supported. A workflow is required, but the workflow 
    is not set for the item
  • Solution: This authoring template is not designed for this feature. It has to be exempted from creating content items off it from toolbar. To do that:
    1. Go to the Content Hub library.
    2. Open Authoring templates, edit the Content Hub authoring template.
    3. Go to the tab Properties.
    4. Open section Profile.
    5. Add this keyword ibm.portal.toolbar.hiddenTemplate to the keywords text box.
    6. Save and close.
  • Problem: XMLAccess in the applyCF command may fail when updating a virtual portal.
  • Solution: The CF install attempts to update the Portal owned pages in the base portal and all virtual portals. Prior to CF11 there were some cases where error conditions were suppressed for both the base portal and all virtual portals. This has changed and error conditions are no longer suppressed. If the CF install (applyCF command) reports an error while running the XMLAccess tool on a virtual portal, there may be a problem with the 4 virtual portal. If virtual portal(s) have been modified and removed some of the Portal owned pages on purpose then these errors can be safely suppressed and allow the CF install to continue. This can be done by adding the ignore-errors-all-vps property to the applyCF command.

    For example:

    applyCF.(sh|bat) ..... -Dignore-errors-all-vps='true'
    If Portal owned pages have not been purposely removed from the virtual portal, then the error should be investigated and attempted to be fixed.

  • Problem: After HCL Portal Version 8.5 CF12 or later is applied, warning messages may be seen in the SystemOut.log similar to the following, and custom themes or custom portlets may stop working:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory 
    matchesWebAppDefault(aResource) Servlet context [/...] does not specify a blackwhite list when accessing resource [...], falling 
    back to the default [[whitelist(null), blacklist(.*)]]...
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF12 (and later) contains a security fix (PI65954). Similar to APAR PI47714, this APAR warns and then blocks access to files in a web application that does not have a blacklist or whitelist. The difference between PI47714 and PI65954 is that the default value for the blacklist is now ".*" instead of "WEB-INF/.*" which is much more restrictive than before. Applications can define a custom blacklist or whitelist list by adding the keys com.ibm.portal.resource.whitelist and com.ibm.portal.resource.blacklist to their web.xml deployment descriptor. Custom themes must be redeployed for these changes to take effect. Alternatively, as a temporary solution, you can define a different default value for the blacklist. Details on the action required for the custom code mentioned in these messages and the setting of a default value for the blacklist can be found when you go to the HCL Software Support page.

  • Problem: If you drag and drop digital assets into:
    1. The Dropzone in the Sharing Tab of Rich Media Edition console, or;
    2. Shared files directory (MBAssets folder and its subfolders) that the Rich Media Edition uses
    then default renditions may not be created automatically.
  • Solution: One of the two workaround approaches can be followed to place assets into Rich Media Edition and generate default renditions.
    1. Use WCM interface to place assets in Rich Media Edition instead of Dropzone or shared files directory of the Rich Media Edition server.
    2. Launch the link from Shared files section in the Sharing tab of Rich Media Edition console in a browser to open the user's workspace. Login and double click the assets for which you want to generate renditions and open Renditions tab. Click on the Regenerate Renditions button.
  • Problem: If HCL Portal is configured with an empty value for the wps context root or an empty value for the "portal" mapping for public pages then the following message may be seen in the browser when a user accesses a page that has a whitespace character or any other special character in the friendly page name:
    Error 400: EJBCD0006E: The resolution of a URI failed. Refer to the SystemOut.log for more detailed information.
    A similar warning in the SystemOut.log may also be seen:
    ExceptionLogg W com.ibm.wps.logging.ExceptionLogger logThrowable An exception occurred: [Illegal character in path at index 43: com.ibm.portal.friendly.name:/Search/Search Center]. Enable traces for [com.ibm.wps.logging.ExceptionLogger=all] to see the exception stack trace.
  • Solution: This is a known issue and is fixed by applying the WAS interim fix for PI67942 for WAS 8.5.5.10. This issue will also be fixed / included in WAS 8.5.5.11.

  • Problem: A decrease in performance may occur when Textbox.io is used with large amounts of HTML (~4000 lines/20,000 characters) that can cause the Firefox browser to stop working when switching between code and design views.
  • Solution: When switching between code and design views, save and close the content and then reopen it before returning to the design view. For other browsers, waiting and allowing processes to continue running eventually returns to the design view and no further action is required.

  • Problem: If WAS 8.5.5.10 is installed with JDK8 for the Portal server profile and CF12 or later is installed, the following errors may be found in the ConfigTrace.log file:
    [exec] SEVERE:
    [exec] java.lang.UnsupportedClassVersionError: JVMCFRE003 bad major version; class=sun/rmi/rmic/Main, offset=6
    and
    [exec] SEVERE: Error generating RMI code: Cannot run RMIC because it is not installed. Expected location of RMIC is the following: /opt/IBM/WebSphere/AppServer/java_1.8_64/lib.
    [exec] Exception in thread "main" java/lang/NullPointerException
  • Solution: There is a problem communicating with RMIC while preparing an Enterprise Application for installation. This is fixed by installing WAS APAR PI67166 and retrying the applyCF.sh command. It can also be avoided by installing WAS APAR PI67166 before the Portal CF update. Note: This APAR is scheduled to be included in WAS 855.11.

  • Problem: On iSeries environments configured with JDK8 and WAS 8.5.5.10, you may get the error message "Java runtime version not supported" in ConfigTrace.log during the installation of HCL Portal CF12 or later. In more detail, an error similar to the following may be seen in the ConfigTrace.log file:
    [exec] Java runtime version not supported
    [exec] 1.8.0
    [exec] Java runtime version not supported
    [exec] com.ibm.etools.ejbdeploy.EJBDeploymentException: 1.8.0
    [exec] Java runtime version not supported
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.ValidateJavaRTenv(EJBDeploy.java:475)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.execute(EJBDeploy.java:74)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.deploy(EJBDeploy.java:349)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.main(EJBDeploy.java:310)
    [exec] EJBDeploy level: @build@
    --- Exception Thrown ---
    /QIBM/ProdData/WebSphere/PortalServer/V85/Server/wcm/prereq.wcm/config/includes/prereq.wcm.fp_cfg.xml:121: exec returned: 1
  • Solution: To use JDK8 with WAS 8.5.5.10 on IBM i, it will be necessary to first run a script to patch the ejbdeploy file. This script should be run after the IM install step and before the applyCF manual step. If this error has already been encountered during the applyCF command, then the script can be run, and then re-run the applyCF command. To run the script:
    1. Open a command prompt in the (profile_root)/PortalServer/bin/ directory.
    2. Run the command:
      i5ejbscript update
    3. To revert to the old state, run the following command to restore the link to the original file:
      i5ejbscript restore
      Note: WAS 8.5.5.11 should contain the fix for the ejbdeploy file, and this action will no longer be needed.
  • Problem: ISAM v7 does not support JDK8. Any customer choosing to move to JDK8 will no longer have compatibility for ISAM v7. ISAM v9 will support JDK8, but Portal does not yet support this level.
  • Solution: This is currently a Portal limitation. Future plans may include ISAM v9 support in a later CF.

  • Problem: No content indexed in default search collection and system was configured with a custom Portal context root.
  • Solution: Prior to CF06 there was a manual step required after changing the context root. The context in the search content source had to be manually adjusted. If this was not performed, then the content source will be using the wrong context root and will not find any content to index. Beginning with CF06 the update of the search content source was included in the task modify-servlet-path. To solve this problem, the context in the search content source can be adjusted manually or just run the task modify-servlet-path again. See also Why do I get an EJPJO0105E error managing search collections on an alternate context root? when you go to the HCL Software Support page.

  • Problem: In the WCM Authoring environment, when the DAM popup is opened to select an image, it may show all the assets available in DAM, but the thumbnails will not be generated properly. This may be due to a particular library which is installed with Linux, but is at a different version to that which MediaBeacon utilizes.
  • Solution: The typical solution is to create a symbolic link, as shown below:
    1. Open Terminal.
    2. Run the command:
      cd /lib64
    3. If Linux is installed with libbz2.so.1.0.4 for example, then run the command:
      
      ln -s libbz2.so.1.0.4 libbz2.so.1.0
  • Problem: In installations with a large number of resources, it is possible that the applyCF script could run out of memory and fail. Such a failure would record a java.lang.OutOfMemoryError: Java heap space message and a Java stack trace in the ConfigTrace.log file.
  • Solution: If this problem is encountered, the amount of memory available to the JVM can be increased, and the configuration can then be re-run.

    For Unix/Linux:

    1. Backup the (AppServer_home)/bin/wsadmin.sh file.
    2. Open this file in a text editor.
    3. Locate the PERF_JVM_OPTIONS= line specific to your operating system.
    4. Modify this line to have an Xmx parameter of 2048, e.g. from:
      PERF_JVM_OPTIONS="-Xms256m -Xmx256m -Xquickstart" ;;
      to:
      PERF_JVM_OPTIONS="-Xms256m -Xmx2048m -Xquickstart" ;;
    5. Save changes.
    6. Re-run applyCF.sh. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
    For Windows:
    1. Backup the (AppServer_home)\bin\wsadmin.bat file.
    2. Open this file in a text editor.
    3. Locate the "set PERFJAVAOPTION" statement.
    4. Modify this line to have an Xmx parameter of 2048, e.g. from:
      set PERFJAVAOPTION=-Xms256m -Xmx256m -Xquickstart
      to:
      set PERFJAVAOPTION=-Xms256m -Xmx2048m -Xquickstart
    5. Save changes.
    6. Re-run applyCF.bat. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
  • Problem: Where Textbox.io is the configured rich text editor in WCM, rapidly opening and resaving an existing item containing a rich text element can cause the loss of existing content if the item is saved before the rich text content is fully loaded. This is only likely to be an issue in environments with high network latency.
  • Solution: Ensure when editing an existing content, site area, template or component containing a rich text element, the rich text content is fully loaded in the editor before re-saving the item.

  • Problem: During the migration process the node may not start after upgrading to WAS 8.5.5.9 and before executing upgrade-profile. The node does not start with error message:
    ADMU3200I: Server launched. Waiting for initialization status.
    ADMU3011E: Server launched but failed initialization. Server logs, startServer.log, and other log files
    The log contains the following error message:
    HMGR0031E: A conflicting IP address and port has been detected for the DCS_UNICAST_ADDRESS end point.
    The node does not start due to conflicting port configuration for the DCS_UNICAST_ADDRESS end point of server1 and nodeagent in the node serverindex.xml.
  • Solution: It is recommended to execute upgrade-profile before upgrading to WAS 8.5.5.9. Alternatively the conflicting port configuration for the DCS_UNICAST_ADDRESS needs to be solved by setting unique port numbers for server1 and nodeagent.

  • Problem: After Rollback from CF09 or above to CF08 or below, the portal toolbar may show extra items that have titles $Banner$, $Teaser$ and $Video$.
  • Solution: These items can be ignored and should not be used. During an update to CF09 or above again, these items will disappear. To have these items no longer visible in the toolbar, go to Applications > Content > Web Content Authoring. Search for each of these pieces of content. Edit each of the pieces of content and go to Properties. In the profile keyword field, remove the keyword ibm.portal.toolbar.NewContent. Save this change.

  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too long 
            on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path. Details on Changing the location of the wstemp temporary workspace directory be found when you go to the HCL Software Support page.

  • Problem: During the upgrade-profile task in migration, there may be an intermittent issue which can cause the uninstall of an application to fail because files from the application are locked and cannot be deleted. This is an example of an error message that may be seen:
    BUILD FAILED
    C:\ibm\wp85\ConfigEngine\config\includes\apply_PTF.xml:172:
    C:\ibm\wp85\PortalServer\jcr\wp.content.repository.install\config\includes\jcr.mig_cfg.xml:382: Exception found when executing wsadmin: (hostname)
    c:/ibm/wp85/ConfigEngine/lib/wkplc.misc.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batchboot.jar;C:/ibm/wp85/AppServer/deploytool/itp/batch2.jar;
    C:/ibm/wp85/PortalServer/base/wp.base/shared/app/wp.base.jar;C:/ibm/wp85/PortalServer/shared/app/wp.base.jar
  • Solution: If this issue is encountered, wait 5 minutes and re-run the upgrade-profile task with the same options to resume the migration. The lock on the file should be released and the uninstall can continue as expected.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (INSTANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page. See Creating page templates topic for more information. This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of the CF the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page. This is a permanent change in behavior effective in CF08 or later.

