HP Business Availability Center 8.07 Readme
for the Windows and Solaris operating systems
Software version: 8.07
Publication date: July 2011
This file provides information about HP Business Availability Center 8.07. It also includes relevant information for all previous 8.0x versions.
Separation of Discovery and Dependency Mapping (DDM) Content into Content Packs
8.07 Version Content and Limitations
8.06 Version Content and Limitations
8.05 Version Content and Limitations
8.04 Version Content and Limitations
8.03 Version Content and Limitations
8.02 Version Content and Limitations
8.01 Version Content and Limitations
HP Business Availability Center Support Matrixes
Documentation Updates
The first page of this release notes document contains the following identifying information:
To check for recent updates or to verify that you are using the most recent edition, visit this URL: HP Software Product Manuals
To retrieve a document, select the:
1. Product name.2. Version list.3. Operating System.4. Preferred Language.5. Document title.6. Click Open or Download.You must have Adobe Reader installed to view files in PDF format (*.pdf). To download Adobe Reader, go to the Adobe web site.
What's New
To view a list of new features for this release, select Help > What's New in HP Business Availability Center, or double-click whatsnew.html from the root directory of the HP Business Availability Center DVD or delivery package.
System Requirements
To view a complete list of system requirements, see the HP Business Availability Center Deployment Guide. Selected Business Availability Center 8.07 support matrixes are located here.
Service Pack Files/Components
Business Availability Center 8.07 includes the following files/components:
Business Availability Center 8.07 setup files
Updated Components
The following updated components have been released in parallel to Business Availability Center 8.07. They must be downloaded separately from the software patches download site.
For instructions on upgrading the Real User Monitor engine, see Upgrade_RUM.pdf, located in the 8.07 Setup package root directory.
For instructions on upgrading the Real User Monitor probe, see Upgrade_RUM.pdf, located in the 8.07 Setup package root directory.
For instructions on installing Business Process Monitor, see the Business Process Monitor Administration Guide.
To install the Discovery Probe, first stop any existing Discovery Probe instances (if you have a previous version running). Uninstall the existing Discovery Probe via Start > Programs > UCMDB or HP DDM > Uninstall UCMDB Discovery Probe or Uninstall HP DDM Probe. Verify that the Probe installation folder (hp\DDM\DiscoveryProbe) is empty and that the MySQL service has been removed (deleted from Windows services list). Then, run HPDiscoveryProbe_v807_win32.exe. Finally, start the 8.07 Discovery Probe.
Note: When DDM Probe is installed on Windows Server 2008, you must use DDM Content Pack 5 or later (older DDM Content Packs are not supported). To use DDM Content Pack 5 on Windows Server 2008, you must apply the DDM_on_Windows2008_ Patch_for_CP5 patch on top of DDM CP5. The patch is provided with the DDM 8.06 package. To apply the patch on top of DDM CP5, deploy the DDM_on_Windows2008_ Patch_for_CP5.zip file through the Package Manager.
For instructions on installing SiteScope, see the SiteScope Deployment Guide.
For instructions on installing Diagnostics, see the Diagnostics Installation and Configuration Guide.
Documentation files
- whatsnew.html
- readme807.html
- pre_upgrade_tool.pdf (a guide to understanding and using the pre-upgrade tool, which helps you prepare your BAC 8.x environment for upgrading to Business Service Management 9.10)
- BAC_8.05_new_content.pdf (addendum to the 8.05 readme, containing in-depth information on certain new functionality that was released in Business Availability Center 8.05)
The document contains information on: the SLM Web Services API, an integration tool for Service Level Management, enabling administration of agreements from outside of Service Level Management; how to modify CMDB connection settings for the Dashboard Changes & Incidents tab; how to configure a Notification SNMP Trap for a CI status alert.
- new_content_803.pdf (addendum to the 8.03 readme, containing in-depth information on new functionality that was released in Business Availability Center 8.03 and is also included in 8.05)
- Upgrade_RUM.pdf (contains instructions on upgrading the Real User Monitor probe and engine)
- HPBAC_804_VMware_Performance_FR_20_7_10.pdf (contains a benchmarking study on Business Availability Center (BAC) 8.x deployment on VMware ESX Server 4.0)
- sms_integration.pdf (provides details on integration between HP Universal CMDB and Microsoft Systems Management Server (SMS)).
- updates to Business Availability Center documentation set
Installation Prerequisites
Business Availability Center 8.07 must be installed on top of Business Availability Center 8.01, 8.02, 8.03, 8.04, 8.05, or 8.06. Customers who do not yet have 8.01 installed must install it on top of Business Availability Center 8.00. For specific details on Business Availability Center 8.00 or 8.01, see the readme file included with the 8.00 or 8.01 package.
Note: The requirement to install Business Availability Center 8.01 as a prerequisite is due to the separation of updated Discovery and Dependency Mapping content into separate content packs. For details, see the next section.
Separation of Discovery and Dependency Mapping (DDM) Content into Content Packs
Discovery and Dependency Mapping (DDM) content packs include new and updated packages needed for activating DDM jobs. The packs also include documentation of the procedures for running the DDM jobs.
The Business Availability Center 8.01 release contained software changes to support the separation of service pack releases from DDM content pack releases. Thus, Business Availability Center 8.01 is a prerequisite to enabling the separation of service pack versus content pack installation.
As a result of the separation of service pack from content pack, starting from Business Availability Center 8.01, customers can choose whether to update their DDM content via content pack release regardless of which service pack release is installed.
Note: DDM Content Pack 1 is built into Business Availability Center 8.01 (there is no separate Content Pack 1 installation).
Note: DDM Content Pack 2 is not supported by Business Availability Center (it is not possible to install Content Pack 2 on top of Business Availability Center).
Installation
Read all the information in this Installation section (including the Upgrade Notes and Limitations section) before starting the installation.
- If you are currently running a version of Business Availability Center earlier than 8.01, you must install 8.01 before installing 8.07 (see instructions below). If you are installing 8.01, directly followed by 8.07, you do not need to repeat the pre-installation procedures (file backups or log setting modifications) listed below before installing 8.07. Further, you should perform all the post-installation procedures after installing 8.07.
- Business Availability Center 8.07 is supported on Windows Server 2008 (see matrix for exact version list). See below for instructions on how to install Business Availability Center on Windows Server 2008.
- Upgrade path summary (for complete details, see subsequent instructions):
- Installation on VMware: Business Availability Center 8.04 was deployed and run in a virtual environment and the benchmark performance results are available in the document HPBAC_804_VMware_Performance_FR_20_7_10.pdf found in the same location as this Readme file. These results are applicable for all 8.x versions.
- Pre-Installation Note: Before installing a new version of Business Availability Center, it is recommended that you back up all Business Availability Center databases (management, CMDB, CMDB history, profile, BPI).
- Pre-Installation Note: Before starting Business Availability Center 8.07 installation, make sure Business Availability Center has been fully stopped on all machines and that there are no open connections (for example, via Windows Explorer) from any machines to the Business Availability Center root directory or any of its subdirectories.
- Pre-Installation Note: If you are installing Business Availability Center 8.07 on machines running Solaris OS, you must install under the "root" user. Then, after installation, change ownership of the /opt/HPBAC directory and all its subdirectories to your Business Availability Center user using the following command: chown -R <BAC_user> /opt/HPBAC
- 8.01/8.02/8.03/8.04/8.05/8.06 à 8.07: If you have Business Availability Center 8.01 or later installed, stop Business Availability Center on all servers and make sure it is stopped before installing the service pack. Install the 8.07 service pack by double-clicking the setup.exe file for Windows (or by running the setup.bin script for Solaris) and following the on-screen instructions. Do this on every Business Availability Center server. After installing 8.07 on all servers, run the database configuration utility to reconnect to existing databases (as per instructions in the Deployment Guide), then restart Business Availability Center on all servers.
- 8.00 à 8.01 à 8.07: If you have Business Availability Center 8.00 installed, stop Business Availability Center on all servers and make sure it is stopped before installing the service pack. Install the 8.01 service pack by double-clicking the setup.exe file for Windows (or by running the setup.bin script for Solaris) and following the on-screen instructions. Do this on every Business Availability Center server. Then install the 8.07 service pack by double-clicking the setup.exe file for Windows (or by running the setup.bin script for Solaris) and following the on-screen instructions. Do this on every Business Availability Center server. After installing 8.07 on all servers, run the database configuration utility to reconnect to existing databases (as per instructions in the Deployment Guide), then restart Business Availability Center on all servers.
7.x à 8.00 à 8.01 à 8.07: If you do not yet have Business Availability Center 8.00 or 8.01 installed, install 8.00 according to the instructions in the HP Business Availability Center 8.07 Deployment Guide (this is a more updated version of the 8.00/8.01 Deployment Guide). After installing 8.00, do not run the database configuration process or start Business Availability Center until after relevant service packs are installed. After installing Business Availability Center 8.00, install 8.01, and then 8.07 on every Business Availability Center server as per instructions above. After 8.07 is installed, run the upgrade wizard and start servers as per instructions in the Deployment Guide.
- Windows Server 2008: If you plan to install Business Availability Center on machines running Windows Server 2008, follow the instructions below. Note that if you are currently running Business Availability Center 8.0x on Windows Server 2003 and want to upgrade to 8.07 and migrate to Windows Server 2008, you will need to reinstall Business Availability Center on the Windows Server 2008 machines using the full 8.00 à 8.01 à 8.07 installation path described below. Note further that because Windows Server 2008 comes with IIS 7 or 7.5, to enable Business Availability Center 8.07 to work correctly with IIS 7.x, certain roles must be defined in IIS running on the Gateway Server. This is necessary due to changes made by Microsoft in the IIS configuration APIs between IIS 6 and IIS 7, and is not connected to runtime factors. The relevant steps are also outlined below:
- To enable Business Availability Center 8.07 to work with IIS 7.x, perform the following steps on all Windows Server 2008 machines running the Business Availability Center Gateway Server:
- Open the Server Manager window (for example, via Start > Settings > Control Panel > Programs and Features > Turn Windows features on or off).
- In the Server Manager tree, select Web Server IIS under the Roles node.
- Locate the Role Services section, and click Add Role Services to open the Select Role Services wizard.
- Using the wizard, add the following roles: ISAPI Extensions; ISAPI Filters; IIS 6 Management Compatibility (including all child roles)
- Install Business Availability Center 8.00 according to the instructions in the HP Business Availability Center 8.07 Deployment Guide (this is a more updated version of the 8.00/8.01 Deployment Guide) with the following exception: do not run the 8.00 setup file included with any 8.00 distribution media (DVD, 8.00 download, and so on). Rather, copy the BAC80Setup_on_Win2008.bat file-located in the 8.07 package in the Win_2008_Setup directory-to the 8.00 setup directory (the same directory where the 8.00 setup.exe file is located). Then double-click the BAC80Setup_on_Win2008.bat file to start the setup process. If your Business Availability Center 8.00 setup files are located on a DVD and not on the network, launch the setup.exe file using the following command line string:
<path to setup.exe file>\setup.exe -W checkRequirementsSequence.active=false
After installing 8.00, do not run the database configuration process or start Business Availability Center until after relevant service packs are installed.
- Install 8.01 by copying the BAC801Setup_on_Win2008.bat file-located in the 8.07 package in the Win_2008_Setup directory-to the 8.01 setup directory (the same directory where the 8.01 setup.exe file is located). Then double-click the BAC801Setup_on_Win2008.bat file to start the setup process. Do this on every Business Availability Center server.
- Install the 8.07 service pack by double-clicking the 8.07 setup.exe file for Windows and following the on-screen instructions. Do this on every Business Availability Center server.
- On the Data Processing Server(s), create temp directories in the following directories:
- After 8.07 is installed, run the database configuration utility or upgrade wizard (depending on whether you are installing 8.x as a new installation or whether you are upgrading from an earlier Business Availability Center version) and start servers as per instructions in the Deployment Guide.
- OMi Note: The following note is relevant for OMi 8.10 customers upgrading to Business Availability Center 8.07.
If you are running OMi 8.10 together with Business Availability Center 8.07, you must install the latest OMi 8.10 patch on top of your OMi 8.10 instance after installing 8.07. Please check the patch documentation for installation instructions. You can download patches from the HP SSO patches site (http://support.openview.hp.com/selfsolve/patches)
Upgrade Notes and Limitations
- Login/Authentication: If you are upgrading to Business Availability Center 8.07 from an earlier version and are using Lightweight SSO (the default authentication method), after 8.07 installation, you must immediately change the initString default value for all HP software applications that are integrated using LW-SSO.
- In applications other than Business Availability Center, locate the lwssofmconf.xml file and change the value directly in that file.
- In Business Availability Center, you can do this directly in the <HP Business Availability Center root directory>\conf\settings\SingleSignOn\lwssofmconf.xml file (only before the first time the service is loaded) or using the JMX Console by searching for the LW-SSO Configuration service and changing the initString value. You can also change the value in Business Availability Center using t he Authentication Management Wizard (Admin > Platform > Users and Permissions > Authentication Management).
For details on initString values, see the "LW-SSO Security Warnings" section of the Platform Administration guide.
- Login/Authentication: If you have disabled Lightweight Single Sign-On in HP Business Availability Center 7.5x and then upgrade to HP Business Availability Center 8.0x, login to HP Business Availability Center fails. To enable login, you must set the SSO Mode to Disabled in the HP Business Availability Center SSO JMX Console (Topaz:SSO) after upgrading. To do so:
- Access the HP Business Availability Center SSO JMX Console by navigating to the following URL (log in using JMX credentials): http://<Gateway Server>:8080/jmx-console/HtmlAdaptor?action=inspectMBean&name=Topaz%3Aservice%3DSSO, where <Gateway Server> is the name of the machine on which HP Business Availability Center Gateway Server is running.
- In the setSingleSignOnMode() MBean, type Disabled in the ParamValue box and click Invoke.
- Login/Authentication: Business Availability Center configured for secure access (Apache + SSL) uses a customized version of the file httpd-ssl.conf. Upgrade to Business Availability Center 8.07 may overwrite this with a standard version of the file, breaking the configuration.
Workaround: Before upgrading, backup the customized file <HP Business Availability Center root directory>\WebServer\conf\extra\httpd-ssl.conf. Then copy it back to its original location after completing the upgrade. (QCCR1I41178).
- Login/Authentication: If you are upgrading from a 7.0x or 7.5x version of Business Availability Center and are accessing via reverse proxy, you may not be able to log into Business Availability Center. Workaround:
- Install 8.00, 8.01, 8.07, and run the full upgrade procedure in either staging or direct mode.
- Before logging into Business Availability Center for verifications, disable the reverse proxy as follows for the upgrade environment:
- Access the JMX console via http://<HP Business Availability Center machine>:8080/jmx-console and enter the user name and password.
- Click service=Infrastructure Settings Manager.
- Invoke the void setGlobalSettingValue() method with the following parameter values:
contextName=platform
settingName=enable.reverse.proxy
newValue=false- Check that you can now log into Business Availability Center. If you still cannot log in, remove the values from the reverse proxy URL settings as follows:
- Access the JMX console via http://<HP Business Availability Center machine>:8080/jmx-console and enter the user name and password.
- Click service=Infrastructure Settings Manager.
- Invoke the void setGlobalSettingValue() method with the following parameter values:
contextName=platform
settingName=default.centers.server.url
newValue=<leave empty>- Invoke the void setGlobalSettingValue() method again but with the following parameter values:
contextName=platform
settingName= default.core.server.url
newValue=<leave empty>- Access Business Availability Center directly and not through the reverse proxy. If you upgraded in staging mode, proceed with your verifications of the upgrade environment. If verified, declare the upgrade environment as the new production environment.
For details on configuring the reverse proxy again, see "Using a Reverse Proxy in HP Business Availability Center" in the Business Availability Center Hardening Guide.
- Login/Authentication: When upgrading to 8.07 from any version prior to 8.04, the LDAP authentication, or any feature that depends on certificates imported into Business Availability Center's cacerts file, fails after upgrade, due to the upgrading of the JRE version. (This issue is not applicable if you are currently running version 8.04 or later and upgrading to 8.07.) (QCCR1I45208)
Workaround: Before beginning the upgrade process to Business Availability Center 8.07 from any version prior to 8.04, backup the certificate files:
<Business Availability Center root directory>\jre\lib\security\cacerts
and
<Business Availability Center root directory>\jre\lib\security\jssecacertsAfter upgrade, paste the backed up certificate files into:
<Business Availability Center root directory>\jre\lib\securityAlias /topaz/WEB-INF "<BAC PATH>\HPBAC/AppServer/webapps/site.war/WEB-INF/"<Directory "<BAC PATH>\HPBAC/AppServer/webapps/site.war/WEB-INF/">#Options Indexes FollowSymLinksOptions FollowSymLinksAllowOverride NoneOrder allow,denyAllow from all<LimitExcept POST GET HEAD>Require valid-user</LimitExcept><FilesMatch "\.(sh|war|ear|bak|xml)$">Order allow,denyDeny from all</FilesMatch></Directory>
- Administration: During the upgrade, the log file <HP Business Availability Center install directory>\log\EJBContainer\upgradeFramework.log may remain empty, though the following other log files catch the log input: <HP Business Availability Center install directory>\log\EJBContainer\admincenter.ejb.log and <HP Business Availability Center install directory>\log\topaz_all.ejb.log
Perform the below workaround before starting the upgrade process to enable writing to the upgradeFramework.log file.