  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site/Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: Exception found when 
    executing wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;
    C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'
    Ear File 
    C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear
    WASX7015E: Exception running command: "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; 
    exception information:
    com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: 
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file:
    1. Linux, AIX, Solaris: /etc/hosts
    2. Windows: %SystemRoot%\System32\drivers\etc\hosts and add a line similar to the following, substituting your actual machine name where indicated:
      127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
      If a line for localhost already exists, you can add your hostname to this existing line.
    3. IBM i: To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.
  • Problem: After the CF is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) 
    Servlet context [/...] does not specify a blackwhite list when accessing resource [...], falling back to the default 
    [[(null), (WEB-INF/.*)]]. Applications can define a custom list by adding the keys [com.ibm.portal.resource.whitelist] and 
    [com.ibm.portal.resource.blacklist] to their web.xml deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 (or later) contained a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the res datasource. Details on the action required for the custom code mentioned in these messages can be found when you go to the HCL Software Support page. You must redeploy your custom theme for these changes to take effect.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the JSP components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting the following link on their support site: https://ephox.zendesk.com/entries/96347398-Java-1-8-u60-Internet-Explorer-Issue.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.io. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it

  • Problem: Error messages can be seen in the SystemOut.log after rollback regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190)
  • Solution: These logging errors are expected if content was created against a Portal feature at an updated CF level and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables
  • Problem: Starting with HCL Portal 8.5.0 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title) (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal 8.5 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal 8.5 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal 8.5.0. Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: For IBM i only: May have the following error when running applyCF.sh when Portal databases have been transferred to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database directory.
  • Solution: Run the following steps then rerun applyCF.sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed.
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches:
         [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh
    --- Exception Thrown ---
    /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named EJBDeploy tool for pre-EJB 3.0 modules.

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing HCL Portal 8.5 GA and before updating to the current CF, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply HCL Portal 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall HCL Portal 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall Portal 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 or later and retry the rollback. Please see the following documentation for more details and resolution.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    "Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback."
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.

Known Issues for HCL Portal Version 8.5 / Version 9.0 Combined CF14 (PI73835)

  • Problem: On a Linux or Unix system, it is possible that the applyCF command may fail, with this error message recorded in ConfigTrace.log:
    Couldn't flush user prefs: java.util.prefs.BackingStoreException: Couldn't get file lock.
  • Solution: The problem concerns Java system and/or user preferences and can potentially be addressed in one of the following ways:
    1. Make sure that Java's preferences directory is accessible. Ensure that the /etc/.java/.systemPrefs directory exists and is read/writable by the user performing the upgrade as well as any alternate user ID used to do WAS administration.
    2. Create a new unique Java preferences path for use by WAS.
      • a. Create a new sub-directory of your choice (for example, under /home/wasadmin/prefs/).
      • b. Assign permissions on this directory so that it is read/writable by the user performing the upgrade as well as any alternate user ID used to do WAS administration.
      • c. Set the following argument on the affected WebSphere JVM: -Djava.util.prefs.userRoot=/home/wasadmin/prefs. To set this argument, go to the administrative console, click Servers > Server Types > WebSphere applicaton servers, and select the server that you want to add the generic argument to.
      • d. Then, in the Server Infrastructure section, click Java and process management > Process definition > Java virtual machine. Then add -Djava.util.prefs.userRoot=/home/wasadmin/prefs to any existing properties in the Generic JVM Arguments field and click OK.
      • e. Click Save to save your changes. Then restart the application server.
  • Problem: An error may occur during applyCF on the primary cluster node, similar to the following:
    [wplc-remove-ear] AppName EphoxTbioServices 
    [wplc-remove-ear] ADMA5017I: Uninstallation of EphoxTbioServices started. 
    [wplc-remove-ear] ADMA5104I: The server index entry for WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin11+WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin09+WebSphere:cell=psvtmigwin08Cell01,node=psvtmigwin10Node01 is updated successfully. 
    [wplc-remove-ear] ADMA5102I: The configuration data for EphoxTbioServices from the configuration repository is deleted successfully. 
    [wplc-remove-ear] ADMA5011I: The cleanup of the temp directory for application EphoxTbioServices is complete. 
    [wplc-remove-ear] ADMA5106I: Application EphoxTbioServices uninstalled successfully. 
    [wplc-remove-ear] WASX7015E: Exception running command: "source "C:/ibm/wp85/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; exception information: 
    [wplc-remove-ear] com.ibm.websphere.management.exception.ConfigServiceException 
    [wplc-remove-ear] com.ibm.ws.sm.workspace.WorkSpaceException 
    [wplc-remove-ear] com.ibm.websphere.management.exception.DocumentIOException: cells/psvtmigwin08Cell01/applications/EphoxTbioServices.ear/EphoxTbioServices.ear
    
  • Solution: Disable auto plugin generation. Go to the HCL Software Support page to find instructions to disable auto plugin generation. After applyCF completes, re-enable the two plugin settings.

  • Problem: The post-was-configuration-update task may fail due to an error in the Portal server startup. SystemOut.log shows an exception related to the wps_scheduler, and the wps_scheduler does not start when the Portal server starts. Error logs may appear similar to the following:
    [6/21/17 23:05:06:112 EDT] 00000080 InternalGetSc E 
    com.ibm.wps.command.scheduler.internal.InternalGetSchedulerTaskCommand AbstractCommand.throwCommandException EJPDD0009E: JNDI 
    naming lookup failed for name = [ejb/wpsSchedulerManager].
    javax.naming.NameNotFoundException: Context: w7944aCell/nodes/w7944aNode/servers/HCL Portal and HCL Web Content Manager, name: ejb/wpsSchedulerManager: 
    First component in name wpsSchedulerManager not found. [Root exception is org.omg.CosNaming.NamingContextPackage.NotFound: 
    IDL:omg.org/CosNaming/NamingContext/NotFound:1.0]
  • Solution: Edit the WAS 9 properties file, for example, AppServer/properties/migration.properties:
    #-------------------------------------------------------------------------
    # Specify the number of apps which can be installed in parallel.  default=3
    #-------------------------------------------------------------------------
    #com.ibm.ws.migration.application.install.threadPoolSize=3
    Uncomment the property and set to a value of 1. For example:
    com.ibm.ws.migration.application.install.threadPoolSize=1
    Save the change and continue with the profile migration. The stop the Portal server, and run the post-was-configuration-update again.

  • Problem: When creating a portal page, there is an option to Create Content. When clicking on that option, it presents available templates to create content from, Content Hub Authoring Template is one of them. If selecting this template, it may show an error:
    "The new content item could not be created. The selected template is not supported. A workflow is required, but the workflow is not set for the item"
  • Solution: This authoring template is not designed for this feature. It has to be exempted from creating content items off it from toolbar. To do that:
    1. Go to the Content Hub library.
    2. Open Authoring templates, edit the Content Hub authoring template.
    3. Go to the Properties tab.
    4. Open section Profile.
    5. Add this keyword ibm.portal.toolbar.hiddenTemplate to the keywords text box.
    6. Save and close.
  • Problem: XMLAccess in the applyCF command may fail when updating a virtual portal.
  • Solution: The CF install attempts to update the Portal owned pages in the base portal and all virtual portals. Prior to CF11 there were some cases where error conditions were suppressed for both the base portal and all virtual portals. This has changed and error conditions are no longer suppressed. If the CF install (applyCF command) reports an error while running the XMLAccess tool on a virtual portal, there may be a problem with the virtual portal. If virtual portal(s) have been modified and removed some of the Portal owned pages on purpose then these errors can be safely suppressed and allow the CF install to continue. This can be done by adding the ignore-errors-all-vps property to the applyCF command. Example:
    applyCF.(sh|bat) ..... -Dignore-errors-all-vps='true'
    If Portal owned pages have not been purposely removed from the virtual portal, then the error should be investigated and attempted to be fixed.

  • Problem: After HCL Portal Version 8.5 CF12 or later is applied, warning messages may be seen in the SystemOut.log similar to the following, and custom themes or custom portlets may stop working:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet 
    context [/...] does not specify a blackwhite list when accessing resource [...], falling back to the default 
    [[whitelist(null), blacklist(.*)]]...
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF12 (and later) contains a security fix (PI65954). Similar to APAR PI47714, this APAR warns and then blocks access to files in a web application that does not have a blacklist or whitelist. The difference between PI47714 and PI65954 is that the default value for the blacklist is now ".*" instead of "WEB-INF/.*" which is much more restrictive than before. Applications can define a custom blacklist or whitelist list by adding the keys [com.ibm.portal.resource.whitelist] and [com.ibm.portal.resource.blacklist] to their web.xml deployment descriptor. Custom themes must be redeployed for these changes to take effect. Alternatively, as a temporary solution, you can define a different default value for the blacklist. Details on the action required for the custom code mentioned in these messages and the setting of a default value for the blacklist can be found in the security bulletin in the HCL Software Support page.

  • Problem: If you drag and drop digital assets into:
    1. The Dropzone in the Sharing Tab of Rich Media Edition console, or
    2. Shared files directory (MBAssets folder and its subfolders) that the Rich Media Edition uses

    then default renditions may not be created automatically.

  • Solution: One of the two workaround approaches can be followed to place assets into Rich Media Edition and generate default renditions.
    • 1. Use WCM interface to place assets in Rich Media Edition instead of Dropzone or shared files directory of the Rich Media Edition server.
    • 2. Launch the link from Shared files section in the Sharing tab of Rich Media Edition console in a browser to open the user's workspace. Login and double click the assets for which you want to generate renditions and open Renditions tab. Click on the Regenerate Renditions button.
  • Problem: If HCL Portal is configured with an empty value for the wps context root or an empty value for the portal mapping for public pages then the following message may be seen in the browser when a user accesses a page that has a whitespace character or any other special character in the friendly page name:
    Error 400: EJBCD0006E: The resolution of a URI failed. Refer to the SystemOut.log for more detailed information.
    A similar warning in the SystemOut.log may also be seen:
    ExceptionLogg W com.ibm.wps.logging.ExceptionLogger logThrowable An exception occurred: [Illegal character in path at index 43: com.ibm.portal.friendly.name:/Search/Search Center]. Enable traces for [com.ibm.wps.logging.ExceptionLogger=all] to see the exception stack trace.
  • Solution: This is a known issue and is fixed by applying the WAS interim fix for PI67942 for WAS 8.5.5.10. This issue will also be fixed / included in WAS 8.5.5.11.

  • Problem: A decrease in performance may occur when Textbox.io is used with large amounts of HTML (~4000 lines/20,000 characters) that can cause the Firefox browser to stop working when switching between code and design views.
  • Solution: When switching between code and design views, save and close the content and then reopen it before returning to the design view. For other browsers, waiting and allowing processes to continue running eventually returns to the design view and no further action is required.

  • Problem: If WAS 8.5.5.10 is installed with JDK8 for the Portal server profile and CF12 or later is installed, the following errors may be found in the ConfigTrace.log file:
    [exec] SEVERE:[exec] java.lang.UnsupportedClassVersionError: JVMCFRE003 bad major version; class=sun/rmi/rmic/Main, offset=6
    and
    [exec] SEVERE: Error generating RMI code: Cannot run RMIC because it is not installed. Expected location of RMIC is the following: /opt/IBM/WebSphere/AppServer/java_1.8_64/lib.[exec] Exception in thread "main" java/lang/NullPointerException
  • Solution: There is a problem communicating with RMIC while preparing an Enterprise Application for installation. This is fixed by installing WAS APAR PI67166 and retrying the applyCF.sh command. It can also be avoided by installing WAS APAR PI67166 before the Portal CF update. Note: This APAR is scheduled to be included in WAS 855.11.