- Open the file <HP Business Availability Center install directory>\conf\core\Tools\log4j\EJB\upgradeFramework.properties.
- Go to the END of LAST FILE LINE (make sure you are on the end of line by pressing the End key).
- Press Enter. Make sure the marker went to next line.
- Save the file. The logger should now work properly and logging will appear in <HP Business Availability Center install directory>\log\EJBContainer\upgradeFramework.log.
- Administration: After service pack installation, files with the extensions .bff, .xmit, and .savf will not be available for download from the Downloads page in Platform Administration until the IIS is restarted on the Gateway Server.
- Administration: When upgrading from HP Business Availability Center 8.x earlier than 8.04 to HP Business Availability Center 8.04 or later, the Sonic directory must not be locked. To verify this, do the following:
If you can rename the directory, it is not locked. If it is locked, check which process is locking it and stop the process. (QCCR1I49304)
- Sonic Installation Troubleshooting: During installation of Business Availability Center 8.07 on top of versions earlier than 8.06, the SonicMQ component is upgraded from version 7.5.2 to 7.6.2. If, after installing 8.07 per the installation instructions described in this file, Business Availability Center services fail to start (specifically the Domain Manager or Message Broker services), try reconfiguring the Domain Manager by making sure that all Business Availability Center services are stopped and then reconnecting to the management database (using the database configuration utility). If after reconnection to the database, the Domain Manager or Message Broker services still fail to start, contact HP Customer Support. When contacting support, have the following information at hand:
- Verify that the backup directory <Business Availability Center root directory>\Sonic7.5.bak exists.
- Verify the installation of Sonic 7.6:
- Verify the configuration of Sonic 7.6:
- Verify the existence of the following boot logs <Business Availability Center root directory>\log directory: DomainManager_boot.log, MessageBroker_boot.log
- Upgrade from 7.x on Solaris: This note is relevant to customers upgrading from 7.x to 8.0x on a Solaris operating system. In such cases, when restoring manually copied files from backup, change the owner of the files from root to topaz.
- Viewing BAC: When logging into Business Availability Center for the first time after an upgrade, delete the browser's temporary Internet files. (This should be done on each browser that accesses Business Availability Center.)
- OMi: The following note is relevant if you are upgrading to Business Availability Center 8.07 from Business Availability Center versions earlier than 8.04 (if upgrading from 8.04 or later, no action needs to be taken) on a system with an installed and configured OMi. After installation of 8.07, you must run:
<HP Business Availability Center root directory>\bin\opr-configuration.bat -setup
In the Configuration Wizard, accept all preset values by clicking Next.
- Dashboard: The following note is relevant if you are upgrading from Business Availability Center 8.00 or 8.01 to 8.02 or above. After you finish the upgrade process and start Business Availability Center, perform the below steps on the Data Processing Server (DPS). If you have more than one DPS, perform the steps on the DPS that runs the KPI ENRICHMENT service.
- If you have more than one DPS, determine which DPS is running the KPI ENRICHMENT service:
- Access the JMX console on any DPS via http://<HP Business Availability Center DPS machine>:8080/jmx-console and enter the user name and password.
- Click service= hac-manager in the Topaz section.
- Locate the first java.lang.String listAllAssignments() operation (the one that requires no parameters) and click Invoke.
- Locate the line showing the KPI_ENRICHMENT service for the relevant customer.
- Check the Process column for the name of the server that runs the KPI ENRICHMENT service.
- Refresh the out-of-the-box repositories:
- Access the JMX console on the DPS via http://<HP Business Availability Center DPS machine>:8080/jmx-console and enter the user name and password.
- Click service=repositories_manager in the Topaz section.
- Locate the java.lang.String cleanRepositoryTypeDBAndCache() operation.
- For the parameter sCustId, enter 0 in the ParamValue box. Do not enter any other number, as doing so will clean any repository customizations that have been made.
- For the parameter repositoryType, enter assignmentRules in the ParamValue box.
- Click Invoke
- Restart the KPI Enrichment Service (KES).
- Return to the JMX Console entry page (http://<HP Business Availability Center DPS machine>:8080/jmx-console).
- Click service=KES Server in the Topaz section.
- Locate void stop()
- Click Invoke.
- Wait for the KPI ENRICHMENT service to stop (should take less than a minute). To determine whether the service is stopped, follow the procedure described above to determine which DPS is running the KPI ENRICHMENT service, and after invoking the java.lang.String listAllAssignments() operation and locating the KPI_ENRICHMENT service line, check the State column for the service status.
- On the KES Server JMX MBEAN View page, locate void start().
- Click Invoke.
- Dashboard: If you are upgrading from Business Availability Center 7.5x to 8.07, you may need to manually synchronize certain KPI assignments (for example, if you added KPI assignments in the previous version). If you do not perform manual synchronization on the relevant assignments, existing CIs may have wrong KPIs. You can ascertain whether you need to perform manual synchronization by checking the log files after upgrade. You can use one of the following files, located on the Business Availability Center Data Processing Server:
The following message is displayed for the relevant assignments in the upgrade.wizard.log file (a similar message is displayed in the kes.upgrade.log file):
"NOTE: Assignment name=[<assignment name>] id=[<assignment ID>] description=[<assignment description>] of Bundle name=[<assignment bundle name>] was added to new persistency. Please consider a manual synchronization for this assignment."
If you decide that manual synchronization is required, perform the synchronization in the Dashboard Administration KPI Assignments tab.
- Alerts: This note is relevant if you are upgrading from Business Availability Center 7.x to 8.0x using the staging procedure. After upgrading to 8.0x, the staged Business Availability Center server will not send email reports and will send alerts only to the staging recipient (defined in alert settings) until staging mode is disabled. This is to prevent duplicate alerts/reports from being sent from both production and staging environments.
- Alerts: Favorite Filters created for CI Status Alert and SLA Status Alert reports in Business Availability Center 7.5x are not upgraded and will not work in Business Availability Center 8.0x. You must recreate the Favorite Filters. (QCCR1I49333)
- BPI: If you customize any process diagram using annotations in Business Process Insight and then upgrade BPI, the customizations need to be restored manually. The installer keeps a backup of the annotations.xml file and images in the following directory:
<BPIInstallDir>\nonOV\jakarta-tomcat-5.0.19\webapps\hpbpi-bpd.bak.0
The annotations should be copied to the following directory to restore the customizations:
<BPIInstallDir>\nonOV\jakarta-tomcat-5.0.19\webapps\hpbpi-bpd
- Problem Isolation: If you use Problem Isolation On-Demand Monitors and have upgraded from HP Business Availability Center version 8.00 or 8.01 to version 8.07, you must reconnect to the management database. For details on setting database parameters, refer to the HP Business Availability Center Deployment Guide.
- UCMDB: The following note is relevant if you are upgrading from Business Availability Center 7.x: As a result of the upgrade process from Business Availability Center 7.x to HP Business Availability Center 8.00, the results displayed in some views may differ from the results displayed before the upgrade. You can use the Upgrade Information Tool to create a log file listing all such views and the problems in each view. The tool should be run on Business Availability Center 7.x before the upgrade. It details the changes to be anticipated during the upgrade process. The tool is not relevant for versions of Business Availability Center other than 7.x. The tool and instructions for its use are located in the <Business Availability Center 8.05 setup package root directory>\CMDB_Upgrade_Info_Tool directory.
- UCMDB: The following note is relevant if you are working with an external UCMDB architecture and are upgrading to Business Availability Center 8.07 and HP Universal CMDB 8.07 from 7.x or earlier versions: As described in the HP Universal CMDB Deployment Guide, during the HP Universal CMDB upgrade to 8.x, you are directed to perform a procedure that involves invoking a JMX that converts instance views into perspective-based views. If you were already working with an external UCMDB architecture prior to the upgrade, during the Business Availability Center upgrade, all instance views are automatically converted into perspective-based views. As such, it is not necessary to run this procedure again during the HP Universal CMDB upgrade. Doing so can result in corrupted views.
- UCMDB: If you are installing Business Availability Center 8.07 and are working with an external UCMDB architecture, you must perform the following procedure to copy and deploy the updated BAC packages to the UCMDB server, as they are not deployed automatically to the external UCMDB server during the installation. Perform the procedure after installing Business Availability Center 8.07 on the Business Availability Center machines and HP Universal CMDB 8.07 on the external UCMDB machine.
- Open the file < Business Availability Center 8.07 setup package root directory>\updated_bac_packages\updated_bac_packages.txt to view a list of the updated packages. The file contains the following lists: updated packages from 8.00 to 8.01 (relevant if you have not previously deployed updated 8.01 packages to the external UCMDB server), updated packages from 8.01 to 8.02, updated packages from 8.02 to 8.03, updated packages from 8.03 to 8.04, updated packages from 8.04 to 8.05, updated packages from 8.05 to 8.06, and updated packages from 8.06 to 8.07.
- Manually copy the relevant packages listed in updated_bac_packages.txt from the <Business Availability Center root directory>\ mam_lib\patch_packages directory on the Business Availability Center Data Processing Server to the root\lib\factory_packages directory on the UCMDB server. Only copy the .zip files and only copy the relevant packages listed in updated_bac_packages.txt, not all the contents of the patch packages directory.
- Start the UCMDB server.
- Deploy the updated packages you copied to the UCMDB server using the UCMDB JMX Console:
- Launch your Web browser and enter the following address: http://<server_name>:8080/jmx-console, where <server_name> is the name of the machine on which HP Universal CMDB is installed.
- Under MAM, click service=MAM Packaging Services to open the JMX MBEAN View page.
- Locate java.lang.String deployPackages.
- In the ParamValue box for the parameter customerId, enter the <customer id>. For non-multi-customer installations, the default customer ID is 1.
- Leave the ParamValue box for the parameter dir empty.
- In the ParamValue box for the parameter packagesNames, enter the names of the packages (separate the package names using commas).
- Select True to override DDM job configurations changed in the DDM user interface.
- Click Invoke to deploy the packages.
For more information on the UCMDB JMX Console, see the Package Manager section of the Model Management guide.
- UCMDB: HP UCMDB and Federated Adapters upgrade notes:
- CMDB RMI adapter: The CmdbRmiAdapter.zip package is overwritten during upgrade. If you modified the CmdbRmiAdapter package follow the below instructions:
- Copy the existing <HP Business Availability Center Data Processing Server root directory>\mam_lib\factory_packages\CmdbRmiAdapter.zip file to a temporary directory before starting the upgrade process.
- After the upgrade process, edit the files in the new CmdbRmiAdapter.zip package to match the changes you had made to the old adapter package. You must make the changes in the new CmdbRmiAdapter.zip and not overwrite with your old package.
- Replication. When upgrading from Business Availability Center 7.0x or 7.50 (but not from 7.51, 7.52, 7.53, 8.00, 8.01, 8.02, 8.03, 8.04, 8.05, or 8.06), it is recommended to perform a diff replication on your current system prior to upgrading to 8.07. This will prevent losing data for removal during the full synch. Then, after upgrading to 8.07, each replication job should be run in full replication mode. For details on running diff and full replication, see the Replication Jobs Tab section in the Model Management guide.
- Service Desk adapter: Perform the following procedures to make sure you do not lose your existing Service Desk adapter configuration information during the upgrade process:
- Proprietary adapter configured in Generic DB Adapter default folder: If you created a proprietary adapter in the default Generic DB Adapter folder, you must perform the following procedures to make sure you do not lose your existing Generic DB Adapter configuration information during the upgrade process.
- Proprietary adapters based on Generic DB Adapter (when upgrading from HP Business Availability Center 7.5.1 or 7.52 to 8.07): If you created proprietary adapters based on the Generic DB Adapter, you must perform the below procedure for each such adapter.
- Upgrade HP Business Availability Center from HP Business Availability Center 7.5.1 or 7.52 to 8.00, then to 8.01, and finally to 8.07.
- Copy all .jar files from <HP Business Availability Center Data Processing Server root directory>\fcmdb\Codebase\GenericDBAdapter to the proprietary adapter folder. Overwrite any existing files with the same name, if such files exist.
- Reload the adapter.
- Proprietary adapters based on Generic DB Adapter (when upgrading from HP Business Availability Center 7.0x or 7.50 to 8.07): If you created proprietary adapters based on the Generic DB Adapter, you must perform the below procedure for each such adapter.
- Upgrade HP Business Availability Center from HP Business Availability Center 7.0x or 7.50 to 8.00, then to 8.01, and finally to 8.07 (if upgrading from 7.51 or 7.52, see previous issue).
- Copy all .jar files from <HP Business Availability Center Data Processing Server root directory>\fcmdb\Codebase\GenericDBAdapter to the proprietary adapter folder. Overwrite any existing files with the same name, if such files exist.
- Copy plugins.txt from <HP Business Availability Center Data Processing Server root directory>\fcmdb\Codebase\GenericDBAdapter\META-INF to <HP Business Availability Center Data Processing Server root directory>\fcmdb\Codebase\<proprietary adapter folder>\META-INF.
- Open <HP Business Availability Center Data Processing Server root directory>\fcmdb\Codebase\<proprietary adapter folder>\META-INF\orm.xml using a text or XML editor.
- In orm.xml, for each CI type that is defined as a multinode in the reconciliation_rules.txt file, change the mapped attribute name from <CI type name>_<attribute name> to <attribute name>.
For example, if "host" is a multinode, change
<basic name="host_host_hostname">
<column updatable="false" insertable="false" name="[column_name]"/>
</basic>to:
<basic name="host_hostname">
<column updatable="false" insertable="false" name="[column_name]"/>
</basic>Note: If a reconciliation rule includes a link, attributes of end2 should still be prefixed by the corresponding CI type name. For example, if one of the reconciliation rules is multinode[host] expression[host.host_hostname or ip.ip_address] end1_type[host] end2_type[ip] link_type[contained], then ip.ipaddress should still be mapped as:
<basic name="ip_ip_address">
<column updatable="false" insertable="false" name="[column_name]"/>
</basic>
- In orm.xml, completely delete any entities whose name starts with IdClass2PK or IdClass3PK. For example, an entity defined as <entity class="generic_db_adapter.IdClass2PK_cpu" name="IdClass2PK_cpu"> should be deleted.
- In orm.xml, delete all lines that are similar to: <id-class class="generic_db_adapter.IdClass2PK_something"/> or <id-class class="generic_db_adapter.IdClass3PK_something"/>.
- Delete the <HP Business Availability Center Data Processing Server root directory>\fcmdb\Codebase\<proprietary adapter folder>\dynamic_classes folder, if it exists.
- Reload the adapter.
- UCMDB: This note pertains to reconciliation rules in UCMDB. During the upgrade procedure from a previous version, and once the server is up, you should activate the RemoveMultipleProcesses enrichment rule. Once Calculate TQL Result Count returns zero results, you can deactivate the rule. For details on activating enrichment rules, see the Rule General Attributes Page section in the Enrichment Manager chapter of the Model Management guide.
- SiteScope: This note is relevant if you are upgrading SiteScope from a pre-10.00 version and are working with the Business Availability Center/ServiceCenter or Service Manager integration. When setting up the ServiceCenter monitor in SiteScope, a file called peregrine.jar is created and placed in the WEB-INF\lib directory on the SiteScope machine. This file must be backed up before upgrading SiteScope or it will be deleted during the upgrade. After the upgrade is complete, copy the backed up peregrine.jar file back to the WEB-INF\lib directory.
- SiteScope: This note is relevant if you are upgrading from Business Availability Center 7.x to 8.07. If you are using Business Availability Center with SiteScope 9.54, you must restart SiteScope to enable any monitors created in SiteScope after the upgrade to report topology.
- Diagnostics: This note is relevant if you are upgrading to Business Availability Center 8.07 and have an HP Diagnostics instance registered in Business Availability Center. To ensure proper source adapter synchronization, after upgrading to 8.07 re-register the Business Availability Center-Diagnostics integration in the Business Availability Center > Admin > Diagnostics page.
Uninstallation
- For instructions on completely uninstalling Business Availability Center, see the HP Business Availability Center Deployment Guide.
- To roll back an installation of Business Availability Center 8.07 to Business Availability Center 8.00, 8.01, 8.02, 8.03, 8.04, 8.05, or 8.06, follow the instructions below.
- Stop the HP Business Availability Center service on all servers and confirm they are stopped.
- Stop the Web server process (IIS Admin Service for IIS; Apache service for Apache) on all servers.
- From Control Panel, select Add/Remove Programs > Change or Remove Programs > HP Business Availability Center 8.07. On Windows 2003 Server, select the Show Updates option to ensure that version 8.06 is displayed beneath the root HP Business Availability Center entry.
- When the uninstall process is complete, restart the machine if requested.
- Remove all subdirectories under the directory: <HP Business Availability Center root directory>\EJBContainer\server\mercury\work
- Restart the Web server process (IIS Admin Service for IIS; Apache service for Apache) on all servers.
- Restart the HP Business Availability Center service on all servers.
- Perform the below procedure to remove an attribute added by the 8.06 SP to LW-SSO settings, whose presence will cause an error when running Business Availability Center:
- Enter the URL of the JMX console (http://<Gateway or Data Processing server name>:8080/jmx-console/) in a Web browser.
- Enter your JMX console authentication credentials. If you do not know your authentication credentials, contact your system administrator.
- Under Foundations, click service=Infrastructure Settings Manager.
- Locate the getGlobalSettingValue() method.
- In the contextName box enter: SingleSignOn
- In the settingName box enter: lw.sso.configuration.xml
- Click Invoke.
- Copy the resulting xml output from the Setting's value column to a text editor.
- In the text editor, locate and remove the mode attribute from the xml code (for example, delete the string mode="production" or mode="demo").
- Copy the updated xml code to the clipboard.
- In the JMX console return to the Infrastructure Settings Manager page and locate the setGlobalSettingValue() method.
- In the contextName box enter: SingleSignOn
- In the settingName box enter: lw.sso.configuration.xml
- In the newValue box, paste the updated xml code.
- Click Invoke.
- Return to the main JMX console page and, under Topaz, click service=SSO.
- Locate the start() method, and click Invoke.
- If you installed 8.07 using the installation path 8.00 à 8.01 à 8.07 and then rolled back the 8.07 installation, you may encounter an issue with the servers (Business Availability Center services) restarting. In this case, stop Business Availability Center on all servers and reconnect to the management database (using the database configuration utility).
- If you need to roll back to an earlier Business Availability Center version (for example, from 8.07 to 8.06), you may also need to roll back any database that changed as a result of the upgrade. If you encounter any problems, contact HP Software Support.
- If you are uninstalling Business Availability Center 8.07 from machines running Solaris OS, you must uninstall under the "root" user. Then, after uninstallation, change ownership of the /opt/HPBAC/conf/bus directory to your Business Availability Center user using the following command: chown -R <BAC_user> /opt/HPBAC/conf/bus
- When uninstalling or rolling back Business Availability Center, make sure to uninstall or rollback the data collectors as well.
8.07 Version Content and Limitations
- See also 8.06 limitations, 8.05 limitations, 8.04 limitations, 8.03 limitations, 8.02 limitations, 8.01 limitations, and 8.00 limitations below. Removed limitations are marked with prefix FIXED IN 8.0x.
- Export/Import of User Information: Users, roles, and permissions can be transferred from a source system to a target system, using the following four steps:
- Configure the contexts that are to be exported. Decide if you want to limit which contexts are exported. You can view a list of available contexts by opening up the JMX console for the Authorization Service MBean, which can be found at: http://<source-host-name>:8080/jmx-console/HtmlAdaptor?action=inspectMBean&name=Topaz%3Aservice%3DAuthorization+Service
To display the list of deployed contexts, invoke listAuthorizationContexts(). To update the contexts included in the export, go to HPBAC\conf\tas\exportedContexts.properies (on BAC Gateway server) and alter the contexts-to-export property. Be sure to save your changes.
Contexts in contexts-to-export property must be separated by commas only, without spaces. Otherwise, roles and permissions will be exported incorrectly, and you will have to check and invoke roles/permissions of imported users/groups manually after the import.
- Export the users, roles, and permissions from the source BAC system to a ZIP file containing XML files with the exported information. Open up the JMX console to the export functionality. The JMX MBean that houses the export functionality can be found at: http://<source-host-name>:8080/jmx-console/HtmlAdaptor?action=inspectMBean&name=Topaz%3Aservice%3DAuthorization+Service+Data+Import+Export
Invoke the loadExportedContexts() method to configure the exporter with those contexts. On the JMX bean page you will find an input parameter for the exportAllTasEntities() method. Enter the fully qualified filename for the export file on the host machine. The file must end in ".zip", for example: C:\HPBAC\export.zip
Invoke the method, and wait for it to complete.
- Transfer the exported ZIP file from the source machine to the target machine: Navigate to the location on the file system, and copy the file to the machine that the target system is running on.
- Import the users, roles and permissions from the ZIP file to the target BAC system: Open the JMX console as above, but with the target system as the hostname: http://<target-host-name>:8080/jmx-console/HtmlAdaptor?action=inspectMBean&name=Topaz%3Aservice%3DAuthorization+Service+Data+Import+Export
In the importAllTasEntities() method, enter the fully qualified filename of the ZIP file that was produced by the export, and invoke the method. When the command completes, it will have imported the permissions for all the contexts in the source machine's exportedContexts.properties file.
Limitations: Before importing users/groups, ensure that the destination environment is clean from any created resources, such as reports, profiles, monitors and others. Otherwise, you have to check and invoke roles/permissions of imported users/groups manually after import. (QCCR1I46743)
- Empty Folder in UCMDB UI: In some cases, the empty Pre-Upgrade folder remains in the UCMDB UI after running the pre-upgrade tool and undeploying the UCMDB views. This can be manually deleted in the UCMDB UI. (QCCR1I59955)
- Dashboard and External UCMDB: In Dashboard 8.x, if you are working with an external UCMDB, or if you have changed the default CMDB connection settings, to view information in the Console Changes and Incidents and Business Impact tabs you must define the related infrastructure settings. Select Admin > Platform > Setup and Maintenance > Infrastructure Settings > Applications, and select Dashboard Application. Within the Dashboard Layout Properties area, edit the following settings:
- SLA Recalculation: The BLE Offline State table contains data for SLA calculations, and it is divided into partitions with start dates and end dates. If you recalculate an SLA, and a partition for the recalculation start date was already purged, an error message appears. The start date of SLA recalculation must be later than the start date of the first available (not purged) partition. (QCCR1I48846)
- Generated Report in PDF Format Formatting Problems: When a report generated in PDF format exceeds one page in size, according to a default page size, it is split into multiple pages. If by chance there is a line of text exactly where the page is split, it may be cut horizontally, so that the top half of the line is on one page, and the bottom half on the next page. (QCCR1I57827)
- Enhanced Alerts Functionality: The following functionality has been added to the Alerts application:
- Overriding Alerts Default Mail Settings: Infrastructure Settings include three new Alerting options: Alerts E-Mail Settings; Alerts Pager Settings; and Alerts SMS Settings. Each of the new options overrides the corresponding settings in Alerts Default Mail Settings.
For example, to use a secure SMS server while keeping emails and pager messages on the default server:
baseline data is used instead of raw data in order to enhance the report's performance. (QCCR1I50009)
- Container Naming Issue in RUM: If there is more than one container on the same level in Monitoring Administration for RUM that have names longer than 34 characters, and for which the first 34 characters are identical, after adapter synchronization there will be only one container in the UCMDB. This container will have the name of the first 34 characters of the source containers and it will contain all the source containers' children.
Workaround: Ensure that container names are no longer than 34 characters. (QCCR1I56988)
Workaround:
- Register (or remove and reregister) SiteScope to the BAC server using the Default Virtual Server for Application Users URL instead of the Default Virtual Server for Data Collectors URL (enter the URL in Admin > System Availability Management > Distributed Settings > Gateway server name/IP address). This URL can be found in Admin > Platform > Setup and Maintenance > Infrastructure Settings > Foundations > Platform Administration > Host Configuration table.
- If you cannot remove SiteScope from BAC because you do not want to lose data, you can change the host name on SiteScope under BAC Integration Preferences (Preferences > Integration Preferences > Business Availability Center machine name/IP address). (QCCR1I58582)
Business Availability Center Bug Fixes
UCMDB Version Content and Limitations
UCMDB
- Incorrect results may appear for some CIs in a Topology report based on a report TQL which has a cardinality of 0.
- A local attribute size of any class from the UCMDB class model can be increased, but cannot be decreased thereafter.
- In the IT Universe Manager, when you search for a CIT and select a CI that has more than 900 related CIs, an error message appears. After closing the error message, select another CI in the results with fewer than 900 CIs. If you then select the original CI again, its Get Related CIs shortcut menu options are disabled. To avoid this, modify your search to search for that specific CI only. The menu options are enabled again.
- If you want to work with the UCMDB Gold Master report, copy the file from the DVD: <Business Availability Center 8.07 setup package root directory>\CSRF_Patch_over_807\AppServer
to: Business Availability Center Gateway Server\HPBAC\Appserver folderDiscovery
Bug Fixes
8.06 Version Content and Limitations
- Authentication: HP Business Availability Center's default authentication strategy is Lightweight Single Sign-On (LW-SSO). When LW-SSO is enabled, you must use a fully qualified domain name (FQDN) to access HP Business Availability Center. If you do not require LW-SSO, you can disable it. If LW-SSO is disabled, you do not need to enter a FQDN to access HP Business Availability Center. For details on disabling LW-SSO, see "Authentication Strategies" in the Platform Administration guide.
- Working with Oracle Database and SCAN: In release 11g, Oracle introduced the Single Client Access Name (SCAN), as a preferred access method for clients connecting to the RAC. In this method clients are not required to configure individual nodes in the RAC, rather they use a single virtual IP known as the SCAN or the SCAN VIP. The SCAN is a single network name defined for the cluster either in your organization's Domain Name Server (DNS) or in the Grid Naming Service (GNS) that rotates between several IP addresses reflecting multiple listeners in the cluster. The SCAN eliminates the need to change clients when nodes are added to or removed from the cluster. The SCAN and its associated IP addresses provide a stable name for clients to use for connections, independent of the nodes that make up the cluster. SCAN addresses, virtual IP addresses, and public IP addresses must all be on the same subnet.
In HP Business Availability Center 8.06 when using Oracle 11g RAC, it is recommended to use the SCAN method. For details on configuring Oracle RAC, see "Real Application Cluster" in the HP Business Availability Center Database Guide. The following changes apply to that section of the documentation:
- When configuring the bac-tnsnames.ora file, the parameter ADDRESS_LIST contains an address entry for each node in the RAC environment. In the case of Oracle 11g using SCAN, it contains only the SCAN virtual IP.
- When creating a new database schema or connecting to an existing schema in the Server and Database Configuration utility, fill all the parameters as you usually would except for the following:
- Working with Oracle Database and partitioning: When working with Oracle Server 11.2.0.2, with either a large or small HP Business Availability Center deployment, set the _PARTITION_LARGE_EXTENTS parameter to FALSE. This hidden parameter, set to TRUE in 11.2.0.2, is affecting the size of partitions in native partitioned tables. Unless set to false, the initial extent allocated for each partition is very large thus causing unwanted growth of database data files.
- Working with external UCMDB architecture: Business Availability Center and HP Universal CMDB must have identical service pack versions installed for the products to work together.
- Configuring LW-SSO to work with client-side authentication certificates: Lightweight Single Sign-On can be configured via the JMX console to accept client-side authentication certificates. Once a certificate is accepted, the user is logged into HP Business Availability Center. Client-side authentication certificates provide a secure alternative to entering user credentials in the login screen. To configure LW-SSO to accept client-side authentication certificates:
- Enter the URL of the JMX console (http://<Gateway or Data Processing server name>:8080/jmx-console/) in a Web browser.
- Enter your JMX console authentication credentials. If you do not know your authentication credentials, contact your system administrator.
- Under Foundations, locate Service: Infrastructure Settings Manager.
- Invoke the operation java.lang.String getGlobalSettingValue() with contextName set as SingleSignOn and settingName as lw.sso.configuration.xml.
- Copy the resulting Setting's value string to a text editor and save the string as a backup of the original value.
- Within the text editor, add the following element string just after the </in-ui-lwsso> element, if it is not already there: <in-clientCertificate enabled="true" userIdentifierRetrieveField="SubjectDN" userIdentifierRetrieveMode="FieldPart" userIdentifierRetrieveFieldPart="EMAILADDRESS"/>
- From the JMX MBean View page, invoke the operation void setGlobalSettingValue() with contextName set as SingleSignOn; settingName as lw.sso.configuration.xml; and newValue as the new Setting's value string, copied from the text editor. Be sure to copy the entire string without line breaks or word wrapping.
- In Business Availability Center, select Admin > Platform > Setup and Maintenance > Infrastructure Settings. Click Foundations and select Single Sign-On from the drop-down list.
- Set Unknown User Handling Mode to Deny.
- RUM Licensing: The RUM Advanced with SLM license key cannot be generated and installed. Instead, the following license keys are installed (among others):
Workaround: After installing RUM licenses, replace RUM Advance with SLM license type assignments with RUM Probe and RUM SPI for Applications license types. (To assign licenses, select End User Management Administration > Settings tab > Real User Monitor Settings tab > RUM Probe Licensing table.)
- Service Level Management: You can modify the Start date, Time zone, and Include Locations options (included on the Define Agreement Properties page of the Agreements Wizard) only during agreement creation, or if the agreement is still in Preliminary or Pending state (in which case, you can modify the agreement in edit mode). Once the agreement is running or terminated, you cannot change these options.
- Service Level Management: You can apply or remove the Include Locations option for an agreement only during agreement creation, or if the agreement is still in Preliminary or Pending state.
- If you require Include Locations functionality for an agreement, select the option during agreement creation. Once the agreement is running, you cannot apply the functionality; the only option is to clone the agreement to create a new one and select the Include Locations option in the new agreement.
- If a running agreement has Include Locations functionality, you cannot remove the functionality; the only option is to clone the agreement to create a new one and deselect the Include Locations option in the new agreement. Therefore, the procedures "To remove Include Locations functionality from a service-based agreement" and "To remove Include Locations functionality from a CI-based agreement" in the "Agreements in Service Level Management" chapter of the Using Service Level Management guide are not supported. (QCCR1I49626)
- CI-based SLA issues: When you create a CI-based SLA, and a CI has more than one parent CI, the following problems may occur:
- When you expand a CI instance that appears twice, the CI may be expanded in the context of another parent.
- When you detach a CI instance from its parent CI, the CI instance is detached from all its related parents.
- When you add a CI to an SLA in the context of one parent CI, the CI is detached from all other parents associated with the CI in the SLA.
Workaround: To prevent these problems, create a service-based rather than a CI-based SLA. (QCCR1I53456)
- Editing KPI thresholds in an SLA: If you edit KPI thresholds in an SLA but you do not recalculate the SLA for the entire period on which SLM forecasting is based, the expected breach date may appear blank. To generate an expected breach date, perform recalculation for the entire period on which forecasting is based. (QCCR1I53078)
- FIXED IN 8.07: SLM reports: Drilldown from CI Summary, CI Impact, and Time Range Comparison reports does not work (displays internal error message).
- PARTIALLY FIXED IN 8.07, See "Container Naming Issue in RUM" in 8.07 Limitations section:
EUM Source Adapter failure: The EUM Source Adapter fails to insert data in the CMDB due to long container names. Workaround: Limit container names to a maximum of 34 characters. (QCCR1I56988)
- Custom KPI issue: When a custom KPI exists but it is not applied to any CIs, the Dashboard Filters tab may not work properly. If you delete the unused KPI within the KPI repository, the Filter tab capability is restored. (QCCR1I52624)
- Working with Report Manager: When using the Report Manager (Application > User Reports) a JavaScript error might be encountered when using Internet Explorer 7.0.5 or earlier. Workaround: Use the latest version of IE.