  • Problem: On iSeries environments configured with JDK8 and WAS 8.5.5.10, you may get the error message "Java runtime version not supported" in ConfigTrace.log during the installation of HCL Portal CF12 or later. In more detail, an error similar to the following may be seen in the ConfigTrace.log file:
    [exec] Java runtime version not supported
    [exec] 1.8.0
    [exec] Java runtime version not supported
    [exec] com.ibm.etools.ejbdeploy.EJBDeploymentException: 1.8.0
    [exec] Java runtime version not supported
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.ValidateJavaRTenv(EJBDeploy.java:475)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.execute(EJBDeploy.java:74)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.deploy(EJBDeploy.java:349)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.main(EJBDeploy.java:310)
    [exec] EJBDeploy level: @build@
    --- Exception Thrown ---
    /QIBM/ProdData/WebSphere/PortalServer/V85/Server/wcm/prereq.wcm/config/includes/prereq.wcm.fp_cfg.xml:121: exec returned: 1
  • Solution: To use JDK8 with WAS 8.5.5.10 on IBM i, it will be necessary to first run a script to patch the ejbdeploy file. This script should be run after the IM install step and before the applyCF manual step. If this error has already been encountered during the applyCF command, then the script can be run, and then re-run the applyCF command. To run the script:
    • 1. Open a command prompt in the (profile_root)/PortalServer/bin/ directory.
    • 2. Run the command:
    i5ejbscript update
    • 3. To revert to the old state, run the following command to restore the link to the original file:
    i5ejbscript restore
    Note: WAS 8.5.5.11 should contain the fix for the ejbdeploy file, and this action will no longer be needed.
  • Problem: ISAM v7 does not support JDK8. Any customer choosing to move to JDK8 will no longer have compatibility for ISAM v7. ISAM v9 will support JDK8, but Portal does not yet support this level.
  • Solution: This is currently a Portal limitation. Future plans may include ISAM v9 support in a later CF.

  • Problem: No content indexed in default search collection and system was configured with a custom Portal context root.
  • Solution: Prior to CF06 there was a manual step required after changing the context root. The context in the search content source had to be manually adjusted. If this was not performed, then the content source will be using the wrong context root and will not find any content to index. Beginning with CF06 the update of the search content source was included in the task modify-servlet-path. To solve this problem, the context in the search content source can be adjusted manually or just run the task modify-servlet-path again. Go to the HCL Software Support page to learn more about Why do I get an EJPJO0105E error managing search collections on an alternate context root?.

  • Problem: In the WCM Authoring environment, when the DAM popup is opened to select an image, it may show all the assets available in DAM, but the thumbnails will not be generated properly. This may be due to a particular library which is installed with Linux, but is at a different version to that which MediaBeacon utilizes.
  • Solution: The typical solution is to create a symbolic link, as shown below:
    1. Open Terminal.
    2. Run the command:
      cd /lib64
    3. If Linux is installed with libbz2.so.1.0.4 for example, then do:
      ln -s libbz2.so.1.0.4 libbz2.so.1.0
  • Problem: In installations with a large number of resources, it is possible that the applyCF script could run out of memory and fail. Such a failure would record a java.lang.OutOfMemoryError: Java heap space message and a Java stack trace in the ConfigTrace.log file.
  • Solution: If this problem is encountered, the amount of memory available to the JVM can be increased, and the configuration can then be re-run.

    For Unix/Linux:

    1. Backup the (AppServer_home)/bin/wsadmin.sh file.
    2. Open this file in a text editor.
    3. Locate the PERF_JVM_OPTIONS= line specific to your operating system.
    4. Modify this line to have an Xmx parameter of 2048, e.g. from:
      PERF_JVM_OPTIONS="-Xms256m -Xmx256m -Xquickstart" ;;
      to
      PERF_JVM_OPTIONS="-Xms256m -Xmx2048m -Xquickstart" ;;
    5. Save changes.
    6. Re-run applyCF.sh. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
    For Windows:
    1. Backup the (AppServer_home)/bin/wsadmin.sh file.
    2. Open this file in a text editor.
    3. Locate the set PERFJAVAOPTION statement.
    4. Modify this line to have an Xmx parameter of 2048, e.g. from:
      set PERFJAVAOPTION=-Xms256m -Xmx256m -Xquickstart
      to
      set PERFJAVAOPTION=-Xms256m -Xmx2048m -Xquickstart
    5. Save changes.
    6. Re-run applyCF.bat. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
  • Problem: Where Textbox.io is the configured rich text editor in WCM, rapidly opening and resaving an existing item containing a rich text element can cause the loss of existing content if the item is saved before the rich text content is fully loaded. This is only likely to be an issue in environments with high network latency.
  • Solution: Ensure when editing an existing content, site area, template or component containing a rich text element, the rich text content is fully loaded in the editor before re-saving the item.

  • Problem: During the migration process the node may not start after upgrading to WAS 8.5.5.9 and before executing upgrade-profile. The node does not start with error message:
    ADMU3200I: Server launched. Waiting for initialization status.
    ADMU3011E: Server launched but failed initialization. Server logs, startServer.log, and other log files
    The log contains the following error message:
    HMGR0031E: A conflicting IP address and port has been detected for the DCS_UNICAST_ADDRESS end point.
    The node does not start due to conflicting port configuration for the DCS_UNICAST_ADDRESS end point of server1 and nodeagent in the node serverindex.xml.
  • Solution: It is recommended to execute upgrade-profile before upgrading to WAS 8.5.5.9. Alternatively the conflicting port configuration for the DCS_UNICAST_ADDRESS needs to be solved by setting unique port numbers for server1 and nodeagent.

  • Problem: After Rollback from CF09 or above to CF08 or below, the portal toolbar may show extra items that have titles $Banner$, $Teaser$ and $Video$
  • Solution: These items can be ignored and should not be used. During an update to CF09 or above again, these items will disappear. To have these items no longer visible in the toolbar, go to Applications > Content > Web Content Authoring. Search for each of these pieces of content. Edit each of the pieces of content and go to properties. In the profile keyword field, remove the keyword ibm.portal.toolbar.NewContent. Save this change.

  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too long on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
    
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path.

  • Problem: During the upgrade-profile task in migration, there may be an intermittent issue which can cause the uninstall of an application to fail because files from the application are locked and cannot be deleted. This is an example of an error message that may be seen:
    BUILD FAILED
    C:\ibm\wp85\ConfigEngine\config\includes\apply_PTF.xml:172:
    C:\ibm\wp85\PortalServer\jcr\wp.content.repository.install\config\includes\jcr.mig_cfg.xml:382: Exception found when executing wsadmin: (hostname)
    c:/ibm/wp85/ConfigEngine/lib/wkplc.misc.jar;C:/ibm/wp85/AppServer/deploytool/itp/batchboot.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batch2.jar;C:/ibm/wp85/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/ibm/wp85/PortalServer/shared/app/wp.base.jar
  • Solution: If this issue is encountered, wait 5 minutes and re-run the upgrade-profile task with the same options to resume the migration. The lock on the file should be released and the uninstall can continue as expected.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (INSTANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page (See Creating page templates topic for more information). This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of the CF the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page. This is a permanent change in behavior effective in CF08 or later.

  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site/Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: 
    Exception found when executing wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'
    Ear File C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear
    WASX7015E: Exception running command: "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; 
    exception information:
    com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: 
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file:
    • - Linux, AIX, Solaris: /etc/hosts
    • - Windows: %SystemRoot%\System32\drivers\etc\hosts and add a line similar to the following, substituting your actual machine name where indicated:
    127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
    • If a line for localhost already exists, you can add your hostname to this existing line.
    • - IBM i: To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.
  • Problem: After the CF is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] 
    does not specify a blackwhite list when accessing resource [...], falling back to the default [[(null), (WEB-INF/.*)]]. 
    Applications can define a custom list by adding the keys [com.ibm.portal.resource.whitelist] and [com.ibm.portal.resource.blacklist] 
    to their web.xml deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 (or later) contained a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the res datasource. Details on the action required for the custom code mentioned in these messages can be found in the HCL Software Support page. You must redeploy your custom theme for these changes to take effect.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the jsp components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting the following link on their support site.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.io. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it

  • Problem: Error messages can be seen in the SystemOut.log after rollback regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190)
  • Solution: These logging errors are expected if content was created against a Portal feature at an updated CF level and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables
  • Problem: Starting with HCL Portal 8.5.0 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title) (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal 8.5 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal 8.5 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal 8.5. Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: For IBM i only: May have the following error when running applyCF.sh when Portal databases have been transferred to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database directory.
  • Solution: Run the following steps then rerun applyCF.sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed.
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Relational database alias: The hostname. Use the short TCP/IP hostname of the database system.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches:
         [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh
    --- Exception Thrown ---
    /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named EJBDeploy tool for pre-EJB 3.0 modules.

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing Portal 8.5 GA and before updating to the current CF, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply Portal 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall Portal 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall Portal 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 or later and retry the rollback. See this link for more details.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback.
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.

Known Issues for V8.5.0.0 Combined CF13 / V9 (PI68557)

  • Problem: When creating a portal page, there is an option to Create Content. When clicking on that option, it presents available templates to create content from, Content Hub Authoring Template is one of them. If selecting this template, it may show an error:
    "The new content item could not be created. The selected template is not supported. A workflow is required, but the workflow is not set for the item"
  • Solution: This authoring template is not designed for this feature. It has to be exempted from creating content items off it from toolbar. To do that:
    1. Go to the Content Hub library.
    2. Open Authoring templates, edit the Content Hub authoring template.
    3. Go to the Properties tab.
    4. Open section Profile.
    5. Add this keyword ibm.portal.toolbar.hiddenTemplate to the keywords text box.
    6. Save and close.
  • Problem: XMLAccess in the applyCF command may fail when updating a virtual portal.
  • Solution: The CF install attempts to update the Portal owned pages in the base portal and all virtual portals. Prior to CF11 there were some cases where error conditions were suppressed for both the base portal and all virtual portals. This has changed and error conditions are no longer suppressed. If the CF install (applyCF command) reports an error while running the XMLAccess tool on a virtual portal, there may be a problem with the virtual portal. If virtual portal(s) have been modified and removed some of the Portal owned pages on purpose then these errors can be safely suppressed and allow the CF install to continue. This can be done by adding the ignore-errors-all-vps property to the applyCF command. Example:
    applyCF.(sh|bat) ..... -Dignore-errors-all-vps='true'
    If Portal-owned pages have not been purposely removed from the virtual portal, then the error should be investigated and attempted to be fixed.

  • Problem: After HCL Portal Version 8.5 CF12 or later is applied, warning messages may be seen in the SystemOut.log similar to the following, and custom themes or custom portlets may stop working:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] 
    does not specify a blackwhite list when accessing resource [...], falling back to the default [[whitelist(null), blacklist(.*)]]...
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF12 (and later) contains a security fix (PI65954). Similar to APAR PI47714, this APAR warns and then blocks access to files in a web application that does not have a blacklist or whitelist. The difference between PI47714 and PI65954 is that the default value for the blacklist is now ".*" instead of "WEB-INF/.*" which is much more restrictive than before. Applications can define a custom blacklist or whitelist list by adding the keys com.ibm.portal.resource.whitelist and com.ibm.portal.resource.blacklist to their web.xml deployment descriptor. Custom themes must be redeployed for these changes to take effect. Alternatively, as a temporary solution, you can define a different default value for the blacklist. Details on the action required for the custom code mentioned in these messages and the setting of a default value for the blacklist can be found in the security bulletin.