- Table presentation in report PDFs: If the table of a report does not expand to the whole width of the page when the report is in PDF format, proceed as follows:
- Triage report missing icons in SSL: When Business Availability Center is running in an SSL environment, the Triage report might be missing status icons when displayed in PDF format. (QCCR1I54860)
- Prevent sending Business Availability Center login credentials in published report URLs: In Reports, when publishing a report, Generate URL constructs a URL with the user's encrypted credentials, which could present a security breach. To prevent this, disable the Publish Report option by doing the following:
- Modify the default character set used to send SNMP alert traps: By default the character set used to send SNMP alert traps is the platform's default character set. If your operating system supports multi-byte characters, it is recommended to use the UTF-8 character set. To modify the default character set used to send SNMP alert traps, select Admin > Platform > Setup and Maintenance > Infrastructure Settings, select Foundations, select Alerting, scroll to the Alerting - Triggered alerts area, and modify the SNMP alerts charset parameter. The change takes effect after 1 minute. (QCCR1I50902)
- Substitute double-quotes in the command in the Executable Files action of an event-based alert: When specifying a command in the Executable Files action of an event-based alert, you can use special tokens that are replaced with actual values when the command is prepared for execution. Those values might include a double quote (") that may cause the resulting command line to be inappropriately interpreted by the operating system. To avoid this misinterpretation, you can modify the default value of the Substitution for quote in command line infrastructure setting (Admin > Platform > Setup and Maintenance > Infrastructure Settings; select Foundations; select Alerting; locate the Substitution for quote in command line parameters entry in the Alerting - Triggered alerts table). Modify the value as listed below. (QCCR1I53267)
In addition, double-quotes used in the Alert Name parameter in the Executable File action definition for CI Status alerts and SLA alerts are substituted according to the value of the Substitution for quote in command line parameters infrastructure setting.
- Open reports in one window or in separate windows: To determine whether reports open in one window or in separate windows, in Admin > User Reports:
- Select a report in the list of available reports and click the View Report button (top left-hand corner) to display the report in a new window. Every subsequent report that you open in the same way, overrides the previous report in the same window.
- Select a report in the list of available reports, press Ctrl and click the View Report button to display the report in a new window. Every subsequent report that you open in the same way, is displayed in a new window. (QCCR1I39732)
- Custom Report URL format: When you create a Custom Report that includes a URL, you must include the protocol (HTTP/HTTPS) in the URL string. (QCCR1I7890)
- Manipulating Time Range in Custom Query Builder URL: When generating a URL using the Custom Query Builder (Applications > User Reports > Custom Query Builder > Generate > Publish Report > Generate URL), it is now possible to manipulate the time range within the URL without having to regenerate the query and the URL. You can add the following parameters to the end of the URL generated to adjust the time range of the report:
These parameters should be added into the URL at the end of the URL string using the '&' separator for each parameter. For example, add the following to the end of the generated URL: &from_time_stamp=1259080209&to_time_stamp=1259080809
- SiteScope: The Business Availability Center Documentation Library released with Business Availability Center 8.06 includes the SiteScope 10.00 documentation. If you install SiteScope 10.13 for use with Business Availability Center 8.06, you can access the SiteScope 10.13 documentation while in SiteScope as follows:
- Business Process Monitor Installation: Before installing Business Process Monitor on a machine, install the Microsoft Visual C++ 2005 Service Pack 1 Redistributable Package ATL Security Update (available from http://www.microsoft.com/downloads/details.aspx?FamilyID=766A6AF7-EC73-40FF-B072-9112BAB119C2&displaylang=en).
- TransactionVision: SSL support for SonicMQ for transports using the OEM SonicMQ Message Broker is not supported in TransactionVision 8.05.
Business Availability Center Bug Fixes
UCMDB Version Content and Limitations
UCMDB
Discovery
Bug Fixes
8.05 Version Content and Limitations
Workaround: Create a duplicate path on the C:\ drive by copying E:\<HP Business Availability Center root directory>\conf\settings to C:\HP Business Availability Center root directory>\conf\settings. (QCCR1I47337)
Workaround: Enable the Compatibility mode in IE 8.0 by selecting Tools > Compatibility View Settings and adding the BAC 8.05 site to the list of Websites shown in the Compatibility view. (QCCR1I49012)
- Real User Monitor: If you want to install the Real User Monitor Engine and/or Repository (MySQL Database) on Windows 2008 using Windows Remote Desktop, you must carry out the following steps on the remote machine on which you are installing them:
- Dashboard Administration: In Microsoft Internet Explorer (IE) 8, Dashboard Administration does not support use of web URLs to create a Custom Map image. (QCCR1I49865)
- Dashboard: Within Dashboard, when you select an OMU event in Top View and choose the OVO Drill Down menu command, the window that opens displays a URL containing the username and password of the OM-BAC integration.
Workaround: To prevent this, enter the following URL in the HP OM integration form (Admin > Integrations > HP OM > HP OM URL):
http://<OMU server>:3443/OvCgi/ito_op_applet_cgi.ovpl?webstart=true
The OVO Drill Down menu command then opens the Java Web Start (JNLP) UI, and the integration username and password will not be visible. (QCCR1I41204)
- Dashboard Reports: The Format-Excel option does not work with Microsoft Excel 2007 for all Dashboard reports (QCCR1I49217)
- My BSM: The documentation for My BSM contains a section with the heading Change the Default Page for a Specific User - Optional. This section is not correct; you cannot change the default page for a specific user.
- Alerts: In CI Status alerts and SLA alerts, you can now use special characters in the Name and Description fields in the General page of the CI Status Alert wizard and the SLA wizard. In addition, the total number of characters allowed in these fields is 200. (QCCR1I48382)
- (See 8.06 Version Content and Limitations for an updated version of this entry. )Service Level Management: If you are using the Include Locations option for an SLA, you cannot then remove the Include Locations functionality. Therefore, the procedure "To remove Include Locations functionality from a service-based agreement" in the "Agreements in Service Level Management" chapter of the Using Service Level Management guide is not supported. If you do not use the Include Locations option while creating an SLA, you will not be able to apply this functionality later (in edit mode); therefore, be sure to select the option, if required, during SLA creation. (QCCR1I49626)
- (See 8.06 Version Content and Limitations for an updated version of this entry. ) Service Level Management: Note that you can only set the SLA start date, time zone, and Include Locations option during SLA creation; you cannot change these values when you open the SLA in edit mode.
- Service Level Management: Note the following regarding the outage functionality:
An outage starts when the criteria specified in the outage rule are met, as follows:
- Outage by samples. An outage starts if the samples for the CI have reached a minimum number of consecutive failures, and the combined duration covered by the failed samples has reached the defined minimum duration.
- Outage based on availability or system availability. An outage starts if the CI's Availability KPI result over the SLA calculation cycle (default = 5 minutes) is less than the defined threshold for the Outages KPI, and the availability has been less than the threshold for the defined minimum duration. (This means that outages have a granularity of 5 minutes, by default.)
An outage ends (and is then included in the reports), when the following occurs:
- Outage by samples. The outage is ended if a successful sample arrives for the CI.
- Outage based on availability or system availability. The outage is ended if the CI's Availability KPI result over the SLA calculation cycle (default = 5 minutes) reaches the threshold defined in the Outages KPI.
- All outages. Any outage is ended if:
Note: The 24x7 calendar continues infinitely, so outages cannot be ended by this calendar. When viewing reports filtered by this calendar, any ongoing outages do not appear; however, for the same period outage may appear when you filter by other calendars, because an ongoing outage is ended when a calendar period ends. (QCCR1I34272)
- Service Level Management: When viewing outage details in the Outage reports, the full duration of the outage is displayed (and not only the outage duration during the selected time period). For example, you display the Outage Distribution report for the month of July. An outage that started on 27th May and terminated on June 6th is shown in the report as having a duration of 10 days, even though only 6 of those days occurred during the selected time period. (QCCR1I34273)
- BAC for Siebel: The option of disabling NTP time sync is not working when running database breakdown (QCCR1I49330)
- SiteScope: The Business Availability Center Documentation Library released with Business Availability Center 8.05 includes the SiteScope 10.00 documentation. If you install SiteScope 10.12 for use with Business Availability Center 8.05, you can access the SiteScope 10.12 documentation while in SiteScope as follows:
- SiteScope: When accessing SiteScope from System Availability Management Administration using Internet Explorer 7, a 408 Request Timeout error is displayed.
Workaround: Configure the browser to accept cookies from the SiteScope server:
TransactionVision
Workaround: Create each job as a new job in the 8.05 UI, entering these values:
Job Name: CMDB Population
Startup Mode: automatic
Class Name: com.bristol.tvision.job.CMDBPopulationJob
Repeat Every: 15 minutes
Job Name: CMDB Update
Startup Mode: automatic
Class Name: com.bristol.tvision.job.CMDBUpdateJob
Repeat Every: 24 hours
- New CICS Task criterion for queries: For queries, a new criterion is available for the CICS technology: CICS Task. This criterion works in the same way as the z/OS CICS Task criterion in the "WebSphere MQ" criteria group. See chapter 11, "Queries" in Using TransactionVision for more information. (QCCR1I38603)
- Configuring the TransactionVision database on Oracle 10g RAC: In order to access an Oracle 10g RAC database, the TransactionVision database must be manually configured as follows:
jdbc_driver=com.mercury.jdbc.oracle.OracleDriver
jdbc_url=jdbc:mercury:oracle:TNSNamesFile=C:\\db\\tnsnames.ora;TNSServerName=TVDB
TNSNamesFile should point to the tnsnames.ora file that has been copied from the Oracle server.
TNSServerName should contain the RAC instance name.
Example of a tnsnames.ora file for an RAC configuration with two Oracle instances:
TVDB =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = labm1db15-vip)(PORT = 1521))
(ADDRESS = (PROTOCOL = TCP)(HOST = labm1db16-vip)(PORT = 1521))
(LOAD_BALANCE = yes)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = TVDB)
(failover_mode=(type=select)(method=basic))
)
)
Note: For connection to an Oracle 11g RAC database, the SCAN (Single Client Access Name) can be used as the database name during TransactionVision configuration, and the manual URL configuration with tnsnames.ora file is not necessary. (QCCR1I39176)
TransactionVision Error(FrameworkInitError): Failed to initialize Analyzer framework.
TransactionVision Error(CouldNotLoadBeanClass): Bean class 'XXXXX' specified in beans config file could not be loaded. Please see error log for details.
Solution: On Windows, copy the "set ANALYZER_ADDITIONAL_CLASSPATH=...." line from the saved SetupEnv.bat.bak file, and paste it in the migrated SetupEnv.bat file. Then add the following line to the SetupEnv.bat file:
set CLASSPATH=%CLASSPATH%;%ANALYZER_ADDITIONAL_CLASSPATH%
On UNIX, perform these steps on the SetupEnv.sh file. (QCCR1I40796)<tv>
<transport type="sonicmq" connectionstring="broker=localhost; port=21111; user=Administrator; password=; configurationqueue=TVISION2.CONFIGURATION.QUEUE" />
<capturehttpheaders enabled="true"/>
<capturecookies enabled="true"/>
</tv>
(QCCR1I41160)
You can specify the maximum size for the user data to be displayed in tree format. Any XML chunk that exceeds this size in the event will be shown as raw text. The new setting is max_userdata_xml_bytes_to_display and is located in Analyzer.properties. The default value is 50000. (QCCR1I44990)
You can now set an Analyzer property that causes the user data to be stored in the USER_DATA table. This is consistent with how user data of other technology events, such as JMS Sensors, store user data. User data can potentially be very large compared to the XML event itself and in these cases should be stored in the USER_DATA blob tables which improves overall efficiency. Also having user data stored separately allows it to be managed separately. For example, you may want to delete only the user data for old events rather than completely deleting old events. The new setting is store_userevent_user_data_separately and is located in Analyzer.properties. The default value is false.
Note: Storing the user data in the USER_DATA table is also required if you want to discard the user data entirely (by setting the Store User Data option of the communication link or the save_user_data option in Analyzer.properties to false). (QCCR1I48727)
Workaround: Add the following line to the <sensor_install_directory>/exit_sensor.deny file:
*/mqm/eclipse*
(QCCR1I49789)
Business Process Insight
Workaround: After upgrading to 8.05, and before you start the BPI components, update the Process Repository Database hostname in the BPI Server Administration Console > Process Repository > <MSSQL_or_Oracle_database> > < MSSQL_or_Oracle> Server hostname field. (QCCR1I37870)
Workaround: Edit the <BAC_installation_dir>\conf\bpi_process_repository\templates\bia-model-repository-db.mssql.properties file as follows.
Change:
jdbc.connection.url jdbc:inetdae7: {{HPBPI_PROCESS_REPOSITORY_MSSQL_HOST}}: {{HPBPI_PROCESS_REPOSITORY_MSSQL_PORT}}?database={HPBPI_PROCESS_REPOSITORY_MSSQL_DATABASE}}&tds9=false
To use a specific database hostname and instance separated by "\\" :
jdbc.connection.url jdbc:inetdae7:<DATABASE_HOST_NAME>\\<INSTANCE>:{{HPBPI_PROCESS_REPOSITORY_MSSQL_PORT}}?database={HPBPI_PROCESS_REPOSITORY_MSSQL_DATABASE}}&tds9=false.
Workaround: Change the Data Definition Name using the Modeler if necessary. (QCCR1I48053)
Business Availability Center Bug Fixes
See table below for details.
UCMDB Version Content and Limitations
Universal CMDB
- Applet Tag Usage: Following Sun's recommendations (http://java.sun.com/j2se/1.5.0/docs/guide/plugin/developer_guide/using_tags.html#object), UCMDB code uses the HTML applet tag and not the object tag to launch the UCMDB applet, in version 8.0x. Therefore, users must not disable the Java plug-in applet tag support (Internet Explorer > Tools > Internet Options > Advanced tab > Use JRE xx for <applet> (requires restart)).
- In the CI Selector, there is no option to search for CIs in all views. The CI search is run on the current view. (24778)
Discovery and Dependency Mapping
To configure the DDM Probe to work against a reverse proxy:
- Edit the discoveryProbe.properties file (located in <DDM Probe root directory>\hp\DDM\DiscoveryProbe\root\lib\collectors).
- Set the serverName property to the reverse proxy server's IP or DNS name.
- Set the serverPort and serverPortHttps properties to the reverse proxy server's ports.
- Save the file. (26659)
I18N
Integration
This field is used for different purposes by different adapters.
To connect to a database, you must specify the database type, and schema or SID. For example, to connect to an MS-SQL database, enter dbtype=sqlserver;dbname=yourdbname. (29626)
Defect Fixes
8.04 Version Content and Limitations
- Platform Administration: You can now use the Customization tab in User Management to hide applications and whole contexts from a user's view, in addition to being able to hide pages and tabs within applications. To hide an application or context for a user, select Admin > Platform > User Management, highlight a user in the list of users and select the Customization tab. In the list of Contexts, clear the checkbox for the applications you want to hide from the selected group or user. In the list of Pages and Tabs, you can hide specific areas per visible application.
- Platform Administration (Windows Server 2008): On Windows Server 2008 SP2 or R2, when viewing files using Windows Explorer, the size of Business Availability Center log files might be displayed as 0 bytes even though there may be information in the logs, or as a smaller number of bytes than represents the quantity of data in the logs. Refreshing the Explorer view may solve the issue.
- Platform Administration (Windows Server 2008): On Windows Server 2008 R2, if the Business Availability Center root directory is defined as a shared folder, when attempting to open Business Availability Center in a browser to log in, an HTTP 401.3 server error is displayed and it is not possible to log into the application. Workaround: Grant full control for IUSR account and IIS_IUSRS group on Business Availability Center root directory (by default, called HPBAC). Then restart IIS using iisreset /RESTART.
- My BSM: When you create a perspective-based view, the CIs in the view contain KPIs that are filtered according to the perspective. For example, an EUM view contains response time KPIs. If a CI in this view has the System KPI assigned, it is not visible since it is filtered out based on the perspective. However, My BSM inverts this KPI filtering and displays KPIs that should not be displayed, and does not display the KPIs that should be displayed. Workaround: Within Admin > Dashboard > KPIs, select the view, and click on the view node. Check a checkbox and then clear it (to enable the Save button), and then click Save. This updates the view settings, and KPIs are displayed properly in My BSM.
- My BSM: A new Infrastructure Setting named Portlet personalization security allows you to enable or disable the personalization option in My BSM. When the setting value is false (default value), all users can personalize their portal. When the setting value is true, personalization is disabled for all users who do not have full control permissions on modules. Full control permissions for a specific user can be enabled using the Permissions Manager > My BSM context, by giving the user Full Control operation on the Modules resource.
- Dashboard and Service Level Management: The Rules API that was introduced in Business Availability Center version 8.00 enables you to create new business rules for Dashboard and Service Level Management, using the Groovy scripting language. The Business Availability Center documentation contains basic examples of API rules. Additional examples of rules based on customer requests can be found in the Self-Solve knowledge base, in document KM778069. This document is intended to be collaborative, and more API rules will be added to it based on customer input. If you have any comments on existing rules or suggestions for new rules, please provide feedback on the document within the knowledge base. Your feedback will then be included within the document.
- Service Level Management: If you are upgrading to Business Availability Center 8.04 from an existing 8.x installation, the EMS Events Availability rule, which was added in Business Availability Center 8.03 (for details click here) is only available after performing the following procedure:
Note: If you are upgrading to Business Availability Center 8.04 from an 8.03 installation, and you already performed the above procedure when installing 8.03, you do not need to perform this procedure again.