  • Problem: If you drag and drop digital assets into:
    1. The Dropzone in the Sharing Tab of Rich Media Edition console, or
    2. Shared files directory (MBAssets folder and its subfolders) that the Rich Media Edition uses,
    then default renditions may not be created automatically.
  • Solution: One of the two workaround approaches can be followed to place assets into Rich Media Edition and generate default renditions:
    • 1. Use WCM interface to place assets in Rich Media Edition instead of Dropzone or shared files directory of the Rich Media Edition server.
    • 2. Launch the link from Shared files section in the Sharing tab of Rich Media Edition console in a browser to open the user's workspace.
    • 3. Login and double click the assets for which you want to generate renditions and open Renditions tab. Click on Regenerate Renditions button.
  • Problem: If Portal is configured with an empty value for the wps context root or an empty value for the portal mapping for public pages then the following message may be seen in the browser when a user accesses a page that has a whitespace character or any other special character in the friendly page name:
    Error 400: EJBCD0006E: The resolution of a URI failed. Refer to the SystemOut.log for more detailed information.
    A similar warning in the SystemOut.log may also be seen:
    ExceptionLogg W com.ibm.wps.logging.ExceptionLogger logThrowable An exception occurred: [Illegal character in path at index 43: com.ibm.portal.friendly.name:/Search/Search Center]. Enable traces for [com.ibm.wps.logging.ExceptionLogger=all] to see the exception stack trace.
  • Solution: This is a known issue and is fixed by applying the WAS interim fix for PI67942 for WAS 8.5.5.10. This issue will also be fixed / included in WAS 8.5.5.11.

  • Problem: A decrease in performance may occur when Textbox.io is used with large amounts of HTML (~4000 lines/20,000 characters) that can cause the Firefox browser to stop working when switching between code and design views.
  • Solution: When switching between code and design views, save and close the content and then reopen it before returning to the design view. For other browsers, waiting and allowing processes to continue running eventually returns to the design view and no further action is required.

  • Problem: If WAS 8.5.5.10 is installed with JDK8 for the Portal server profile and CF12 or later is installed, the following errors may be found in the ConfigTrace.log file:
    [exec] SEVERE:
    [exec] java.lang.UnsupportedClassVersionError: JVMCFRE003 bad major version; class=sun/rmi/rmic/Main, offset=6
    and
    [exec] SEVERE: Error generating RMI code: Cannot run RMIC because it is not installed. Expected location of RMIC is the following: /opt/IBM/WebSphere/AppServer/java_1.8_64/lib.
    [exec] Exception in thread "main" java/lang/NullPointerException
  • Solution: There is a problem communicating with RMIC while preparing an Enterprise Application for installation. This is fixed by installing WAS APAR PI67166 and retrying the applyCF.sh command. It can also be avoided by installing WAS APAR PI67166 before the Portal CF update.
    Note: This APAR is scheduled to be included in WAS 855.11.

  • Problem: On iSeries environments configured with JDK8 and WAS 8.5.5.10, you may get the error message "Java runtime version not supported" in ConfigTrace.log during the installation of Portal CF12 or later. In more detail, an error similar to the following may be seen in the ConfigTrace.log file:
    [exec] Java runtime version not supported
    [exec] 1.8.0
    [exec] Java runtime version not supported
    [exec] com.ibm.etools.ejbdeploy.EJBDeploymentException: 1.8.0
    [exec] Java runtime version not supported
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.ValidateJavaRTenv(EJBDeploy.java:475)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.execute(EJBDeploy.java:74)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.deploy(EJBDeploy.java:349)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.main(EJBDeploy.java:310)
    [exec] EJBDeploy level: @build@
    --- Exception Thrown ---
    /QIBM/ProdData/WebSphere/PortalServer/V85/Server/wcm/prereq.wcm/config/includes/prereq.wcm.fp_cfg.xml:121: exec returned: 1
  • Solution: To use JDK8 with WAS 8.5.5.10 on IBM i, it will be necessary to first run a script to patch the ejbdeploy file. This script should be run after the IM install step and before the applyCF manual step. If this error has already been encountered during the applyCF command, then the script can be run, and then re-run the applyCF command. To run the script:
    1. Open a command prompt in the (profile_root)/PortalServer/bin/ directory.
    2. Run the command:
      i5ejbscript update
    3. To revert to the old state, run the following command to restore the link to the original file:
      i5ejbscript restore
      Note: WAS 8.5.5.11 should contain the fix for the ejbdeploy file, and this action will no longer be needed.

  • Problem: ISAM v7 does not support JDK8. Any customer choosing to move to JDK8 will no longer have compatibility for ISAM v7. ISAM v9 will support JDK8, but Portal does not yet support this level.
  • Solution: This is currently a Portal limitation. Future plans may include ISAM v9 support in a later CF.

  • Problem: No content indexed in default search collection and system was configured with a custom Portal context root.
  • Solution: Prior to CF06 there was a manual step required after changing the context root. The context in the search content source had to be manually adjusted. If this was not performed, then the content source will be using the wrong context root and will not find any content to index. Beginning with CF06 the update of the search content source was included in the task modify-servlet-path. To solve this problem, the context in the search content source can be adjusted manually or just run the task modify-servlet-path again.

  • Problem: In the WCM Authoring environment, when the DAM popup is opened to select an image, it may show all the assets available in DAM, but the thumbnails will not be generated properly. This may be due to a particular library which is installed with Linux, but is at a different version to that which MediaBeacon utilizes.
  • Solution: The typical solution is to create a symbolic link, as shown below:
    1. Open Terminal.
    2. Run the following command:
      cd /lib64
    3. If Linux is installed with libbz2.so.1.0.4 for example, then do:
      ln -s libbz2.so.1.0.4 libbz2.so.1.0
  • Problem: In installations with a large number of resources, it is possible that the applyCF script could run out of memory and fail. Such a failure would record a "java.lang.OutOfMemoryError: Java heap space" message and a Java stack trace in the ConfigTrace.log file.
  • Solution: If this problem is encountered, the amount of memory available to the JVM can be increased, and the configuration can then be re-run. In Unix/Linux:
    1. Backup the (AppServer_home)/bin/wsadmin.sh file.
    2. Open this file in a text editor.
    3. Locate the PERF_JVM_OPTIONS= line specific to your operating system.
    4. Modify this line to have an Xmx parameter of 2048, e.g. from:
      PERF_JVM_OPTIONS="-Xms256m -Xmx256m -Xquickstart" ;;
      to:
      PERF_JVM_OPTIONS="-Xms256m -Xmx2048m -Xquickstart" ;;
    5. Save changes.
    6. Re-run applyCF.sh. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
    For Windows:
    1. Backup the (AppServer_home)\bin\wsadmin.bat file.
    2. Open this file in a text editor.
    3. Locate the set PERFJAVAOPTION statement.
    4. Modify this line to have an Xmx parameter of 2048, e.g. from:
      set PERFJAVAOPTION=-Xms256m -Xmx256m -Xquickstart
      to:
      set PERFJAVAOPTION=-Xms256m -Xmx2048m -Xquickstart
    5. Save changes.
    6. Re-run applyCF.bat. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
  • Problem: Where Textbox.io is the configured rich text editor in WCM, rapidly opening and resaving an existing item containing a rich text element can cause the loss of existing content if the item is saved before the rich text content is fully loaded. This is only likely to be an issue in environments with high network latency.
  • Solution: Ensure when editing an existing content, site area, template or component containing a rich text element, the rich text content is fully loaded in the editor before re-saving the item.

  • Problem: During the migration process the node may not start after upgrading to WAS 8.5.5.9 and before executing upgrade-profile. The node does not start with error message:
    ADMU3200I: Server launched. Waiting for initialization status.
    ADMU3011E: Server launched but failed initialization. Server logs, startServer.log, and other log files
    The log contains the following error message:
    HMGR0031E: A conflicting IP address and port has been detected for the DCS_UNICAST_ADDRESS end point.
    The node does not start due to conflicting port configuration for the DCS_UNICAST_ADDRESS end point of server1 and nodeagent in the node serverindex.xml.
  • Solution: It is recommended to execute upgrade-profile before upgrading to WAS 8.5.5.9. Alternatively the conflicting port configuration for the DCS_UNICAST_ADDRESS needs to be solved by setting unique port numbers for server1 and nodeagent.

  • Problem: After Rollback from CF09 or above to CF08 or below, the portal toolbar may show extra items that have titles $Banner$, $Teaser$ and $Video$.
  • Solution: These items can be ignored and should not be used. During an update to CF09 or above again, these items will disappear. To have these items no longer visible in the toolbar, go to Applications > Content > Web Content Authoring. Search for each of these pieces of content. Edit each of the pieces of content and go to properties. In the profile keyword field, remove the keyword ibm.portal.toolbar.NewContent. Save this change.

  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too long on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path.

  • Problem: During the upgrade-profile task in migration, there may be an intermittent issue which can cause the uninstall of an application to fail because files from the application are locked and cannot be deleted. This is an example of an error message that may be seen:
    BUILD FAILED
    C:\ibm\wp85\ConfigEngine\config\includes\apply_PTF.xml:172: 
    C:\ibm\wp85\PortalServer\jcr\wp.content.repository.install\config\includes\jcr.mig_cfg.xml:382: 
    Exception found when executing wsadmin: (hostname)
    c:/ibm/wp85/ConfigEngine/lib/wkplc.misc.jar;C:/ibm/wp85/AppServer/deploytool/itp/batchboot.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batch2.jar;C:/ibm/wp85/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/ibm/wp85/PortalServer/shared/app/wp.base.jar
  • Solution: If this issue is encountered, wait 5 minutes and re-run the upgrade-profile task with the same options to resume the migration. The lock on the file should be released and the uninstall can continue as expected.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (INSTANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page (See Creating page templates topic for more information). This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of the CF the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page. This is a permanent change in behavior effective in CF08 or later.

  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site/Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: 
    Exception found when executing wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;
    C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'
    Ear File C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear
    WASX7015E: Exception running command: "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; 
    exception information:
    com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: 
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file:
    • 1. Linux, AIX, Solaris: /etc/hosts
    • 2. Windows: %SystemRoot%\System32\drivers\etc\hosts and add a line similar to the following, substituting your actual machine name where indicated:
    127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
    • If a line for localhost already exists, you can add your hostname to this existing line.
    • 3. IBM i: To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.
  • Problem: After the CF is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque 
    W com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] 
    does not specify a blackwhite list when accessing resource [...], falling back to the default [[(null), (WEB-INF/.*)]]. 
    Applications can define a custom list by adding the keys [com.ibm.portal.resource.whitelist] and [com.ibm.portal.resource.blacklist] 
    to their web.xml deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 (or later) contained a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the res datasource. Details on the action required for the custom code mentioned in these messages can be found in the security bulletin. You must redeploy your custom theme for these changes to take effect.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the JSP components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting the following link on their support site.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.IO. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it

  • Problem: Error messages can be seen in the SystemOut.log after rollback regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190)
  • Solution: These logging errors are expected if content was created against a Portal feature at an updated CF level and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables
  • Problem: Starting with HCL Portal 8.5.0 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title) (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal 8.5.0 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal 8.5.0 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal 8.5.0. Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: For IBM i only: May have the following error when running applyCF.sh when Portal databases have been transferred to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database directory.
  • Solution: Run the following steps then rerun applyCF.sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed.
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Relational database alias: The hostname. Use the short TCP/IP hostname of the database system.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches:
         [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh
    --- Exception Thrown ---
    /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named EJBDeploy tool for pre-EJB 3.0 modules.

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing Portal 8.5 GA and before updating to the current CF, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply Portal 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall Portal 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall Portal 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 or later and retry the rollback. See this link for more details.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback.
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.

Known Issues for V8.5.0.0 Combined CF12 (PI64037)

  • Problem: After 8500 CF12 is applied, warning messages may be seen in the SystemOut.log similar to the following, and custom themes or custom portlets may stop working:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque 
    W com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] 
    does not specify a blackwhite list when accessing resource [...], falling back to the default [[whitelist(null), blacklist(.*)]]...
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF12 (and later) contains a security fix (PI65954). Similar to APAR PI47714, this APAR warns and then blocks access to files in a web application that does not have a blacklist or whitelist. The difference between PI47714 and PI65954 is that the default value for the blacklist is now ".*" instead of "WEB-INF/.*" which is much more restrictive than before. Applications can define a custom blacklist or whitelist list by adding the keys com.ibm.portal.resource.whitelist and com.ibm.portal.resource.blacklist to their web.xml deployment descriptor. Custom themes must be redeployed for these changes to take effect. Alternatively, as a temporary solution, you can define a different default value for the blacklist. Details on the action required for the custom code mentioned in these messages and the setting of a default value for the blacklist can be found in the security bulletin.