- Alerts: The Transaction Error parameter is not supported in 8.04. The Transaction Error parameter, which can be used as a URL parameter in the Access URL dialog box, represents a description of the error that the script generated if an error occurred at the time of the event.
- Custom Query Builder: When generating a URL using the Custom Query Builder (Applications > User Reports > Custom Query Builder > Generate > Publish Report > Generate URL), it is now possible to manipulate the time range within the URL without having to regenerate the query and the URL. You can add the following parameters to the end of the URL generated to adjust the time range of the report:
These parameters should be added into the URL at the end of the URL string using the '&' separator for each parameter. For example, add the following to the end of the generated URL: &from_time_stamp=1259080209&to_time_stamp=1259080809
- SiteScope: The Business Availability Center Documentation Library released with Business Availability Center 8.04 includes the SiteScope 10.00 documentation. If you install SiteScope 10.11 for use with Business Availability Center 8.04, you can access the SiteScope 10.11 documentation while in SiteScope as follows:
TransactionVision
- TransactionVision: Drill-downs from the Transaction Tracking Details report to Diagnostics may not work correctly. Solution: Use Java 1.6_16 with the Diagnostics UI.
- TransactionVision: On WebLogic application server platforms, in rare scenarios, it is possible that some transactions have abnormal transaction duration time due to the JMS API acknowledge calls that take many minutes to complete. Solution: Change the Data Collection filter of your Communication Links to filter out these JMS acknowledge events.
- TransactionVision: In an environment where the Agents/Sensors can only connect to the Analyzer box via a NAT IP address, the NAT IP needs to be sent in the Configuration Messages for the Time Server address. At the same time, the Analyzer has to know that it doesn't need to check the Time Skew from the Time Server since it's on the same host--it cannot connect to the NAT IP.
Solution: In order to allow a different IP to be specified for the Time Server while still maintaining the "Use the Time Server running in the Analyzer" option in the CommLink, the "Host name" field is now editable with the option checked.
In this case, the different IP will be sent in the Configuration messages, and the Analyzer will still be aware that it should not bother checking time skew from the Time Server since it is on the same host.
<java_agent_install_dir>/TransactionVisionAgent/config/sensor/SensorConfiguration.xml
To enter the required information (see bold text):
...
<!-- for object resolution, we need admin right to access servers
the incoming url could be tcp://heineken:7222 or
tcp://heineken.bristol.com:7222 or using ssl -->
<TibcoEMSServers>
<TibcoEMSServer name="ems_server_name" loginName="XXX" password="YYY">
<ServerUrl>tcp://ems_server_host_name:7222</ServerUrl>
</TibcoEMSServer>
</TibcoEMSServers>
...
- TransactionVision: In some cases, events continue to be collected from the WMQ Sensor despite the fact that collection has been stopped from the TransactionVision user interface. This may be due to orphaned configuration messages on the config queues that had infinite timeout, which would cause them to stay enabled indefinitely. Shutting down the queue managers would resolve the problem, but this is not possible in a production environment.
Solution: To clear out the orphaned messages from the configuration queue, and thereby stop event collection use the following utility to issue an emergency stop message:
In general, the intent of this modification is to create and make available dummy interface modules for IMS and/or CICS by assembling and linking them into the TransactionVision SSLDLOAD library and then make all references to the SDFHLOAD (CICS) or SDFSRESL (IMS) libraries refer to the TV SSLDLOAD library.
Select one of the following dependent on the type of install.
SMP/E Installation:
- Within installation JCL member SLDDDDEF, adjust the %cicshlq and %imshlq substitution values and suffix constants .SDFHLOAD and .SDFSRESL respectively, replacing these values and constants as necessary to generate the dataset name produced by the concatenation %tvhlq..SSLDLOAD.
- Immediately preceding the execution of JCL member SLDDDDEF, execute JCL members DUMYCICS and DUMYIMS as needed (if for example CICS is installed, but IMS is not, then only execute DUMYIMS). Load modules output from either of the DUMY* jobs need to target the library %tvhlq..SSLDLOAD. Also, if DUMYIMS is being used, please insert the following two lines immediately after the line starting with '//L.SYSLMOD DD ':
//L.SYSIN DD *
ALIAS DFSLI000Non-SMP/E Installation:
- Within installation JCL member SLDINSTL, after the DEFINE step, insert JCL members DUMYIMS and DUMYCICS as needed (if for example CICS is installed, but IMS is not, then only insert DUMYIMS). Tailor the DUMY* JCL member(s) to output load modules targeting library %tvhlq..SSLDLOAD. Also, if DUMYIMS is being used, please insert the following two lines immediately after the line starting with '//L.SYSLMOD DD ':
//L.SYSIN DD *
ALIAS DFSLI000
- Within installation JCL member SLDINSTL, after the DEFINE step, insert JCL members DUMYIMS and DUMYCICS as needed (if for example CICS is installed, but IMS is not, then only insert DUMYIMS). Tailor the inserted JCL to output load modules to target library %tvhlq..SDFHLOAD.
- Tailor SLDINSTL replacing references to %cicshlq and %imshlq substitution values and suffix constants .SDFHLOAD and .SDFSRESL respectively, replacing these values and constants as necessary to generate the dataset name produced by the concatenation %tvhlq..SSLDLOAD.
- Execute JCL member SLDINSTL.
- Repeat the above steps when performing non-SMP/E installation of the SLM product code, tailoring SLMINSTL, and substituting SLM for SLD as necessary.
Business Process Insight
- Business Process Insight: The BPI Server has been updated to support Windows Server 2008.
- Business Process Insight: The Process Simulator is a contributed component that is a test tool for Business Process Insight, which is placed in the contrib directory of the installed BPI server. For 8.04, the manual and its PDF names for using the Process Simulator have changed to HP Business Process Insight Using the HP BPI Process Simulator guide (procsim.pdf), also located in the contrib directory.
- Business Process Insight troubleshooting tip for 8.0x versions: Instance violation counts are always zero on the Health page. The Business Health tab in the lower- right pane of the BPI Health page shows how many BPI monitor threshold violations there have been within the most recent sample period. The sample period corresponds to the period at which data samples are sent to BAC. If the sample period is too short, or the clocks are not synchronized between the database and BPI Server hosts, the violation counts may always be zero. If the violation counts on the BPI Health page remain zero when you expect there to be violations, check the following:
- Business Process Insight: Deadline thresholds for Business Process monitors cannot be calculated if the name of the deadline property has greater than ten characters and the BPI Instance Database is on an Oracle server.
An error can occur when calculating the deadline or status for deadline thresholds on Business Process monitors if the BPI Instance Database is on an Oracle server. This error appears in one of two ways:
WARNING: Monitor generation failed for... ORA-00972: identifier is too long: ...
WARNING: Monitor generation failed for... ORA-00904: ... invalid identifier...
Workaround: Change the name of the deadline property to 10 or fewer characters.
Unlike other monitors, those with a type of Absolute Duration raise violations before an instance has completed. When there is a combination of a short absolute threshold duration and the event which completes the instance is constantly delayed, violations are incorrectly raised for every instance. For example, a combination of a critical threshold of 30 seconds and a latency of five minutes for the end events cause a violation.
Workaround: Disable the violation detection for instance duration monitors that have yet to complete. This is achieved by altering the relevant stored procedures (METRIC_THRESH_ACTIV_INSTS_OVER and METRIC_THRESH_ACTIVE_OVERDUE) to not match on active instances:
- For MS SQL, this can be done by doing the following:
- Run Microsoft SQL Server Management Studio (or equivalent).
- Connect to the BPI Instance database (default is hpbpischema).
- Navigate to the Stored Procedures (for example, Programmability > Stored Procedures).
- Select the stored procedure called METRIC_THRESH_ACTIV_INSTS_OVER.
- Select Modify (or equivalent) to modify the stored procedure.
- Replace Active with Never in the stored procedure.
- Click Execute on the toolbar (or equivalent) to apply the change.
- For Oracle, this can be done by doing the following:
- Run Oracle Enterprise Manager (or equivalent) and connect to the BPI Instance database.
- Navigate to the Stored Procedures section (for example, Administration >Procedures) and display the list of stored procedures for the BPI schema.
- Select the stored procedure called METRIC_THRESH_ACTIV_INSTS_OVER.
- Select Edit (or equivalent) to modify the stored procedure.
- Replace Active with Never in the stored procedure.
- Click Apply (or equivalent) to apply the change.
Note: The caveat to the above workaround is that if there are process instances experiencing issues, the absolute duration violations will not be triggered until after the monitor scope for the process instances is complete. Therefore, the absolute duration violations may not be triggered if the active process instances are blocked on a particular step within the monitor scope and none complete the monitor scope. This workaround should only be used if the implications of this are acceptable.
UCMDB Version Content and Limitations
Universal CMDB
- Reduced memory consumption: The compound relationship calculation in the TQL engine now consumes less memory.
- Only an integration user can connect to the UCMDB through the UCMDB Java API set. A connection attempt by other types of users may cause errors, even when using LDAP verification.
- FIXED IN 8.05 (Office Excel 2003 SP3 or Excel 2007 SP2 must be installed for the fix to work): CSV Format Export Troubleshooting: If one of the following reports includes words in the Japanese language and is exported using the CSV format, when the report is opened in Microsoft Excel, the Japanese words are shown as gibberish:
- Export search results:
- Export reports:
- Export CI properties:
- Discovery - Run Discovery - Statistic Results
- Modeling - IT Universe Manager
- Modeling - IT Universe Manager - CI Instances
- Modeling - Modeling Studio
- Modeling - View Manager - Preview
- Modeling - View Manager - Show Element's Instances
- Modeling - Correlation Manager - Preview
- Modeling - Correlation Manager - Element's Instances
- Modeling - CI Type Manager - Show CIT Instances
- Modeling - Topology Report Manager - Preview
- Modeling - Topology Report Manager - Element's Instances
- Modeling - Query Manager - Preview
- Modeling - Query Manager - Element's Instances
- Modeling - Enrichment Manager - Preview
- Topology View
- Compliance - Compare CIs
- Compliance - Compare Snapshots
- Export Discovery ranges:
- Export required permissions:
- Export history:
Workaround: Import the CSV list into Microsoft Excel, instead of opening the file directly. For details, see "Report Issues" in Reference Information.
Discovery and Dependency Mapping
Operational Improvements
- You can now choose to ignore changes on attribute values and focus on real changes occurring in your infrastructure. A managed resource may report a slightly different value for a given attribute, compared to a previous report (for example, a small disk size change from 8.00008 to 8.00009 GB). A new DDM qualifier (ALLOW_VOLATILITY) now automatically ignores these "false" changes.
By default, this qualifier affects the discovered values. To deactivate it, add a model.allow.volatility parameter to the <Business Availability Center Data Processing Server root directory>\conf\settings\cmdb.xml file, as follows::
<setting name="model.allow.volatility"
sectionResource="section.model"
refreshRate="Reboot"
displayInUI="true"
nameResource="model.allow.volatility.name"
descResource="model.allow.volatility.desc"
settingType="customer">
<value type="boolean">false</value>
</setting>>
See also "The ALLOW_VOLATILITY Qualifier" in Model Management.
Modifications to the Support Matrix
Note: When DDM Probe is installed on Windows Server 2008, you must use DDM Content Pack 5 or later (older DDM Content Packs are not supported). To use DDM Content Pack 5 on Windows Server 2008, you must apply the DDM_on_Windows2008_ Patch_for_CP5 patch, provided with the DDM 8.04 package. To apply the patch on top of DDM CP5, deploy the DDM_on_Windows2008_ Patch_for_CP5.zip file through the Package Manager.
Other Improvements
- Discovery and Dependency Mapping Content Packs available on HP Live Network. The HP Live Network Web site has been extended to provide customers with a convenient way to obtain the latest Discovery and Dependency Mapping Content Pack. You need an HP Passport user name and password to log in. The URL for this Web site is: https://h20090.www2.hp.com/.
- Export and Import the domainRangesDocument File. You can export and import the domainRangesDocument.xml file between multiple Business Availability Center instances.
To export a domainRangesDocument file:
- Enter the URL of the JMX console (http://<Data Processing server name>:8080/jmx-console/) in a Web browser.
- Enter your JMX console authentication credentials. If you do not know your authentication credentials, contact your system administrator.
- Under MAM, click service=Discovery manager.
- Locate the exportProbesRanges operation.
- In the customerId parameter, enter 1 (the default).
- In the xmlFileName parameter, enter the name of the file which is to contain the exported Probe ranges. The file is created in the <Business Availability Center Data Processing Server root directory>\mam_lib\server\discovery\<customer_dir> directory.
To import a domainRangesDocument file:
- Enter the URL of the JMX console (http://< Data Processing server name>:8080/jmx-console/) in a Web browser.
- Enter your JMX console authentication credentials. If you do not know your authentication credentials, contact your system administrator.
- Under MAM, click service=Discovery manager.
- Locate the importProbesRanges operation.
- In the customerId parameter, enter 1 (the default).
- In the xmlFileName parameter, enter the name of the file to be imported. The file should reside in the <Business Availability Center Data Processing Server root directory>\mam_lib\server\discovery\<customer_dir> directory.
- Attribute value truncation change: If DDM is configured to truncate the attribute value (when the DDM_AUTOTRUNCATE qualifier is specified), the autoTruncateDbEncoding property now determines which encoding is used for byte transformation.
The autoTruncateDbEncoding property resides in the following file: UCMDB Administration > Discovery > Manage Discovery Resources > Discovery Packages > AutoDiscoveryContent > Configuration Files > globalSettings.xml: <property name="autoTruncateDbEncoding">UTF8</property>
Valid entries are UTF8 (the default) and UTF-16, depending on which encoding is used by the UCMDB database.
jvm 1 | <2010-02-15 12:45:13,843> [WARN ] [WrapperSimpleAppMain] (LocalTopology.java:67) - Could not create network topology for Probe Manager using WBEM
This warning has no impact on the DDM Probe's work and signifies that the Probe's IPv6 address cannot be parsed.
Applet Tag Usage
- Following Sun's recommendations (http://java.sun.com/j2se/1.5.0/docs/guide/plugin/developer_guide/using_tags.html#object), UCMDB code uses the HTML applet tag and not the object tag to launch the UCMDB applet, in version 8.0x. Therefore, users must not disable the Java plug-in applet tag support (Internet Explorer > Tools > Internet Options > Advanced tab > Use JRE xx for <applet> (requires restart)).
Defect Fixes
8.03 Version Content and Limitations
- SiteScope: The Business Availability Center Documentation Library released with Business Availability Center 8.03 includes the SiteScope 10.00 documentation. If you install SiteScope 10.10 for use with Business Availability Center 8.03, you can access the SiteScope 10.10 documentation while in SiteScope in one of two ways:
- Service Level Management: In Business Availability Center 8.03, Service Level Management includes the ability to customize the first day of the month used in agreement calculations, and a new business rule, the EMS Events Availability rule. For information on this new functionality, see new_content_803.pdf located in the 8.03 Setup package root directory.
If you are upgrading to Business Availability Center 8.03 from an existing 8.x installation, the EMS Events Availability rule is not available. Workaround: Run the following JMX operation to reload the repositories:- System Availability Management: Added the ability to control the location of the ellipsis ("...") used in long labels in System Availability Management Reports. This enables you to move the ellipsis from the end to the middle of the label. To use it, add the following properties to the <HP Business Availability Center root directory>\AppServer\webapps\site.war\WEB-INF\rfw\conf\applications\sam-presentation.xml file:
<property key="stringBreakFormattingPolicyStr" value="Break_Middle"/>
<property key="stringBreakCutLocation" value="2" />TransactionVision
- TransactionVision Java Agent extensions: The TransactionVision Java Agent in release 8.03 offers an extension to instrumentation in which code snippets are used to gather additional event data and put that data into user events. A callback mechanism can also be used in this release that allows the user to define additional fields and otherwise control the events being sent to the TransactionVision Analyzer by implementing a callback interface in Java. For more information, see new_content_803.pdf located in the 8.03 Setup package root directory.
- Support for Lightweight Single-Sign-On (LW-SSO) 2.2 has been added to the UI/Job Server.
- User will not be able to access the TransactionVision application from Business Availability Center if they have disabled LW-SSO from Business Availability Center. TransactionVision must also be configured to disable LW-SSO. See chapter 27 in the TransactionVision 8.03 Deployment Guide for details.
- Support for SonicMQ 7.6.2 for event transport has been added.
- The Transaction Flow table is now always ordered by Start Time. This table appears as part of the Transaction Detail page, accessed from the Transaction Tracking report.
- A new option has been added that prevents the WMQ Sensor from attempting to resolve remote cluster queues or queues which are opened remotely through a transmission queue. This option is relevant only if MQ Event Messaging is enabled in your environment, and need to reduce LOCALEV messages being sent to the SYSTEM.ADMIN.QMGR.EVENT queue because of failed MQOPEN operations during WMQ Sensor queue resolution.
To specify this option, set the Do not attempt to resolve cluster or remotely opened queues checkbox. This checkbox is located on the Edit Data Collection Filters > WebSphere MQ > Object > Criterion Settings for Object > New Object group. By default, this option is not specified.
All active data collection filters must have this option checked in order for this resolution to be stopped in the WMQ Sensor. By stopping this resolution, cluster queues cannot be resolved.
- The TVisionSetupInfo script prompts for a database user name/password that is used to access the Oracle, DB2, or SQL Server database to create the TransactionVision tables. Previously, that user had to have Administrative-level privileges; now the user needs less access privileges. See the TransactionVision 8.03 Deployment Guide for more information.
- The Database Tablespace dropdown list on the Create a New Database Schema Page has been changed to a standard text field. The user will not be able to choose from a list of valid tablespace names.
- The WebSphere Business Integration (WBI) Sensor (previously known as the MQSI Sensor) is no longer supported. Please ignore any references to the WBI Sensor that appear in the 8.03 documentation.
Business Process Insight
- Business Process Insight: The Business Availability Center Database Guide has been updated to include information for setting up and configuring the Business Process Insight Process Repository database on the Microsoft SQL and Oracle Servers, and instructions for creating a BPI user schema manually on both servers (see Chapter 5 "Manually Creating the Microsoft SQL Server Database" and "Chapter 12 "Setting Up the Database Environment").
- Business Process Insight: The Transaction tab on the Business Process Insight Health page does not display the related transactions for processes that have steps using TranasctionVision events in their progression rules.
For the Business Process Insight process definitions that rely on event definitions exported from TransactionVision, make sure that the related transaction classes (and CIs) have been updated within the process definitions.
You can verify this by opening the process definitions in the Business Process Insight Modeler and looking at the Properties > Related CIs page for each step that uses those events' definitions. The related CIs page should show all the related Business Transaction CI types. If any of these related Business Transaction CI types appear with the error message, Transaction CI does not exist in CMDB, then the relevant CIs have not been correctly identified and updated in the CMDB.
Workaround: To fix the Business Transaction CI types that have not been correctly identified and updated in the CMDB:
- In Business Availability Center, re-export the event definitions from the Admin > TransactionVision > HP Business Process Insight Settings page to the Business Process Insight Modeler.
- Close and then re-open the relevant process definitions in the Business Process Insight Modeler and verify that all the related CIs of type Business Transaction have been fixed.
- Make a slight content change to the process definitions description and save the changes.
- Re-deploy the process definitions.
- HP Business Availability Center bug fixes - see table below for details:
UCMDB Version Content and Limitations
Universal CMDB
- SMS Adapter. Business Availability Center 8.03 supports integration between HP Universal CMDB (UCMDB) and Microsoft Systems Management Server (SMS). The SmsDBAdapter adapter is based on the Generic Database Adapter. This adapter supports both full and differential replication of all source CIs that are defined in the SmsDBAdapter package, as well as the Service Manager integration flow. For information on implementing this integration, see sms_integration.pdf located in the 8.03 Setup package root directory.
- Service Manager Integration. Improvement in the error handling of locked CIs in SM that used to fail the entire replication job. Changes to these CIs are now saved and replicated in the next replication run. (This process is similar to the change made in version 7.54.)
Discovery and Dependency Mapping
- Client-side SSL certificates. Support of client-side SSL certificates for communication between DDM Probes and the BAC Gateway server. Mutual authentication can now be configured between the DDM Probe and the BAC Gateway server.
- Enhancements to the hardening procedures for DDM and the DDM Probe have been updated in the Discovery and Dependency Mapping Guide PDF only. For changes to the DDM documentation generally, see below.
- Domain Scope Document. This document now provides better security with improved credentials encryption. The document is encrypted using an AES algorithm with a default key of 192 bytes, but customers can decide the length of the encryption/decryption key and other security parameters. In addition, the encryption key is now stored in a more secure manner.
- Domain Scope Document: importing from a previous version. Because the method for importing or exporting the domainScopeDocument.bin file has changed in version 8.03 (users must now provide a password that is used for the DSD encryption or decryption), you should use the new JMX importDomainScopeDocumentWithKey method to import a DSD that has been exported from a previous version. This method also retrieves the name of the encryption key file that was used for the DSD's creation. The encryption key file must be located in the <HP Business Availability Center Data Processing server root directory>\mam_lib\server\discovery\customer_1 directory.
- Domain Scope Document: Limitations.
- The exportDomainsScopeDocument and importDomainScopeDocument methods import only credentials and probe data. IP ranges are not imported with importDomainScopeDocument. To work around this limitation, you should run the importProbeRanges and exportProbeRanges methods for importing and exporting IP ranges.
- During the procedure for changing or generating an encryption key, DSD may become corrupted if you enter a wrong parameter value in the JMX console, for example, wrong JCE provider name or key length. To be able to rectify this problem, export the DSD file before changing or generating the encryption key. If the DSD file does become corrupt, import the DSD file you exported previously. For details on exporting and importing a DSD file, see the Discovery and Dependency Mapping Guide PDF.
- When defining several JCE providers, you must update the files in the following folders: <HP Business Availability Center server root directory>\JRE and <HP Business Availability Center server root directory>\JRE64 (for a 64-bit environment). Perform this update on both the Gateway and the Data Processing servers.
- DDM Credentials Hardening. The result of running the generateEncryptionKey method is a new generated encryption key. This key is stored only in the secured_storage.bin file and its name and details are not known. If you reinstall an existing Probe, or connect a new Probe to the BAC server, this new generated key is not recognized by the new Probe. In these cases, it is preferable to use the changeEncryptionKey method to change encryption keys. This way, when you reinstall a Probe or install a new Probe, you can import the existing key (whose name and location you know) by running the importEncryptionKey method on the Probe JMX console.
- DDM Probe: Manually changing the encryption key when the Probe Manager and Probe Gateway are installed on separate machines:
- On the Probe Manager machine, start the Probe Gateway service: Start > Programs > HP DDM > DDM Gateway.
- Import the key from the server, using the Probe Gateway JMX. For details, see "Generate or Update the Encryption Key" in the Discovery and Dependency Mapping Guide PDF.
- After the encryption key is imported successfully, stop the Probe Gateway service.
- Discovery and Dependency Mapping (DDM) documentation. The documentation now consists of two guides:
The following chapters have moved from this book to the Discovery and Dependency Mapping Content Guide :
The Security section has been updated with the version 8.03 features.
The Hardening the Probe chapter now includes the procedure to update the Probe's MySQL database password.
The HP Discovery and Dependency Mapping Web Service API chapter has moved to the Administration section.
This guide can be accessed from the Discovery Portal in the Business Availability Center Documentation Library, or directly from here: <Business Availability Center Gateway Server>\HPBAC\AppServer\webapps\site.war\amdocs\eng\pdfs\DDMContent.pdf
The following chapters have been added to this book:
Defect Fixes
8.02 Version Content and Limitations
- Dashboard: In Dashboard, the OK threshold value for the Availability and Performance KPIs is set as >=100% for Application Sessions CIs. As a result, if availability or performance is any lower than 100%, OK status is immediately breached. In BAC 8.02, the threshold was corrected to >=90%. Important: This fix is only relevant for new BAC installations (installing 8.00, 8.01, and then 8.02 or later). If you are upgrading to BAC 8.02 or later from an existing 7.5.x/8.x installation, a manual workaround is required. Contact HP Software Support for details.
- Dashboard: If you are upgrading to BAC 8.02 or later from an existing 8.x installation, the Downtime PNR status image is not available. Workaround: Enter the following URL to perform a manual invocation of the Clean Repositories service: http://<GATEWAY_SERVER>:8080/jmx-console/HtmlAdaptor?action=invokeOpByName&name=Topaz%3Aservice%3Drepositories_manager&methodName=cleanRepositoryTypeDBAndCache&argType=java.lang.String&arg0=0&argType=java.lang.String&arg1=dimensions
- Service Manager/Service Center Integration: If you are integrating with HP ServiceCenter 6.2.6, note the below corrections to the Service Manager Monitor topics in the System Availability Management documentation. These corrections are also available in the System Availability Management Guide in PDF format (SystemAvailabilityManagement.pdf) that is located on the HP Software Manuals site (details here):
- Correction to HP Service Manager Monitor User Interface section, HP Service Manager Web Service Endpoint GUI element description: The URL syntax when integrating with ServiceCenter 6.2.6 is: <protocol>://<SM host>:<SM port>/sc61server/PWS/
- Correction to the HP Service Manager Integration Workflow topic, Create the JAR File section, fourth step, description of 6.2.6 URL syntax:
clocks.wsdl.url=http://<SM host>:<SM port>/sc61server/PWS/Clocks?wsdl
prob.wsdl.url=http://<SM host>:<SM port>/sc61server/PWS/IncidentManagement?wsdl
- Diagnostics Integration: In BAC 8.00, a Location selector was added to the Diagnostics Application KPI. However, the Diagnostics sample does not have a Location parameter, so the KPI is always grey. From BAC 8.02, the Location selector was removed from the Diagnostics Application KPI. As a result, even when working in transaction by location mode, the Application status is calculated according to all locations. Important: This fix is only relevant for new BAC installations (installing 8.00, 8.01, and then 8.02 or later). If you are upgrading to BAC 8.02 or later from an existing 7.5.x/8.x installation, a manual workaround is required. Contact HP Software Support for details.
- TransactionVision: Exception errors may occur when executing the Export BPI Definitions command. If a BPI action is modified in the Transaction Definition Editor, be sure that no required parameters are removed. If they are, an exception error will occur when the Export BPI Definitions command is executed.
- TransactionVision: Java Agent installation and Setup Module response files generated with version 8.00 cannot be used to perform a silent installation of the 8.01 or 8.02 version of the Java Agent installation and Setup Module. Response files generated with 8.01 will work with 8.02 and later releases.
- FIXED IN 8.03: TransactionVision: After upgrading TransactionVision from 7.50 to 8.0x, there is no flow map showing in Aggregate Topology or Instance Topology reports. Workaround: After you perform an upgrade install of TransactionVision Analyzer from 7.50 to 8.0x, add the following lines into the <TVISION_HOME>/config/services/Beans.xml file (right before the </BeanFramework> line):
<Module name="TxnUpdateCtx" type="Context">
<!-- This context contains beans that perform transaction updates -->
<!-- Only one bean in the context is called (based on its type) -->
<Module class="com.bristol.tvision.services.analysis.txnupdate.RUMTxnUpdateBean" type="Bean">
<Attribute name="type" value="RUM"/>
</Module>
</Module><Module name="com.bristol.tvision.services.analysis.transactiontopology.TransactionTopologyBeanCtx" type="Context">
<Module class="com.bristol.tvision.services.analysis.transactiontopology.TransactionTopologyWMQBean" type="Bean"/>
<Module class="com.bristol.tvision.services.analysis.transactiontopology.TransactionTopologyJMSBean" type="Bean"/>
<Module class="com.bristol.tvision.services.analysis.transactiontopology.TransactionTopologyServletBean" type="Bean"/>
<Module class="com.bristol.tvision.services.analysis.transactiontopology.TransactionTopologyEJBBean" type="Bean"/>
<Module class="com.bristol.tvision.services.analysis.transactiontopology.TransactionTopologyJDBCBean" type="Bean"/>
</Module>If you integrate Business Process Insight 8.0x with TransactionVision 8.0x where the minor versions of Business Process Insight and TransactionVision do not match, and you get an error message that the client version does not match the server version when exporting definitions to Business Process Insight, use the following workaround.
Workaround: To make TransactionVision and Business Process Insight compatible, stop the TransactionVision UI/Job server, copy the following file from the Business Process Insight server machine to the TransactionVision UI/Job Server machine and start the TransactionVision UI/Job server after copying the file.
<BPI-install-dir>\nonOV\jakarta-tomcat-5.0.19\webapps\hpbpi-bpd\WEB-INF\lib\bia-flowviewer.jar
<TV-install-dir>\java\lib\bia-flowviewer.jar
- FIXED IN 8.04: Business Process Insight: With Business Availability Center 8.02 or later, if you launch the BPI modeler (from the BPI > Admin > Modeler application), navigate to the Process Editor tab for any process step, and launch the CI Selector, the Business Availability Center user interface displays an error message such as, "An internal error has occurred." Workaround: Select the Site Map menu in the Business Availability Center user interface to restore the Business Availability Center user interface to its normal behavior--the errors are no longer encountered. Contact HP Software Support for details on when a fix to this issue will be available.
- Business Process Insight: Implemented a way to link sub-processes to parent processes.
A minimal solution has been implemented with the 8.02 release, which will be released in the contrib directory of the installed Business Process Insight server. See the readme.htm file in the contrib directory for instructions.
- (SEE UPDATED NOTE IN 8.04): Business Process Insight: The Process Simulator is a contributed component that is a test tool for Business Process Insight, which is placed in the contrib directory of the installed BPI server. For details on using the Process Simulator, see the HP Business Process Insight Using the HP BPI Flow Simulator guide (flowsim.pdf), also located in the contrib directory.
- Business Process Insight: If you specify a Business Availability Center user name other than admin during the Business Process Insight installation it may incorrectly report: The supplied user name or password is invalid. You may also see this error if the Business Availability Center admin password has been changed from the default.
Workaround: Choose the continue option to complete the installation. The Process Repository database details need to be entered manually but the installation should complete successfully and Business Process Insight should operate without problems as long as the name and password that you entered are valid credentials for the Business Availability Center server.
- Business Process Insight: The processing and display of Business Process Insight monitor instance values is only done to a precision of 6 or 7 digits, and not to the precision of 15 digits used for process values. This can mean that large integer values may not show their exact value. For example, $123,000,000 could show as $123,000,064.
- Business Process Insight: Business Process Insight Landing pages are not displaying monitor values and statistics to double precision.
- Business Process Insight: Business Process Insight instance pages fail when first accessed in a multi-domain LWSSO configuration.
Workaround: Access the Process Repository Explorer under the Business Process Insight Admin area in Business Availability Center (Admin->Business Process Insight->Process Repository Explorer). After doing this, the process instance pages can be accessed normally.
- Business Process Insight: If you integrate Business Process Insight 8.0x with Business Availability Center 8.0x where the minor versions of Business Availability Center and Business Process Insight do not match, and when launching the Business Process Insight Modeler you get an error message that the Modeler version does not match the repository version, use the following workaround.
Workaround: To make Business Availability Center and Business Process Insight compatible, stop the Business Availability Center Gateway servers, copy the following files from the Business Process Insight server machine to the Business Availability Center Gateway machines and start the Business Availability Center Gateway server after copying the files.
<BPI-install-dir>\nonOV\jakarta-tomcat-5.0.19\webapps\hpbpi-bpd\WEB-INF\lib\bia-flowviewer.jar
<BPI-install-dir>\msg\bia\model_repository.properties
<BAC-install-dir>\AppServer\webapps\bpi.war\WEB-INF\lib\bia-flowviewer.jar
<BAC-install-dir>\AppServer\webapps\bpi.war\jsps\bpi\admin\modeler\lib\bia-flowviewer.jar
<BAC-install-dir>\AppServer\resources\model_repository.properties
<BAC-install-dir>\conf\bpi_process_repository\model_repository.properties
UCMDB Version Content and Limitations
Universal CMDB
- Aging Mechanism. The mechanism is now more user-friendly and mature, and is now configured and managed through the CI Lifecycle page (Application > CI Lifecycle). (By removing CIs that no longer exist in the IT environment, aging reduces the load on the UCMDB and database servers, and improves server usability and stability.) For details, refer to the "CI Lifecycle and Aging Mechanism" chapter in the Model Management guide.
- The UML Tool now enables the export of documentation about the CITs and their attributes in the class model. The tool can export to files with the XMI format.
Federation and Replication
- Get full layout plug-in. The Generic Database Adapter default implementation now includes a get full layout plug-in in full synchronization, for replication. For more information, refer to the "Generic Database Adapter" chapter in the Model Management guide.
- Using Inline SQL Statements Instead of Table Names. For details on this functionality, refer to the "Generic Database Adapter" chapter in the Model Management guide.
Security Manager
- You can now grant permissions to users for the following general actions:
- CI Search. Grants permissions to users to access the CI Selector search option. Only users with these permissions can search for CIs.
- Get Related. Grants permissions to users to view data of related CIs. Only users with these permissions can view related CIs.