  • Problem: XMLAccess in the applyCF command may fail when updating a virtual portal.
  • Solution: The CF install attempts to update the Portal owned pages in the base portal and all virtual portals. Prior to CF11 there were some cases where error conditions were suppressed for both the base portal and all virtual portals. This has changed and error conditions are no longer suppressed. If the CF install (applyCF command) reports an error while running the XMLAccess tool on a virtual portal, there may be a problem with the virtual portal. If virtual portal(s) have been modified and removed some of the Portal owned pages on purpose then these errors can be safely suppressed and allow the CF install to continue. This can be done by adding the ignore-errors-all-vps property to the applyCF command. Example:
    applyCF.(sh|bat) ..... -Dignore-errors-all-vps='true'
    If Portal owned pages have not been purposely removed from the virtual portal, then the error should be investigated and attempted to be fixed.

  • Problem: If you drag and drop digital assets into:
    1. The Dropzone in the Sharing Tab of Rich Media Edition console, or
    2. Shared files directory (MBAssets folder and its subfolders) that the Rich Media Edition uses
    then default renditions may not be created automatically.
  • Solution: One of the two workaround approaches can be followed to place assets into Rich Media Edition and generate default renditions:
    • Use WCM interface to place assets in Rich Media Edition instead of Dropzone or shared files directory of the Rich Media Edition server.
    • Launch the link from Shared files section in the Sharing tab of Rich Media Edition console in a browser to open the user's workspace. Login and double click the assets for which you want to generate renditions and open Renditions tab. Click on Regenerate Renditions button.
  • Problem: If Portal is configured with an empty value for the wps context root or an empty value for the portal mapping for public pages then the following message may be seen in the browser when a user accesses a page that has a whitespace character or any other special character in the friendly page name:
    Error 400: EJBCD0006E: The resolution of a URI failed. Refer to the SystemOut.log for more detailed information.
    A similar warning in the SystemOut.log may also be seen:
    ExceptionLogg W com.ibm.wps.logging.ExceptionLogger logThrowable An exception occurred: [Illegal character in path at index 43: 
    com.ibm.portal.friendly.name:/Search/Search Center]. Enable traces for [com.ibm.wps.logging.ExceptionLogger=all] to see the exception stack trace.
  • Solution: This is a known issue and is fixed by applying the WAS interim fix for PI67942 for WAS 8.5.5.10. This issue will also be fixed / included in WAS 8.5.5.11.

  • Problem: There is a known problem in CF12 to activate a temporary RME license. The button to click, Click to permanently activate WCM_RME MB does not appear, so you can not activate the license.
  • Solution: Please create a PMR with HCL Software Support to activate the key. Once activated, enter the new license information in the screen you see at the point of 6c and then continue from 6d. This will be fixed in the next CF13.

  • Problem: A decrease in performance may occur when Textbox.io is used with large amounts of HTML (~4000 lines/20,000 characters) that can cause the Firefox browser to stop working when switching between code and design views.
  • Solution: When switching between code and design views, save and close the content and then reopen it before returning to the design view. For other browsers, waiting and allowing processes to continue running eventually returns to the design view and no further action is required.

  • Problem: If WAS 8.5.5.10 is installed with JDK8 for the Portal server profile and CF12 or later is installed, the following errors may be found in the ConfigTrace.log file:
    [exec] SEVERE:
    [exec] java.lang.UnsupportedClassVersionError: JVMCFRE003 bad major version; class=sun/rmi/rmic/Main, offset=6
    and
    [exec] SEVERE: Error generating RMI code: Cannot run RMIC because it is not installed. Expected location of RMIC is the following: /opt/IBM/WebSphere/AppServer/java_1.8_64/lib.
    [exec] Exception in thread "main" java/lang/NullPointerException
  • Solution: There is a problem communicating with RMIC while preparing an Enterprise Application for installation. This is fixed by installing WAS APAR PI67166 and retrying the applyCF.sh command. It can also be avoided by installing WAS APAR PI67166 before the Portal CF update. Note: This APAR is scheduled to be included in WAS 855.11.

  • Problem: On iSeries environments configured with JDK8 and WAS 8.5.5.10, you may get the error message "Java runtime version not supported" in ConfigTrace.log during the installation of Portal CF12 or later. In more detail, an error similar to the following may be seen in the ConfigTrace.log file:
    [exec] Java runtime version not supported
    [exec] 1.8.0
    [exec] Java runtime version not supported
    [exec] com.ibm.etools.ejbdeploy.EJBDeploymentException: 1.8.0
    [exec] Java runtime version not supported
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.ValidateJavaRTenv(EJBDeploy.java:475)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.execute(EJBDeploy.java:74)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.deploy(EJBDeploy.java:349)
    [exec] at com.ibm.etools.ejbdeploy.EJBDeploy.main(EJBDeploy.java:310)
    [exec] EJBDeploy level: @build@
    --- Exception Thrown ---
    /QIBM/ProdData/WebSphere/PortalServer/V85/Server/wcm/prereq.wcm/config/includes/prereq.wcm.fp_cfg.xml:121: exec returned: 1
  • Solution: To use JDK8 with WAS 8.5.5.10 on IBM i, it will be necessary to first run a script to patch the ejbdeploy file. This script should be run after the IM install step and before the applyCF manual step. If this error has already been encountered during the applyCF command, then the script can be run, and then re-run the applyCF command. To run the script:
    1. Open a command prompt in the (profile_root)/PortalServer/bin/ directory.
    2. Run the command:
      i5ejbscript update
    3. To revert to the old state, run the following script to restore the link to the original file:
      i5ejbscript restore
      Note: WAS 8.5.5.11 should contain the fix for the ejbdeploy file, and this action will no longer be needed.
  • Problem: While using JDK 8, the side navigation of the web help search function in the online help dialog may fail with error message
    "Error 500: org.apache.jasper.JasperException: org.apache.jasper.JasperException: Unable to load class for JSP"
    displayed in the browser. You may also find the following message in the SystemOut.log:
    "Compilation error with 'An error occurred at line: 314 in the jsp file: /advanced/searchView.jsp."
  • Solution: There is no known workaround. This will be fixed in the next CF13.

  • Problem: ISAM v7 does not support JDK8. Any customer choosing to move to JDK8 will no longer have compatibility for ISAM v7. ISAM v9 will support JDK8, but Portal does not yet support this level.
  • Solution: This is currently a Portal limitation. Future plans may include ISAM v9 support in a later CF.

  • Problem: No content indexed in default search collection and system was configured with a custom Portal context root.
  • Solution: Prior to CF06 there was a manual step required after changing the context root. The context in the search content source had to be manually adjusted. If this was not performed, then the content source will be using the wrong context root and will not find any content to index. Beginning with CF06 the update of the search content source was included in the task modify-servlet-path. To solve this problem, the context in the search content source can be adjusted manually or just run the task modify-servlet-path again.

  • Problem: In the WCM Authoring environment, when the DAM popup is opened to select an image, it may show all the assets available in DAM, but the thumbnails will not be generated properly. This may be due to a particular library which is installed with Linux, but is at a different version to that which MediaBeacon utilizes.
  • Solution: The typical solution is to create a symbolic link, as explained below:
    1. Open Terminal.
    2. Run the following command:
      cd /lib64
    3. If Linux is installed with libbz2.so.1.0.4 for example, then do:
      ln -s libbz2.so.1.0.4 libbz2.so.1.0
  • Problem: In installations with a large number of resources, it is possible that the applyCF script could run out of memory and fail. Such a failure would record "java.lang.OutOfMemoryError: Java heap space" message and a Java stack trace in the ConfigTrace.log file.
  • Solution: If this problem is encountered, the amount of memory available to the JVM can be increased, and the configuration can then be re-run.

    For Unix/Linux:

    1. Backup the (AppServer_home)/bin/wsadmin.sh file.
    2. pen this file in a text editor.
    3. Locate the PERF_JVM_OPTIONS= line specific to your operating system.
    4. Modify this line to have an Xmx parameter of 2048, e.g. from:
      PERF_JVM_OPTIONS="-Xms256m -Xmx256m -Xquickstart" ;;
      to:
      PERF_JVM_OPTIONS="-Xms256m -Xmx2048m -Xquickstart" ;;
    5. Save changes.
    6. Re-run applyCF.sh. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
    For Windows:
    1. Backup the (AppServer_home)\bin\wsadmin.bat file.
    2. Open this file in a text editor.
    3. Locate the set PERFJAVAOPTION statement.
    4. Modify this line to have an Xmx parameter of 2048, e.g. from:
      set PERFJAVAOPTION=-Xms256m -Xmx256m -Xquickstart
      to:
      set PERFJAVAOPTION=-Xms256m -Xmx2048m -Xquickstart
    5. Save changes.
    6. Re-run applyCF.bat. After some initialization, the script will skip configuration steps that have already completed and pick up where it left off before the crash.
  • Problem: Where Textbox.io is the configured rich text editor in WCM, rapidly opening and resaving an existing item containing a rich text element can cause the loss of existing content if the item is saved before the rich text content is fully loaded. This is only likely to be an issue in environments with high network latency.
  • Solution: Ensure when editing an existing content, site area, template or component containing a rich text element, the rich text content is fully loaded in the editor before re-saving the item.

  • Problem: During the migration process the node may not start after upgrading to WAS 8.5.5.9 and before executing upgrade-profile. The node does not start with error message:
    ADMU3200I: Server launched. Waiting for initialization status.
    ADMU3011E: Server launched but failed initialization. Server logs, startServer.log, and other log files
    The log contains the following error message:
    HMGR0031E: A conflicting IP address and port has been detected for the DCS_UNICAST_ADDRESS end point.
    The node does not start due to conflicting port configuration for the DCS_UNICAST_ADDRESS end point of server1 and nodeagent in the node serverindex.xml.
  • Solution: It is recommended to execute upgrade-profile before upgrading to WAS 8.5.5.9. Alternatively the conflicting port configuration for the DCS_UNICAST_ADDRESS needs to be solved by setting unique port numbers for server1 and nodeagent.

  • Problem: After Rollback from CF09 or above to CF08 or below, the portal toolbar may show extra items that have titles $Banner$, $Teaser$ and $Video$.
  • Solution: These items can be ignored and should not be used. During an update to CF09 or above again, these items will disappear. To have these items no longer visible in the toolbar, go to Applications > Content > Web Content Authoring. Search for each of these pieces of content. Edit each of the pieces of content and go to properties. In the profile keyword field, remove the keyword ibm.portal.toolbar.NewContent. Save this change.

  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too long on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path.

  • Problem: During the upgrade-profile task in migration, there may be an intermittent issue which can cause the uninstall of an application to fail because files from the application are locked and cannot be deleted. This is an example of an error message that may be seen:
    BUILD FAILED
    C:\ibm\wp85\ConfigEngine\config\includes\apply_PTF.xml:172:
    C:\ibm\wp85\PortalServer\jcr\wp.content.repository.install\config\includes\jcr.mig_cfg.xml:382: 
    Exception found when executing wsadmin: (hostname)
    c:/ibm/wp85/ConfigEngine/lib/wkplc.misc.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batchboot.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batch2.jar;
    C:/ibm/wp85/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/ibm/wp85/PortalServer/shared/app/wp.base.jar
  • Solution: If this issue is encountered, wait 5 minutes and re-run the upgrade-profile task with the same options to resume the migration. The lock on the file should be released and the uninstall can continue as expected.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (INSTANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page (See Creating page templates topic for more information). This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of the CF the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page. This is a permanent change in behavior effective in CF08 or later.