- Data Modifier. Grants permissions to users to change CIs in ITU. Only users with these permissions or users with edit rights on a view (as currently) can modify data. Furthermore, users with these permissions can delete CIs from the CI Instances grid.
- Users can access View Manager and Modeling Studio only if they are given permissions to do so.
- A user without any permissions or roles is no longer able to log in to Business Availability Center.
- Security Roles
- Security Upgrade
- You do not need to perform any special steps to upgrade security from version 8.00 and 8.01 to 8.02.
- All roles are granted rights for the Get Related and CI Search actions.
- All roles are granted access rights to View Manager and the Modeling Studio.
- After you upgrade to version 8.02, users can continue working as usual. All current permissions are maintained.
- Security Fixes:
Discovery and Dependency Mapping
# Fix used to run DiscoveryAnalyzer from Eclipse (should be true if running from eclipse else should be false)
appilog.agent.local.discoveryAnalyzerFromEclipse = trueIntegration Documentation
Defect Fixes
8.01 Version Content and Limitations
- Authentication: When HP Business Availability Center is installed with an Oracle Database and User Synchronization enabled with an LDAP Active Directory server, ensure that you login to HP Business Availability Center with the correct-case UID (uppercase or lowercase), as configured on the LDAP server. This is because while the Oracle Database is case-sensitive, the LDAP Active Directory is not case sensitive, and logging in with an incorrect case UID can create undesirable results.
For example, if a user called testuser exists on the LDAP Active Directory server and logs into HP Business Availability Center, he is automatically created as HP Business Availability Center user testuser, who can be assigned permissions in the HP Business Availability Center User Management interface. If you then login to HP Business Availability Center as Testuser, the LDAP Active Directory server sends an acknowledgement that the user exists (because Active Directory is not case sensitive) and he is allowed entry to HP Business Availability Center. However, since the Oracle Database does not identify this user as testuser (because the Oracle Database is case sensitive), the user Testuser is treated as a new user, without the permissions that were assigned to testuser.
- Universal CMDB: When selecting First Service View in Universal CMDB View Manager, the following error is displayed: "Selected Configuration Item could not be loaded." Workaround: Go to JMX Console > UCMDB > Service=Views Services and invoke upgradeOldInstanceView() with parameters CustomerID=1 and bacViewName=First Service View.
- SiteScope: In SiteScope, Solution Templates can only be deployed from the template tree.
- TransactionVision: Custom correlation based on lookup key and time range match up available:
In the area of event correlation, there are certain scenarios where perfect correlation data is not available. While there may be enough correlation information--based on either standard technology context or customer specific business data--to triage the events and limit the matching event candidates to a minimum set, additional factors need to be considered to complete the correlation process and result in one-to-one event relationship.
One such factor is event time stamps. In certain situations, TransactionVision can correlate specific event candidates by considering the respective event execution time. One example: two events A and B representing EJB X and Y method calls respectively are reported to TransactionVision, with EJB X method invoking the EJB Y method. In this case TransactionVision can reasonably assume that the two events' exit timestamps are very close based on the call latency nature. Thus TransactionVision can deduce that by truncating the respective event time stamps to a precision consistent with the expected latency, the two events would have the same truncated timestamp, and this can in turn be considered as a matching key in a lookup key based correlation algorithm.
TransactionVision provides a correlation component that supports this type of time based correlation enhancement. One can consider this as a correlation algorithm based on a mix of time and payload/technology specific data keys. Contact HP Software Support for more information about using this component.
- TransactionVision: Java Agent installation and Setup Module response files generated with version 8.00 cannot be used to perform a silent installation of the 8.01 Java Agent. Generate new response files to be used with the 8.01 version.
- TransactionVision: On Solaris hosts, the TransactionVision Analyzer or UI/Job Server may fail to start upon installation. This occurs when either of these components is installed by a root user and Business Availability Center was started under another user/group on the same host. Therefore the Business Availability Center nanny process does not have permission to execute the TransactionVision components.
Workaround: Change the user/group of the /opt/HP/TransactionVision directory to the same as that being used by Business Availability Center by using the following commands: chown -R <bacuser> /opt/HP/TransactionVision and chgrp -R <bacgroup> /opt/HP/TransactionVision (where <bacuser> and <bacgroup> are the user/group of the Business Availability Center installation directory). Alternatively, you can grant write permissions for all users/groups to the TransactionVision directory by using the command: chmod -R a+w /opt/HP/TransactionVision, but this alternative would open up read/write access for the TransactionVision installation files to all users and may pose a security risk.
- FIXED IN 8.03: TransactionVision: Out-of-memory errors can occur in a .NET application being monitored. This occurs when the .NET Agent is configured in "TV" mode to send events to the TransactionVision Analyzer using SonicMQ as the messaging middleware provider and the SonicMQ Broker is not configured with security.
Work Around : To avoid this problem, enable security in the SonicMQ Broker by requiring authentication. Specify the user id\password credentials in the <sensor_install_dir>/etc/probe_config.xml file. Information about implementing security in SonicMQ Broker is described in the Progress SonicMQ Deployment Guide. This manual is available on any host on which the Analyzer is installed; see <analyzer_install_directory>\Sonic\Docs7.5\books\mq.deploy.pdf.
TransactionVision Error(FrameworkInitError): Failed to initialize Analyzer framework.
TransactionVision Error(CouldNotLoadBeanClass): Bean class 'XXXXX' specified in beans config file could not be loaded. Please see error log for details.
set ANALYZER_ADDITIONAL_CLASSPATH=<service_additional_classpath_value>
In addition, add the following line to the SetupEnv.bat file:
set CLASSPATH=%CLASSPATH%;%ANALYZER_ADDITIONAL_CLASSPATH%
On UNIX, copy and paste the similar entry from the backup SetupEnv.sh file to the new one, and export the updated CLASSPATH.
set CLASSPATH=%CLASSPATH%;%ANALYZER_ADDITIONAL_CLASSPATH%
On UNIX, copy and paste the similar entry from the backup SetupEnv.sh file to the new one, and export the updated CLASSPATH.
Workaround: Add the following lines to the end of the <TVISION_HOME>/exit_sensor.deny file:
*\WebSphere MQ\java\jre\bin\javaw
*/mqm/java/jre/bin/javaw
- FIXED IN 8.02: Business Process Insight: Incorrect backlogs result when monitor instances complete before they start.
- FIXED IN 8.02: Business Process Insight: In Business Availability Center versions 8.00 and 8.01, sometimes the annotations on the Business Availability Center Dashboard flow map tab's process diagram read No Data even though the data samples were being sent to Business Availability Center from Business Process Insight.
Workaround: Edit the \HPBAC\conf\settings\bpiSettings.xml infrastructure settings file and change the setting with the name attribute: settings.bpi.profile.sample.multiple to have a displayInUI attribute value of true and a value attribute value of 10. Restart Business Availability Center once these changes have been made. After Business Availability Center has been restarted the Multiple of sample periods queried property can be modified in the Business Process Insight Infrastructure Settings Profile Database Data Sample Retrieval Settings section.
- Business Process Insight: If you specify a Business Availability Center username other than admin during the BPI installation it may incorrectly report "The supplied user name or password is invalid." You may also see this error if the Business Availability Center admin password has been changed from the default. If you see this, you should choose the option to continue with the installation. The Process Repository database details will need to be entered manually but the installation should complete successfully and BPI should operate without problems as long as the name and password that you entered are valid credentials for the Business Availability Center server.
- Business Process Insight: Business Process Insight cannot use MS SQL named instances, such as <hostname>\<instance>, to specify the database.
Workaround: During the installation of Business Process Insight, identify the host and specific port on which the named instance is running and enter the following during the Business Process Insight installation:
MS SQL Server hostname: enter the host
MS SQL Server port: enter the specific instance port number
For example: Instead of entering (for MS SQL Server hostname) myhost\fred and (for MS SQL Server port) 1433, you find that the fred instance is running on port 1244 and use:
MS SQL Server hostname: myHost
MS SQL Server port: 1244
Alternatively, you can use forward slashes (/) when specifying the named instance. For example, instead of entering (for MS SQL Server hostname) myhost\fred you can use:
MS SQL Server hostname: myHost/fred
MS SQL Server port: 1244
- FIXED IN 8.02: I18n: Exporting a report to Microsoft Excel format may display data with the wrong values if the browser locale is set to a language in which a number's decimal point is represented by a comma (",") and not a period ("."). This can happen with languages such as Hungarian, German, or French. In these cases, the user is able to see the number in the correct format in Business Availability Center (for example 33,333) but if the data is exported to Microsoft Excel, the wrong value is exported (for example 33333).:
- Workaround Option 1: Change the browser language to English and export the file to Microsoft Excel. The correct values appear but in English format (with a period instead of a comma). Once the data has been exported to Excel, you can change the decimal value format to a comma by selecting Control Panel > Regional and Language Options > Regional Options tab > Customize button > Decimal symbol enter a comma (,).
- Workaround Option 2: Use the export to XML option. In XML format, the values are correct and in the correct locale format.
- I18n: When using a language that reads right to left (for example Arabic or Hebrew), exporting a report to PDF format does not work properly. The letters appear backwards in the PDF. Viewing the report in Business Availability Center works properly.
- HP Business Availability Center bug fixes - see table below for details:
8.00 Notes and Limitations
General, Deployment, and Upgrade
- See the upgrade notes in this section for important upgrade information regarding Universal CMDB (UCMDB).
- Upgrading to version 8.00 on a Solaris operating system is not yet supported and will be delivered after the general release of 8.00. Contact HP software support for details on when this will be available.
- HP Business Availability Center comes with Lightweight Single Sign-On (LW-SSO) capability, which enables you to log in to a single application and then permits you to access other supported applications configured in the original application's group. For limitations, security warnings, and general reference on using LW-SSO, see the Lightweight Single Sign-On Authentication - General Reference section in the Platform Administration guide of the HP Business Availability Center Documentation Library.
- If you are using Solaris version 9.0, run HP Business Availability Center using the Sun Java System Web Server and not Apache. You may encounter problems running HP Business Availability Center on Solaris version 9.0 with Apache. (The Apache web server installed during HP Business Availability Center server installation has been adapted for use when running on Solaris version 10.0.)
- After upgrading from HP Business Availability Center 7.5x to HP Business Availability Center 8.0x, any Favorite Filters for Trend Reports configured in HP Business Availability Center 7.5x cannot be viewed in HP Business Availability Center 8.0x. Once you have completed the upgrade, you must recreate your Trend Report Favorite Filters in HP Business Availability Center 8.0x.
- Favorite filters created for CI Status Alert and SLA Status Alert reports in previous versions are not upgraded and will not work in Business Availability Center 8.0. You must recreate the favorite filters.
- If you cleared any of the Include in View and Include in CI Status options for the KPIs of an instance view in Dashboard > Admin > KPIs > KPI Selection, before upgrading to 8.00, any customization of the Zoom feature in Dashboard > Admin > Geographical Map is not upgraded, the Zoom feature is reset, and the Include in View and Include in CI Status options are reset to be selected for all KPIs in the instance view. You must clear the appropriate Include in View and Include in CI Status options for the KPIs in the view and customize the Geographical Map Zoom feature for the view after the upgrade.
- If the BPI Modeler fails to deploy business processes and reports CMDB synchronization error you may need to manually remove business processes and related CIs (steps and all BPI monitors) from the CMDB that were created by BPI 2.20 on Business Availability Center 6.x (from 6.8) or 7.0. Business Availability Center 8.00 (and Business Availability Center 7.5x) do not support upgrade of business processes created in the CMDB either manually or by BPI 2.20. Any business processes created either manually or by BPI on a Business Availability Center 6.x (from 6.8) or 7.0. system should be removed, and the Business.zip UCMDB package redeployed, before business processes can be deployed on Business Availability Center 8.00 from the BPI Modeler.
- When viewing Business Availability Center in Internet Explorer 7.0, it is recommended to disable the address bar to avoid a vertical scroll bar appearing on some pages. To disable the address bar in Internet Explorer 7.0, select Tools > Internet Options > Security tab > Internet icon > Custom Level and enable the Allow websites to open windows without address or status bars option.
- An Internet Explorer exception error message is thrown inside Business Availability Center pages (usually on login page), indicating Operation aborted. To resolve this issue, remove the Skype add-ons on your system. For further details, see http://support.microsoft.com/default.aspx/kb/927917
- After performing an upgrade, the tooltips of cloned and overridden KPIs are not updated with the colors used in the new version. The new colors can be applied by cloning or overriding the KPI again, and redefining the custom KPI's fields.
- The following configuration is required if a SOAP adapter is used for replication via reverse proxy into a secure (hardened) HP Business Availability Center server architecture:
Requests for /axis2/* on the Reverse Proxy Server should be served by the following URL: http://<HP Business Availability Center GATEWAY server>/axis2/*
- The following issue is relevant if you are performing an upgrade in staging mode in an HP Business Availability Center environment in which HP Diagnostics is integrated and you want to register a different Diagnostics server to the staged Business Availability Center instance. During the staging process, do not unregister the Diagnostics server from within the Diagnostics registration page in the staging environment. Doing so will unregister the Diagnostics server from the production environment. Instead, change the Diagnostics server host registration directly in the Business Availability Center management database. Update the CUSTOMER_PROP table with the correct Diagnostics server name by running the following query against the management database: WAITING FOR ANSWER FROM YISRAEL
Update CUSTOMER_PROP set CPP_PROP_VALUE='<<<New Diagnostics server host name>>>' Where CPP_PROP_NAME=' OPAL_FULL_HOST'
- When opening the Upgrade Wizard for the second time, HP Business Availability Center must be up for the wizard to be able to access the relevant pages.
- On a Solaris platform, when you click the Help button in the Upgrade Wizard, the Help is not displayed.
- The instructions in the HP Business Availability Center 8.00 upgrade wizard, and the license information displayed when installing HP Business Availability Center in a non-English locale, are displayed only in English.
- On machines with four or more CPUs, the offline aggregation engine may crash due to a bug in the Sun Hotspot JVM. When this happens, the event will be logged in the file < HP Business Availability Center root directory>\log\offline_engine_boot.log with the following JVM Internal error ID:
# An unexpected error has been detected by HotSpot Virtual Machine:
#
# Internal Error (4E4D4554484F440E4350500529)If this problem occurs on your system, apply the following workaround:
For example:
set SERVICE_MANAGER_OPTS=-DhacProcessName=%PROCESS_NAME% -Dlog.folder.path.output=%PROCESS_NAME% -Dlog.folder=%PROCESS_NAME% -DuseCustomClassLoader=true -XX:ParallelGCThreads=1
- FIXED IN 8.01: If while running the upgrade wizard, the step for upgrading the CMDB fails, this could be due to invalid business view enrichments. Perform the following procedure in the JMX console to determine whether there are invalid enrichments and to remove them from the CMDB.
- Open the JMX console on the Data Processing Server machine (http://<Data Processing Server name>:8080/jmx-console/).
- Locate the service=CMDB Dal Services JMX and invoke the getCmdbVersion() method. If the version of the BusinessViewEnrichmentsUpgrader.PackageManager upgrader is 8.0.0.0, skip the rest of this procedure as this is not the reason for the failure in the upgrade wizard. Refer to the HP Business Availability Center Deployment Guide for troubleshooting upgrade CMDB issues. If the version is not 8.0.0.0 or blank, continue with the procedure.
- Locate the service=CMDB Business View Enrichment Services JMX and invoke the retrieveAllEnrichmentNames() method. Save the results.
- Locate the service=MAM Views Services JMX and invoke the getAllViewsNamesWithTqlNames() method. Search for views with the same names as the business view enrichments saved in step 3 and retrieve their TQL names from the result.
- Locate the service=CMDB Tql Services JMX and invoke the retrieveTqlNames() method. Search for the TQL names retrieved in step 4.
- If there are TQLs missing, the corresponding business view enrichments are invalid and must be removed. Remove the invalid business view enrichments as follows:
- Restart the Data Processing Server machine.
My BSM
- The Topology Map portlet has been discontinued, and is no longer available from Dashboard or My BSM. To view a topological perspective of CIs, select Applications > Universal CMDB > Topology View.
- The infrastructure setting Maximum height in pixels, which you use to standardize the height of portlets, is applied to only a limited number of portlets. For example Dashboard portlets have their own height parameter in filter settings.
- When viewing EMS events in My BSM, the Clear Event option on the context menu does not work. If required, clear the event from Dashboard.
- In My BSM, in Personalize, if you right-click a page and do not select an option, the context menu stays open. To close it, click anywhere in the left pane.
Dashboard and Reporting
- Business Availability Center 8.00 includes Dashboard Ticker 7.50. Dashboard Ticker 7.50 is compatible with Business Availability Center 8.00. Dashboard Ticker is being discontinued as of the next major Business Availability Center release and is being replaced by the Dashboard Widget, already available in Business Availability Center 8.00.