  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site/Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: Exception found when executing 
    wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'
    Ear File 
    C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear
    WASX7015E: Exception running command: "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; 
    exception information:
    com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: 
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file:
    • - Linux, AIX, Solaris: /etc/hosts
    • - Windows: %SystemRoot%\System32\drivers\etc\hosts and add a line similar to the following, substituting your actual machine name where indicated:
    127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
    • If a line for localhost already exists, you can add your hostname to this existing line.
    • - IBM i: To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.
  • Problem: After the CF is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque 
    W com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] 
    does not specify a blackwhite list when accessing resource [...], falling back to the default [[(null), (WEB-INF/.*)]]. 
    Applications can define a custom list by adding the keys [com.ibm.portal.resource.whitelist] and [com.ibm.portal.resource.blacklist] 
    to their web.xml deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 (or later) contained a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the res datasource. Details on the action required for the custom code mentioned in these messages can be found in the security bulletin. You must redeploy your custom theme for these changes to take effect.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the jsp components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting this on their support site: https://ephox.zendesk.com/entries/96347398-Java-1-8-u60-Internet-Explorer-Issue.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.io. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it.

  • Problem: Error messages can be seen in the SystemOut.log after rollback regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190)
  • Solution: These logging errors are expected if content was created against a Portal feature at an updated CF level and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables
  • Problem: Starting with HCL Portal 8.5.0 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title) (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal 8.5 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal 8.5 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal 8.5 Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: For IBM i only: May have the following error when running applyCF.sh when Portal databases have been transferred to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database 
    directory.
  • Solution: Run the following steps then rerun applyCF.sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed.
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Relational database alias: The hostname. Use the short TCP/IP hostname of the database system.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches:
         [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh
    --- Exception Thrown ---
    /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named EJBDeploy tool for pre-EJB 3.0 modules.

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing Portal 8.5 GA and before updating to the current CF, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply Portal 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall Portal 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall Portal 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 or later and retry the rollback. See this link for more details.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback.
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.

Known Issues for V8.5.0.0 Combined CF11 (PI60304)

  • Problem: Where Textbox.io is the configured rich text editor in WCM, rapidly opening and re-saving an existing item containing a rich text element can cause the loss of existing content if the item is saved before the rich text content is fully loaded. This is only likely to be an issue in environments with high network latency.
  • Solution: Ensure when editing an existing content, site area, template or component containing a rich text element, the rich text content is fully loaded in the editor before re-saving the item.

  • Problem: XMLAccess in the applyCF command may fail when updating a virtual portal.
  • Solution: The CF install attempts to update the Portal owned pages in the base portal and all virtual portals. Prior to CF11 there were some cases where error conditions were suppressed for both the base portal and all virtual portals. This has changed and error conditions are no longer suppressed. If the CF install (applyCF command) reports an error while running the XMLAccess tool on a virtual portal, there may be a problem with the virtual portal. If virtual portal(s) have been modified and removed some of the Portal owned pages on purpose then these errors can be safely suppressed and allow the CF install to continue. This can be done by adding the ignore-errors-all-vps property to the applyCF command. Example:
    applyCF.(sh|bat) ..... -Dignore-errors-all-vps='true'
    If Portal-owned pages have not been purposely removed from the virtual portal, then the error should be investigated and attempted to be fixed.

  • Problem: During the migration process the node may not start after upgrading to WAS 8.5.5.9 and before executing upgrade-profile. The node does not start with error message:
    ADMU3200I: Server launched. Waiting for initialization status.
    ADMU3011E: Server launched but failed initialization. Server logs, startServer.log, and other log files
    The log contains the following error message:
    HMGR0031E: A conflicting IP address and port has been detected for the DCS_UNICAST_ADDRESS end point.
    The node does not start due to conflicting port configuration for the DCS_UNICAST_ADDRESS end point of server1 and nodeagent in the node serverindex.xml.
  • Solution: It is recommended to execute upgrade-profile before upgrading to WAS 8.5.5.9. Alternatively the conflicting port configuration for the DCS_UNICAST_ADDRESS needs to be solved by setting unique port numbers for server1 and nodeagent.

  • Problem: After Rollback from CF09 or above to CF08 or below, the portal toolbar may show extra items that have titles $Banner$, $Teaser$ and $Video$.
  • Solution: These items can be ignored and should not be used. During an update to CF09 or above again, these items will disappear. To have these items no longer visible in the toolbar, go to Applications > Content > Web Content Authoring. Search for each of these pieces of content. Edit each of the pieces of content and go to properties. In the profile keyword field, remove the keyword ibm.portal.toolbar.NewContent. Save this change.

  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too long on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path.

  • Problem: During the upgrade-profile task in migration, there may be an intermittent issue which can cause the uninstall of an application to fail because files from the application are locked and cannot be deleted. This is an example of an error message that may be seen:
    BUILD FAILED
    C:\ibm\wp85\ConfigEngine\config\includes\apply_PTF.xml:172:
    C:\ibm\wp85\PortalServer\jcr\wp.content.repository.install\config\includes\jcr.mig_cfg.xml:382: 
    Exception found when executing wsadmin: (hostname)
    c:/ibm/wp85/ConfigEngine/lib/wkplc.misc.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batchboot.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batch2.jar;
    C:/ibm/wp85/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/ibm/wp85/PortalServer/shared/app/wp.base.jar
  • Solution: If this issue is encountered, wait 5 minutes and re-run the upgrade-profile task with the same options to resume the migration. The lock on the file should be released and the uninstall can continue as expected.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (INSTANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page (See Creating page templates topic for more information). This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of the CF the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page. This is a permanent change in behavior effective in CF08 or later.

  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site/Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: 
    Exception found when executing wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;
    C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'
    Ear File 
    C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear
    WASX7015E: Exception running command: "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; 
    exception information:
    com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: 
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file:
    • - Linux, AIX, Solaris: /etc/hosts
    • - Windows: %SystemRoot%\System32\drivers\etc\hosts and add a line similar to the following, substituting your actual machine name where indicated:
    127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
    • If a line for localhost already exists, you can add your hostname to this existing line.
    • - IBM i: To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.
  • Problem: After the CF is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] does not specify a 
    blackwhite list when accessing resource [...], falling back to the default [[(null), (WEB-INF/.*)]]. Applications can define a 
    custom list by adding the keys [com.ibm.portal.resource.whitelist] and [com.ibm.portal.resource.blacklist] to their web.xml 
    deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 (or later) contained a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the 'res' datasource. Details on the action required for the custom code mentioned in these messages can be found in the security bulletin. You must redeploy your custom theme for these changes to take effect.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the JSP components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting this on their support site: https://ephox.zendesk.com/entries/96347398-Java-1-8-u60-Internet-Explorer-Issue.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.io. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it.

  • Problem: Error messages can be seen in the SystemOut.log after rollback regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190)
  • Solution: These logging errors are expected if content was created against a Portal feature at an updated CF level and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables
  • Problem: Starting with HCL Portal 8.5 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal 8.5 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal 8.5 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal 8.5 Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: For IBM i only: May have the following error when running applyCF.sh when Portal databases have been transferred to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database directory.
  • Solution: Run the following steps then rerun applyCF.sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed.
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Relational database alias: The hostname. Use the short TCP/IP hostname of the database system.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches:
         [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh
    --- Exception Thrown ---
    /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named "EJBDeploy tool for pre-EJB 3.0 modules".

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing Portal 8.5 GA and before updating to the current CF, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply Portal 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall Portal 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall Portal 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 or later and retry the rollback. See this link for more details.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback.
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.

Known Issues for V8.5.0.0 Combined CF10 (PI55632)

  • Problem: The 8500 CF10 .zip file extraction contains a readme file for CF06.
  • Solution: The proper readme file for CF10 is also present and should be used. The readme file for CF06 is extraneous and can be disregarded.

  • Problem: During the migration process the node may not start after upgrading to WAS 8.5.5.9 and before executing upgrade-profile. The node does not start with error message:
    ADMU3200I: Server launched. Waiting for initialization status.
    ADMU3011E: Server launched but failed initialization. Server logs, startServer.log, and other log files
    The log contains the following error message:
    HMGR0031E: A conflicting IP address and port has been detected for the DCS_UNICAST_ADDRESS end point.
    The node does not start due to conflicting port configuration for the DCS_UNICAST_ADDRESS end point of server1 and nodeagent in the node serverindex.xml.
  • Solution: It is recommended to execute upgrade-profile before upgrading to WAS 8.5.5.9. Alternatively the conflicting port configuration for the DCS_UNICAST_ADDRESS needs to be solved by setting unique port numbers for server1 and nodeagent.

  • Problem: After Rollback from CF09 or above to CF08 or below, the portal toolbar may show extra items that have titles $Banner$, $Teaser$ and $Video$
  • Solution: These items can be ignored and should not be used. During an update to CF09 or above again, these items will disappear. To have these items no longer visible in the toolbar, go to Applications > Content > Web Content Authoring. Search for each of these pieces of content. Edit each of the pieces of content and go to properties. In the profile keyword field, remove the keyword ibm.portal.toolbar.NewContent. Save this change.

  • Problem: You may see Document Conversion Services errors (Stellent Conversion Error) in the log file as the system is indexing WCM content. The error may look like the following:
    SystemOut     O Stellent Conversion Error code:
    SystemOut     O I/O error while communicating with native process
  • Solution: The root of the problem is that there are some 32 bit libraries that are not available on your system. You can verify this by logging on to Portal Server machine and going to oiexport folder (e.g wp_profile_root/PortalServer/config/oiexport) and running task exporter. If you have this problem you will get error messages like:
    [root@(systemname) oiexport]# ./exporter
    -bash: ./exporter: /lib/ld-linux.so.2: bad ELF interpreter: No such file or directory
    The solution is to install all the missing 32 bit libraries that are listed when you run the exporter command. An explanation of this can be found here. Here are some examples, on a RedHat OS:
    yum install ld-linux.so.2
    yum install libstdc++-4.8.5-4.el7.i686  --setopt=protected_multilib=false
    yum install zlib-1.2.7-15.el7.i686  --setopt=protected_multilib=false
    yum install libgcc_s.so.1
    yum -y install glibc.i686
  • Problem: Clicking for help or to Learn More on some new functions may display "Topic not found".
  • Solution: This happens only on secondary nodes in a vertical cluster. The node contains outdated cached data for in-line help. To clear it:
    1. Go to this directory: (profile_root)/temp/(node_name)/(secondary_server). For example: /opt/IBM/WebSphere/wp_profile/temp/myhostNode/HCL Portal and HCL Web Content Manager_2
    2. Delete all files and sub-directories in this directory.
    3. Restart the IEHS_war application from the WAS administrative console. This is scheduled to be fixed in an upcoming CF.
  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too long on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path.

  • Problem: During the upgrade-profile task in migration, there may be an intermittent issue which can cause the uninstall of an application to fail because files from the application are locked and cannot be deleted. This is an example of an error message that may be seen:
    BUILD FAILED
    C:\ibm\wp85\ConfigEngine\config\includes\apply_PTF.xml:172: 
    C:\ibm\wp85\PortalServer\jcr\wp.content.repository.install\config\includes\jcr.mig_cfg.xml:382: Exception found when executing wsadmin: 
    (hostname)
    c:/ibm/wp85/ConfigEngine/lib/wkplc.misc.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batchboot.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batch2.jar;
    C:/ibm/wp85/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/ibm/wp85/PortalServer/shared/app/wp.base.jar
  • Solution: If this issue is encountered, wait 5 minutes and re-run the upgrade-profile task with the same options to resume the migration. The lock on the file should be released and the uninstall can continue as expected.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (INSTANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page (See Creating page templates topic for more information). This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of the CF the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page. This is a permanent change in behavior effective in CF08 or later.

  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site/Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: Exception found when executing 
    wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;
    C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'Ear File 
    C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear
    WASX7015E: Exception running command: "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; 
    exception information:
    com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: 
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file:
    • - Linux, AIX, Solaris: /etc/hosts
    • - Windows: %SystemRoot%\System32\drivers\etc\hosts and add a line similar to the following, substituting your actual machine name where indicated:
    127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
    • If a line for localhost already exists, you can add your hostname to this existing line.
    • - IBM i: To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.
  • Problem: After the CF is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] 
    does not specify a blackwhite list when accessing resource [...], falling back to the default [[(null), (WEB-INF/.*)]]. 
    Applications can define a custom list by adding the keys [com.ibm.portal.resource.whitelist] and [com.ibm.portal.resource.blacklist] 
    to their web.xml deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 (or later) contained a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the 'res' datasource. You must redeploy your custom theme for these changes to take effect.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the JSP components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting this on their support site: https://ephox.zendesk.com/entries/96347398-Java-1-8-u60-Internet-Explorer-Issue.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.io. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it.

  • Problem: Error messages can be seen in the SystemOut.log after rollback regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190)
  • Solution: These logging errors are expected if content was created against a Portal feature at an updated CF level and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables
  • Problem: Starting with HCL Portal 8.5.0 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title) (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal 8.5 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal 8.5 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal 8.5 Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: For IBM i only: May have the following error when running applyCF.sh when Portal databases have been transferred to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database directory.
  • Solution: Run the following steps then rerun applyCF.sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Relational database alias: The hostname. Use the short TCP/IP hostname of the database system.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches: [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh --- Exception Thrown --- /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named EJBDeploy tool for pre-EJB 3.0 modules.

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing Portal 8.5 GA and before updating to the current CF, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply Portal 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall Portal 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall Portal 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 or later and retry the rollback. See this link for more details.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback.
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.

Known Issues for V8.5.0.0 Combined CF09 (PI50956)

  • Problem: Clicking for help or to Learn More on some new functions may display "Topic not found".
  • Solution: This happens only on secondary nodes in a vertical cluster. The node contains outdated cached data for in-line help. To clear it:
    1. Go to this directory: (profile_root)/temp/(node_name)/(secondary_server). For example: /opt/IBM/WebSphere/wp_profile/temp/myhostNode/HCL Portal and HCL Web Content Manager_2.
    2. Delete all files and sub-directories in this directory.
    3. Restart the IEHS_war application from the WAS administrative console.
  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too long on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
    
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path.

  • Problem: The applyCF command may fail during the CF update process if the PortalAdminGroupId DN contains a space. The error may be similar to the following:
    [wsadmin] WASX7015E: Exception running command: 
    "source "/opt/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; exception information:
    [wsadmin] com.ibm.ws.scripting.ScriptingException: WASX7110E: Incorrect number of elements in option: "[Crawler, No, No, , 
    CN=CNNAME,OU=OUNAME, WITH, SPACES,OU=GROUPS,OU=NAME2, WITH, SPACES,OU=OUNAME,DC=DCNAME1,DC=DCNAME2]" for task: "MapRolesToUsers;" 
    number of elements should be 8.
    where for instance the original value of the DN is:
    CN=CNNAME,OU=OUNAME WITH SPACES,OU=GROUPS,OU=NAME2 WITH SPACES,OU=OUNAME,DC=DCNAME1,DC=DCNAME2
  • Solution: It is recommended to avoid using spaces in the DN name of the HCL Portal Administrator group. This is scheduled to be fixed in HCL Portal Version 8.5 CF10.

  • Problem: During the upgrade-profile task in migration, there may be an intermittent issue which can cause the uninstall of an application to fail because files from the application are locked and cannot be deleted. This is an example of an error message that may be seen:
    BUILD FAILED
    C:\ibm\wp85\ConfigEngine\config\includes\apply_PTF.xml:172: 
    C:\ibm\wp85\PortalServer\jcr\wp.content.repository.install\config\includes\jcr.mig_cfg.xml:382: 
    Exception found when executing wsadmin: (hostname)
    c:/ibm/wp85/ConfigEngine/lib/wkplc.misc.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batchboot.jar;
    C:/ibm/wp85/AppServer/deploytool/itp/batch2.jar;
    C:/ibm/wp85/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/ibm/wp85/PortalServer/shared/app/wp.base.jar
  • Solution: If this issue is encountered, wait 5 minutes and re-run the upgrade-profile task with the same options to resume the migration. The lock on the file should be released and the uninstall can continue as expected.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (INSTANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page (See Creating page templates topic on how to set it up.) This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of CF08 the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page.

  • Problem: After a CF update, it may be possible for anonymous or unauthenticated users to access the Site Toolbar, even if this access was restricted pre-CF update.
  • Solution: The access settings for the Site Toolbar would need to be modified again as the CF update reapplies default role mappings. The Customizing toolbar layouts topic shows how to set the access. This issue is scheduled to be fixed in HCL Portal Version 8.5 CF10.

  • Problem: It may not be possible to navigate away from the timeout screen to another Portal page once the timeout has activated.
  • Solution: The "Log In" link can be clicked again to simply access Portal. Alternatively, the following workaround is available:
    1. Modify the navigation.jsp of the custom theme as such:
      (a href="${fn:escapeXml(node.urlGeneration.setThemeTemplate(''))}" 
      class="wpthemeLeft ${node.isHidden ? 'wpthemeHiddenPageText' : ''} ${node.isDraft? 'wpthemeDraftPageText' : ''}">
      Note: The setThemeTemplate('') is added to the end of the node.urlGeneration.
    2. Update the theme ear with the new version, then restart the server and retest the session time out issue. This issue is scheduled to be fixed in 8500 CF10.
  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site/Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: 
    Exception found when executing wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'Ear File 
    C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear>
    WASX7015E: Exception running command: 
    "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; exception information:
    com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: 
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file:
    • - Linux, AIX, Solaris: /etc/hosts
    • - Windows: %SystemRoot%\System32\drivers\etc\hosts and add a line similar to the following, substituting your actual machine name where indicated.
    127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
    • If a line for localhost already exists, you can add your hostname to this existing line.
    • - IBM i: To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.
  • Problem: After the CF is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] 
    does not specify a blackwhite list when accessing resource [...], falling back to the default [[(null), (WEB-INF/.*)]]. 
    Applications can define a custom list by adding the keys [com.ibm.portal.resource.whitelist] and [com.ibm.portal.resource.blacklist] 
    to their web.xml deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 (or later) contained a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the res datasource. You must redeploy your custom theme for these changes to take effect.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the JSP components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting this on their support site: https://ephox.zendesk.com/entries/96347398-Java-1-8-u60-Internet-Explorer-Issue.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.io. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it.

  • Problem: Error messages can be seen in the SystemOut.log after rollback regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190)
  • Solution: These logging errors are expected if content was created against a Portal feature at and updated CF level and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables
  • Problem: A segment, stored in a segment group, is used to target content. This means that a content item is associated with this segment in a content targeting rule. When this content targeting rule is executed, i.e. someone loads the page containing that web content viewer portlet, the segment group is cached by Personalization and thereafter the cached segment group is used until the cached entry times out. Any edits to the segment group do not invalidate the cache so therefore any updates to any segments will not be reflected in the content targeting execution until the cached entry has expired due to timeout, which is approximately 5 minutes by default.
  • Solution: This current limitation will be addressed in a future HCL Portal Version 8.5 CF cycle. The cache time out can be changed in the Personalization settings file.

  • Problem: Starting with HCL Portal 8.5 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title) (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal 8.5 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal 8.5 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal 8.5 Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: For IBM i only: May have the following error when running applyCF.bat|sh when Portal databases have been transfer to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database directory.
  • Solution: Run the following steps then rerun applyCF.bat|sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed.
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Relational database alias: The hostname. Use the short TCP/IP hostname of the database system.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches:
         [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh
    --- Exception Thrown ---
    /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named EJBDeploy tool for pre-EJB 3.0 modules.

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing Portal 8.5 GA and before updating to CF01, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply Portal 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall Portal 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall Portal 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 or later and retry the rollback.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback.
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.

Known Issues for V8.5.0.0 Combined CF08 (PI45447)

  • Problem: During the update on IBM i, the following error may occur:
    --- Exception Thrown ---
    WKSP0006E Error while saving context -java.io.IOException: The system cannot find the specified file, either the filename is too long on Windows system or run out of file descriptor on UNIX platform. java.io.FileNotFoundException: /QIBM/UserData/WebSphere/AppServer/V85/ND/profiles/wp_profile/wstemp/anonymous1358164268290/workspace/templates/.repositoryContext (The file access permissions do not allow the specified action.)
  • Solution: If this exception occurs after configuration has already successfully completed, then it can be safely ignored. However, if it causes the update to fail, then you will need to change the location of the wstemp directory to shorten the path. See Changing the location of the wstemp temporary workspace directory for instructions.

  • Problem: It may be noticed that a page created from a page template does not have the access controls that are expected or that the access controls for such a page have recently changed.
  • Solution: APAR PI39231 (INSTANTIATING A PAGE TEMPLATE LEADS TO INCONSISTENT ACCESS CONTROL SETTINGS ON THE TARGET PAGE) was introduced in CF08. With this APAR, access control settings applied to a page template are consistently copied onto the target page (See Creating page templates topic learn more). This means that all access rights assigned explicitly on a page template are also applied to the created page. If, for example, the All Authenticated Portal Users group is assigned to USER@PageTemplate, then the created page will be visible to every authenticated user, independent of the access control configuration of the parent page. During the installation of CF08 the out of the box page templates have also been changed, including those in existing virtual portals, to no longer have explicit role mappings and it is recommended that all customers do the same for custom code templates unless the use case really is to copy that role mapping onto the target page.

  • Problem: After a CF update, it may be possible for anonymous or unauthenticated users to access the Site Toolbar, even if this access was restricted pre-CF update.
  • Solution: The access settings for the Site Toolbar would need to be modified again as the CF update reapplies default role mappings. Click here to learn how to set the access. This issue is scheduled to be fixed in HCL Portal Version 8.5 CF10.

  • Problem: It may not be possible to navigate away from the timeout screen to another Portal page once the timeout has activated.
  • Solution: The "Log In" link can be clicked again to simply access Portal. Alternatively, the following workaround is available:
    1. Modify the navigation.jsp of the custom theme as such:
      (a href="${fn:escapeXml(node.urlGeneration.setThemeTemplate(''))}" 
      class="wpthemeLeft ${node.isHidden ? 'wpthemeHiddenPageText' : ''} ${node.isDraft? 'wpthemeDraftPageText' : ''}")
      Note: The setThemeTemplate('') is added to the end of the node.urlGeneration.
    2. Update the theme ear with the new version, then restart the server and retest the session time out issue. This issue is scheduled to be fixed in HCL Portal Version 8.5 CF10.
  • Problem: Non-English characters are not supported when exporting a Site/Section template from Site Builder. When importing Site/Section templates into the Site Builder application, you may encounter an error that the template PAA file could not be imported. This could be due to using non-English characters in the Site/Section template PAA import object file name. You will see the following messages on the site builder interface:
    "Could not import the template. The solution installer import of the PAA file failed. Review the server logs for more information."
  • Solution: Use only English characters in the PAA file name.

  • Problem: If local hostname resolution is not configured properly, the applyCF command may fail with an error like this:
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\config\includes\wp.setup.themedev_cfg.xml:52: 
    Exception found when executing wsadmin: jacl
    C:/IBM/WebSphere/ConfigEngine/lib/wkplc.misc.jar;C:/IBM/WebSphere/AppServer/deploytool/itp/batchboot.jar;
    C:/IBM/WebSphere/AppServer/deploytool/itp/batch2.jar;C:/IBM/WebSphere/PortalServer/base/wp.base/shared/app/wp.base.jar;
    C:/IBM/WebSphere/PortalServer/shared/app/wp.base.jar
    'HCL Portal and HCL Web Content Manager(cells/winwab09Cell/nodes/winwab09Node/servers/HCL Portal and HCL Web Content Manager|server.xml#Server_1452022426983)'
    Ear File 
    C:/IBM/WebSphere/PortalServer/theme/wp.setup.themedev/installedApps/ThemeDevAssets.ear
    WASX7015E: Exception running command: 
    "source "C:/IBM/WebSphere/wp_profile/ConfigEngine/config/work/was/curJaclScript.jacl""; exception information:
    com.ibm.websphere.management.filetransfer.client.TransferFailedException
    java.io.FileNotFoundException: java.io.FileNotFoundException: 
    C:\IBM\WebSphere\PortalServer\theme\wp.setup.themedev\installedApps\ThemeDevAssets.ear (Access is denied)
  • Solution: This indicates a SOAP issue on the Web server where the hostname does not resolve as expected. The solution is to make sure that all host names of the machine are associated with the loopback in the hosts file. Edit the following file:
    • - Linux, AIX, Solaris: /etc/hosts
    • - Windows: %SystemRoot%\System32\drivers\etc\hosts and add a line similar to the following, substituting your actual machine name where indicated:
    127.0.0.1 yourhostname yourhostname.yourdomain localhost localhost.localdomain
    • If a line for localhost already exists, you can add your hostname to this existing line.
    • - IBM i: To update the hosts table on IBM i, run the cfgtcp command and choose the option Work with TCP/IP host table entries.

  • Problem: After HCL Portal Version 8.5 CF08 is applied, warning messages may be seen in the SystemOut.log similar to:
    [10/5/15 8:00:00:000 EDT] 0000000a AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/...] does not specify a blackwhite list when accessing resource [...], falling back to the default [[(null), (WEB-INF/.*)]]. Applications can define a custom list by adding the keys [com.ibm.portal.resource.whitelist] and [com.ibm.portal.resource.blacklist] to their web.xml deployment descriptor.
    There might be many of these warning messages which could cause a significant performance issue.
  • Solution: HCL Portal Version 8.5 CF08 contains a security fix (PI47714 / CVE-2014-8912), which requires specification of black and white lists for custom themes and portlets in certain usage scenarios involving the res datasource. You must redeploy your custom theme for these changes to take effect.

  • Problem: After HCL Portal Version 8.5 CF08 is applied, warning messages related to /wps_semanticTag may be seen in the SystemOut.log similar to:
    [10/28/15 14:55:52:343 EDT] 000000f9 AbstractReque W 
    com.ibm.wps.resolver.resource.AbstractRequestDispatcherFactory matchesWebAppDefault(aResource) Servlet context [/wps_semanticTag]...
  • Solution: HCL Portal Version 8.5 CF08 contains a security fix (PI47714 / CVE-2014-8912), and the black and white list specification for this out of the box content was missing. This will be addressed by HCL Portal Version 8.5 CF09. As a temporary workaround, the following custom properties in the Resource Environment Provider WP ConfigService can be used:
    Name Value
    com.ibm.portal.resource.wps_semanticTag.contextroot /wps_semanticTag
    com.ibm.portal.resource.wps_semanticTag.whitelist .*
    com.ibm.portal.resource.wps_semanticTag.blacklist WEB-INF/.*
  • Problem: After applying HCL Portal Version 8.5 CF08, it may be noted that the search crawler fails with a NPE after content is created underneath a Portal page via the new Site Manager.
  • Solution: An iFix has been produced for this issue and can be applied on top of 8500 CF08. This is planned to be fixed in the next HCL Portal Version 8.5 CF. The iFix to use is 8.5.0.0-WCM-IFPI51020-CF08.zip. You can find more about this on the HCL Software Support page.

  • Problem: When following steps to modify site URLs for the Search Engine Optimization (SEO) workflow some portlets and labels may not be available in clustered environments. For example, you cannot perform the Change the JSP components in the Web Resources v70 Library step.
  • Solution: If you do not have a web server, you must resynchronize the nodes and restart the cluster before you can change the JSP components in the Web Resources library.

  • Problem: There may be a performance issue seen (in EditLive) when using Java 1.8 Update 60 / Internet Explorer.
  • Solution: Oracle and Ephox have acknowledged this issue and are documenting this on their support site: https://ephox.zendesk.com/entries/96347398-Java-1-8-u60-Internet-Explorer-Issue.
    Note: The Ephox EditLive! Java based Rich Text Editor has been deprecated as of HCL Portal 8.5 CF11 and replaced by Textbox.io. HCL Portal 9.5 requires a minimum of CF17 and thus offers Textbox.io. The Ephox EditLive! feature is only available in HCL Portal 9.5 if you explicitly enable it.

  • Problem: Error messages can be seen in the SystemOut.log after rollback from CF08 regarding ManagedRules feature updates. These features would no longer be available after the rollback to a level earlier than CF08. The following error example may be seen in the log:
    java.lang.RuntimeException: java.lang.ClassNotFoundException: com.aptrix.pluto.pzn.Segment
    at com.aptrix.identity.IdentityReference.getTypeClass(IdentityReference.java:190) 
  • Solution: These logging errors are expected if content was created against a feature of 8500 CF08 and a rollback occurs to a level that does not support it. The logging exceptions could be removed by running the following two ConfigEngine tasks:
    ./ConfigEngine.sh drop-wcm-recents-and-favorites-tables
    ./ConfigEngine.sh create-wcm-recents-and-favorites-tables
  • Problem: A segment, stored in a segment group, is used to target content. This means that a content item is associated with this segment in a content targeting rule. When this content targeting rule is executed, i.e. someone loads the page containing that web content viewer portlet, the segment group is cached by Personalization and thereafter the cached segment group is used until the cached entry times out. Any edits to the segment group do not invalidate the cache so therefore any updates to any segments will not be reflected in the content targeting execution until the cached entry has expired due to timeout, which is approximately 5 minutes by default.
  • Solution: This current limitation will be addressed in a future HCL Portal Version 8.5 CF cycle. The cache time out can be changed in the Personalization settings file.

  • Problem: Starting with HCL Portal 8.5 Cumulative Fix 5, a new check has been added to the XML configuration interface. When the XML configuration interface processes update requests, now all locale identifiers are checked if they represent a locale available on the system. If this is not the case, then the update request fails and the XML configuration interface returns with an error. For example, language-dependent strings are marked with a locale:
    {code:} (localedata locale="en") (title)Home(/title) (/localedata){code}
    Here, for en which stands for the English locale, a title is defined. For other locales, other strings can be defined. Now, if the attribute locale contains an empty value or an identifier of a locale which does not exist on the system, then the file cannot be imported. As on levels prior to HCL Portal 8.5 Cumulative Fix 5, this check has not been performed, XML files exported from Portal systems can lead to errors when trying to import them on a HCL Portal 8.5 system with at least Cumulative Fix 5. Even HCL Portal systems migrated or upgraded to at least HCL Portal 8.5 Cumulative Fix 5 can contain configuration data with invalid locales. Exporting such data with the XML configuration interface is possible without problem. However, trying to import this data can lead to the error described above.
  • Solution: To fix this problem you can remove the XML element localedata with its child elements or you can change the locale attribute to an existing locale identifier.

  • Problem: The upgrade of HCL Script Application PAA, from a previous version, may fail to import the new WCM HCL Script Application Library. The following error may be seen:
    Invalid XML. javax.jcr.ItemExistsException: RT7028E: A node or property already exists at path:
    /contentRoot/icm:libraries[13]/Content/HCL Script Application library site area/HCL Script Application content item
    The WCM Library Import task may fail if importing a library over an existing library with the same library and artifact names, but where the unique id of one of the artifacts (component, content item, etc.) differs.
  • Solution: Investigate the ConfigTrace.log and Portal SystemOut.log and SystemErr.log for details of the cause of the error. Determine the HCL Script Application Library artifact causing the collision, via the error information obtained from the logs, if possible, and delete or rename that artifact and then try the install again. To avoid seeing the issue, it is recommended to not delete or rename and then recreate any of the out of the box artifacts that are installed with the HCL Script Application Library.

  • Problem: For IBM i only: May have the following error when running applyCF.bat|sh when Portal databases have been transfer to IBM i DB2 using type 4 driver:
    [importNodeTypes] SEVERE: RT0022E: Failed to initialize pool with minimum number of free connections
    [importNodeTypes] com.ibm.db2.jdbc.app.DB2SQLSyntaxErrorException: Relational database "hostname"  not in relational database directory.
  • Solution: Run the following steps then rerun applyCF.bat|sh:
    1. Start a 5250 session on the database machine.
    2. Type the IBM i command WRKRDBDIRE to display the Relational Database Directory Entry for Location *LOCAL and make a note of the value displayed.
    3. Sign off from the 5250 session.
    4. Start a 5250 session on the machine where HCL Portal is installed.
    5. Create a Relational Database Directory Entry on the system where HCL Portal is installed for the database system using IBM i command WRKRDBDIRE.
    6. Add an entry with the following values:
      • - Relational database: The relational database. Use the value noted from the prior step.
      • - Relational database alias: The hostname. Use the short TCP/IP hostname of the database system.
      • - Remote location: The domain qualified hostname. Use the full TCP/IP hostname of the database system.
      • - Type: IP
      • - Port number or service name: DRDA
      • - Remote authentication method: Preferred method: ENCRYPTED
      • - Allow lower authentication: ALWLOWER
  • Problem: The drag and drop feature of the Toolbar does not work with Microsoft Windows Internet Explorer version 9.
  • Solution: The Microsoft Windows Internet Explorer issue will be resolved in the next cumulative fix.

  • Problem: The Live Repository will not show any available packages to be updated when selecting the IIM Preferences checkbox Search service repositories during installation and updates.
  • Solution: IBM Installation Manager (IIM) should be updated to the latest level, at least IIM 1.8, for the packages to be displayed.

  • Problem: The applyCF.bat|sh task may fail with an error like the following:
    action-clear-was-caches:
         [echo] /usr/IBM/WebSphere/AppServer/deploytool/scripts/install/ejbdeploy-clear-cache.sh
    --- Exception Thrown ---
    /usr/IBM/WebSphere/PortalServer/installer/wp.update/config/includes/cumulative_fix.xml:302: The directory you specified does not exist. 
  • Solution: Portal Server requires the optional WebSphere Application Server EJBDeploy tool, and the above error will occur if this tool is not installed. To resolve the error, use IBM Installation Manager to modify your copy of WebSphere Application Server and add the feature named EJBDeploy tool for pre-EJB 3.0 modules.

  • Problem: The IBM Installation Manager GUI may hang or crash on Red Hat Enterprise Linux v6.6 during the Portal update process.
  • Solution: There are potentially two issues at work. The first is an eclipse bug caused by an infinite loop in code that tries to enumerate printers. This first issue can be resolved by updating your Installation Manager client to v1.8 or later. If updating your client is not a viable option at the moment, you can alternatively work around the problem by adding the following line to the /opt/IBM/InstallationManager/eclipse/IBMIM.ini file:
    -Dorg.eclipse.swt.internal.gtk.disablePrinting
    The second issue is caused by a different eclipse bug, related to a mismatch between the GTK version and the cairo library. The resolution is to update the cairo libraries.

  • Problem: Starting Portal Farm node may result in Error 500 when accessing Portal.
  • Solution: Go and log in to HCL Software Support page to find related information about this issue.

  • Problem: If you do a database transfer to z/OS DB2 after installing Portal 8.5 GA and before updating to CF01, the schema name used for JCR indices will be the config user name instead of the designated schema name, resulting in potential duplicate entries if more than one portal uses the same z/OS DB2. The schema name is supposed to be sufficient for uniquely configuring multiple Portals to the same z/OS DB2 and it is very possible that the same config user could be used for both Portals.
  • Solution: Apply Portal 8.5 CF01 or later prior to doing a database transfer to z/OS DB2.

  • Problem: After installing and rolling back the Cumulative Fix, you may observe error messages similar to the following in the Installation Manager if you subsequently uninstall Portal 8.5:
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    Basedir /usr/IBM/WebSphere/wp_profile does not exist
    org.apache.tools.ant.Project.setBaseDir(Project.java:844)
    org.apache.tools.ant.Project.setBasedir(Project.java:829)
  • Solution: The product will be successfully uninstalled despite these messages, so you can safely choose to ignore them. The errors will be avoided if you uninstall Portal 8.5 without first rolling back the Cumulative Fix.

  • Problem: IBM Installation Manager version 1.7.1 crashes with JVM Error when running HCL Portal rollback in GUI Mode.
  • Solution: Upgrade Installation Manager to version 1.7.3 and retry the rollback.

  • Problem: After installing or rolling back the cumulative fix, you may observe error messages similar to the following in the Installation Manager log file:
    Installation Manager cannot remove feature Portal Server Profile during the Update or Rollback.
  • Solution: These messages do not indicate a problem. To prevent a configuration error, you are prevented from changing your feature selections when performing an update. If you want to add or drop a portal, you must select the Modify function of Installation Manager instead.