- Some KPIs that should not be propagated to parent CIs are automatically propagated for CIs in hierarchies whose links were created by adapters or DNode.
- The Topology Map portlet has been discontinued, and is no longer available from Dashboard or My BSM. To view a topological perspective of CIs, select Applications > Universal CMDB > Topology View.
- When editing a parameter for a rule in the Business Rules repository, changing the value in the Name box actually changes the value of the reference key for the parameter. Since the reference key is used for the rule calculation in the Business Logic Engine, changing the value may cause the calculation to fail. Do not edit the display name for a parameter.
- The PNR KPI in Dashboard does not have persistency in case of a Data Processing Server failure or a system restart. The PNR KPI shows an initialized status until the SLA is recalculated, at which time the PNR value is updated.
- FIXED IN 8.01: All HP Business Availability Center perspectives include all KPIs (including non-relevant KPIs). The problem can be seen, for example, when creating a model that includes end-user CIs and SiteScope CIs. You then generate a view from the model by applying the End User Monitors perspective; the view displays End User CIs without the SiteScope CIs. In Dashboard, the view displays all the KPIs of the end user CIs as well as those of the SiteScope CIs. Workaround: Access the view in Dashboard > Admin and exclude the non-relevant KPIs from the relevant views.
- Reports that include the time period during which a Daylight Savings Time (DST) change took place (change to or change from DST) will either be missing one hour (in cases where clocks are moved forward) or have two sets of data for the same hour (in cases where clocks are moved back).
- To view reports, Adobe Flash Player 8.0 or later is required.
- If a report's time range includes the time of a Daylight Savings Time (DST) change, the data for the period after the DST time change is incorrect. To obtain correct data for the entire time range, print two separate reports, one for the time range before the DST time change and one for the time range after the DST time change.
- Do not create a Dynamic Node Factory CI that includes Business Process Monitor transactions; If Business Process Monitor transactions are included in the Dynamic Node Factory, there may be a conflict between the adapter and the Dynamic Node Factory mechanism that could lead to an unpredictable state of the CIs.
Service Level Management
- During SLA creation or editing using the Agreement Wizard in CI-based mode, the Define KPIs page may not display CIs that were added in the previous Select CIs page. To update the display, click the Refresh button in the View Explorer pane.
- In Service Level Management, you cannot attach more than one PNR KPI (defined in Dashboard) to a CI.
- If the KPI Definition window has not been closed by clicking the OK or Cancel button, Service Level Management does not display the correct objectives for a KPI on the next occasion that the window is accessed.
- If the location name is changed for a Business Process Monitor after the associated CI has been added to an SLA, the CI is no longer included in the SLA's calculations and the CI's descendants are not displayed in the SLA hierarchy. Furthermore, Service Level Management cannot recalculate the SLA. To prevent this occurring, avoid making changes to the location name of a Business Process Monitor for which SLAs are defined. Also, configure the Business Process Monitor source (in the Source Manager) to use the "Regular" hierarchy structure and not "Transaction/Location."
- In the Agreement Wizard in CI-based mode (on the Select CIs page), Service Level Management may take a few minutes to process an SLA that includes many configuration items, before continuing to the Define KPIs page.
- You cannot add a CI to an SLA when the CI includes a copy of itself in the original definition in the CMDB. This is because nesting is meaningless in the SLA calculation (Service Level Management cannot calculate a CI using the CI result as a value). If you add such a CI to an SLA, Service Level Management truncates the CI at the point at which the CI appears for the second time in the SLA hierarchy. However, you can add the same CI to an SLA more than once if the CI does not include a descendant of itself. In this case, Service Level Management calculates the CI only once, so the objectives are the same no matter where the CI appears in the SLA hierarchy.
- Service Level Management KPIs:
- The "No data time out" parameter is relevant only when "Time Based" is chosen for the "Calculation Method" parameter (that is, the rule calculates the KPI using the sample duration).
- The External Source Average Value rule includes a further parameter: "Performance field." This is the name of the performance field in the external source sample.
End User Management
Business Process Monitor
- Business Availability Center 8.00 is being released with Business Process Monitor 7.52.
- When you run a script written in VuGen that uses proxy settings from an automatic configuration script (.pac file) configured in the run-time settings of the script, if one of the proxies in the .pac file is unavailable, replaying the script in either VuGen or Business Process Monitor may result in a high response time and, in some instances, the script may fail due to a timeout. The script can be run in Internet Explorer without this problem, as it ignores the unavailable proxy.
- Business Process Monitor does not support TCP-based WebTrace (default) on Windows XP SP2 platforms. On Windows XP SP2 platforms, traceroute (ICMP-based WebTrace) is supported. (WebTrace is supported for Windows 2000 or 2003, and Windows XP without SP2.)
- Business Process Monitor registers with HP Business Availability Center using http://<HP Business Availability Center Gateway Server name>/topaz, and not with HPBAC or MercuryAM at the end of the URL. You must use this format when creating instances for Business Process Monitor.
- Large scripts (larger than 10 MB) cannot be uploaded to the database.
- When running a page component breakdown (from End User Management or Business Process Monitor), the target URL is limited to 144 characters.
Real User Monitor
- In the Custom Report Manager, report components of the Real User Monitor (Old) type may be removed in forthcoming versions. It is recommended to remove these report components from your custom reports and to replace them with the corresponding new report component of the Real User Monitor type.
- When Business Availability Center is working with SSL and requires client-side certificate, Real User Monitor session replay will not work
System Availability Management
- If a SiteScope is restarted while being accessed in System Availability Management Administration, you may receive an error and have to renavigate to the SiteScope to properly access the SiteScope server.
- If you are copying from a SiteScope version 10.00 to another SiteScope version 10.00 using the copy/paste functionality in System Availability Management Administration, there may be a delay of up to a minute until the pasted object appears in the target SiteScope. This is because the monitor tree in the new SiteScope user interface refreshes by default five seconds after a paste operation completes. If the paste operation takes longer than 5 seconds, it will miss this automatic refresh and the pasted objects appear only after the regularly scheduled refresh (by default 60 seconds).
- When creating a Technology Integration monitor for integrating measurements into HP Business Availability Center (by selecting Measurements as the Sample type in the Field Mapping area of the monitor properties), SiteScope reports system topology data for the monitor to HP Business Availability Center. This is the standard host topology SiteScope reports to HP Business Availability Center. This cannot be changed by adding a script in the Topology Settings area. If you add a topology script, the script will fail.
- In System Availability Management Administration, when copying objects from one SiteScope to another, there is no dependency check. This means that you can copy an object that depends on another object to a different SiteScope, even though the object on which it depends does not exist on the target SiteScope.
- In System Availability Management Administration, when the SiteScope service is down, instead of receiving an error to this effect, you will see a Page cannot be displayed error.
- If a SiteScope is being reset when HP Business Availability Center is not running, the SiteScope will appear as registered with a configurable profile in System Availability Management Administration once HP Business Availability Center is running again. The SiteScope must be reset when HP Business Availability Center is running.
- On the New Monitor page, the Recently Used Monitors list is not saved when you exit System Availability Management Administration.
- When creating a SiteScope profile in System Availability Management Administration, if you change the profile display name after having received any error during the profile creation process, the default profile name is not changed accordingly. Ensure that the profile name and SiteScope agent machine location fields contain the desired values.
- When working in languages other than English, some event detail information in the Event Log is still in English.
- In System Availability Management Administration, there is no notification for successful actions in the user interface.
TransactionVision
- The TransactionVision Analyzer is installed under /opt/HP/TransactionVision on Solaris and RedHat Enterprise Linux, and under /usr/lpp/HP/TransactionVision on AIX. While it is not possible to change the installation path, it is possible to create soft-links to ensure that the product files are placed on a drive or partition with available space. Under normal circumstances, customers may create a soft-link from /opt/HP/TransactionVision or /usr/lpp/HP/TransactionVision to a new location prior to running the tvinstall_800_unix.sh installation script. However, due to a limitation in the SonicMQ installation with respect to soft-links, the SonicMQ files must be installed under the default /opt/HP/TransactionVision/Sonic or /usr/lpp/HP/TransactionVision/Sonic directory. If space is a concern, customers may still create soft-links for other higher capacity directories (such as jre, jre64, java, docs, and logs) under the TransactionVision installation directory.
- On AIX systems, after the Java Agent installation, the following message will appear:
The Installation Wizard has successfully installed HP Diagnostics/TransactionVision Agent for Java. Choose next to continue the wizard.
However, on some systems, the following warning is also generated: o null.
In this case, the installation has succeeded and the message can be ignored.
- If your HP TransactionVision UI URL setting is configured using https:// in HP Business Availability Center Infrastructure Settings Manager, you must also access HP Business Availability Center using https://. Logging into HP Business Availability Center using http:// in this case would produce "mixed mode content" errors on each page.
- When starting HP Business Availability Center on Solaris systems using the "run_topaz start" command, "Bad String" errors may be seen if the LANG environment variable is not set properly. These errors can be eliminated by setting LANG to the proper locale prior to starting BAC. For example, run the command "export LANG=en_US.UTF-8" to set LANG for U.S. English.
- FIXED IN 8.01: Basic Authentication is not supported when using a TransactionVision UI/Job Server with a BAC Gateway server.
- TransactionVision does not currently support IDM-SSO environments.
- If transaction class definitions are imported from a TransactionDefinition.xml file that does not have a dbschema tag defined, they will have an invalid schema assignment. In this case, classes need to be manually assigned to specific schemas in the TransactionVision Transaction Definition. This applies to the upgrade scenario as well. If class definitions are migrated from an earlier version of TransactionVision, make sure to manually assign those classes to one or more of your project schemas.
- FIXED IN 8.01: If the TransactionVision UI/Job Server is switched to point to a different Business Availability Center server or one that had its CMDB re-initialized, applications may not be associated with their respective transactions in the top dashboard view. Re-syncing the CMDB does not fix this problem since CMDB updates only affect business transaction classes and not system model information. One solution is to drop and re-create the project schema. Another solution is to stop the TransactionVision UI/Job Server and execute the following SQL after logging into the TransactionVision database:
update <project_name>.SYS_MDL_OBJECT set CMDB_ID=null
update project_name>.SYS_MDL_OBJECT_RELATION set CMDB_ID=null;
commit;
After executing these SQL commands, restart the TransactionVision UI/Job Server and ensure that the CMDB population job is running.
- FIXED IN 8.01: In this TransactionVision release, SiteScope Configuration Item status is not shown in the Aggregated Topology Flowmap.
- Customers who plan to enable the TransactionVision IMS sensor on z/OS should ensure that they are not already using the IMS OTMA Input/Output exit (DFSYIOE0) for some other non-TransactionVision purpose before proceeding with the IMS sensor configuration. If DFSYIOE0 is already being used for customer internal purposes or by another vendor product, please contact TransactionVision customer support for detailed instructions on how to modify DFSYIOE0 source code to invoke the TransactionVision IMS sensor (SLMYIOE0).
- Starting with z/OS version 1.9, the default setting for system parmlib DIAGnn member statement VSM ALLOWUSERKEYCSA has changed to "NO." This parameter disallows User Key CSA storage for security reasons. In order for TransactionVision 8.00 Sensors to execute successfully under z/OS, the VSM ALLOWUSERKEYCSA parameter value must be set to "YES."
- When using the Diagnostics/TransactionVision Agent to monitor WebLogic on Linux, the following error may be seen due to a problem with the version of the JVM shipped with that version of WebLogic, which causes the server to not start:
<Dec 3, 2008 5:41:17 PM PST> <Critical> <WebLogicServer> <BEA-000386> <Server subsystem failed. Reason: java.lang.NullPointerException
java.lang.NullPointerException
at weblogic.t3.srvr.SubsystemRequest.action(SubsystemRequest.java:79)
at weblogic.t3.srvr.SubsystemRequest.start(SubsystemRequest.java:51)
at weblogic.t3.srvr.ServerServicesManager.startService(ServerServicesManager.java:372)
at weblogic.t3.srvr.ServerServicesManager.startInStandbyState(ServerServicesManager.java:125)
at weblogic.t3.srvr.T3Srvr.initializeStandby(T3Srvr.java:630)
at weblogic.t3.srvr.T3Srvr.startup(T3Srvr.java:402)
at weblogic.t3.srvr.T3Srvr.run(T3Srvr.java:361)
at weblogic.Server.main(Server.java:67)
To work around this problem, the following steps should be performed:
1. Edit the all-in-one-rule.xml file under the directory MercuryDiagnostics/JavaAgent/TransactionVisionAgent/config/sensor/instrumentDef and remove the following section:
<ClassName name="weblogic.management.configuration.DomainMBeanImpl">
<Action name="ForwardMethodCall">
<SensorFlagName name="com.bristol.tvision.sensor.CheckSensor.servletSensorInstalled"/>
<SensorClass name="com.bristol.tvision.sensor.servlet.ServletSensor"/>
<SensorMethodPrefix name="DomainMBeanImpl_"/>
<MethodSignature name="void setName(java.lang.String)"/>
</Action>
</ClassName>
2.Edit the TV.properties file under MercuryDiagnostics/JavaAgent/DiagnosticsAgent/etc directory to add the following at the end of the tvProperties line.
; com.hp.j2ee.domain.name=your_weblogic_domain_name
Note that the WebLogic domain name is usually the directory containing your server.
3. Restart your WebLogic server.
Syntax
nanny -s <hostname> -p <port> <cmd>
Options
- FIXED IN 8.01: The Aggregated Topology page may incorrectly show a blank table. Click the Generate button to populate this page correctly.
- FIXED IN 8.01: Tooltips may flicker in the Flowmap, Aggregated Topology, and Instance Topology. This occurs if the tooltip is too close to the bottom of the screen. If you experience this, you can either slowly move the cursor off the tooltip or move the line or icon higher up in the topology so that the tooltip is not touching the bottom of the screen. This is a known bug with Sun's JRE. See Sun's Java bug report for more information: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6308296.
- FIXED IN 8.01: System Health monitors TransactionVision components such as the Analyzer, SonicMQ Domain Manager, SonicMQ Message Broker, and UI/Job Server. On UNIX, System Health does not correctly monitor the SonicMQ Domain Manager. This results in a bad status being shown for SonicMQ Domain Manager process in the System Health Dashboard.
To fix this problem, perform the following steps:
- Stop the nanny on the Analyzer host using the command:
<TVISION_HOME>/bin/run_topaz stop.- Edit the file: <TVISION_HOME>/conf/supervisor/nanny/SonicMQDomainManager.nanny and replace the following line:
procName=SonicMQDomainManager
with the line:
procName=SonicMQDM- Start the nanny on the Analyzer host using the command: <TVISION_HOME>/bin/run_topaz start.
To verify the fix, click the Analyzer node in the System Health Dashboard and observe the green status of the SonicMQ Domain Manager process.
- FIXED IN 8.01: Appendix B of the TransactionVision Deployment Guide is incomplete; the following configuration files are supported at this release and should be included in the appendix:
- License.properties: The <TVISION_HOME>/config/license/License.properties file specifies the TransactionVision license code supplied by HP.
- Performance.properties: The <TVISION_HOME>/config/services/Performance.properties file contains the following settings for performance logging:
- performance. Specifies whether performance logging is on or off. The default is off.
- count_interval. Specifies the number of events after which performance data is logged. The default is 500.
- detail. Specifies whether to generate detailed performance and statistics logs. The default is off.
- start_at. Defines how many events have to get processed before the performance logging will start. The default is 500.
- get_queuedepth. Specifies whether to retrieve the current event queue depth for each log interval. The default is off.
- The flow map will not load for the Aggregated Topology, Instance Topology or Dashboard if you do not have the correct Java JRE installed - you must have JRE 1.5.x or above. Instead of the flow map, a red X icon displays. To fix this, you can do one of the following:
- FIXED IN 8.01: The Search function has been removed from the Active Filters dialog for both the Aggregated Topology and Instance Topology pages, since it was not functioning properly.
- Release 7.50 Java Agents do not collect the data required by the CMDB Population and CMDB Update jobs when used with the 8.00 Analyzer and UI/Job Server. In this case, the UI /Job Server logs contain messages such as:
TransactionVision Error (SystemModelMissingCIAttributes: Data required to populate CI not present in system model.
- If your TransactionVision deployment environment consists of 7.5 Java Agents only, we recommend that the CMDB Population and CMDB Update jobs be stopped.
- If your TransactionVision deployment environment consists of both release 7.50 Java Agents and 8.00 Java Agents reporting to the 8.00 Analyzer and UI/Job Server, the CMDB Population and CMDB Update jobs can continue to run and the above error message can be ignored. In this case, only the infrastructure entities corresponding to the 8.00 Agents are reported to the uCMDB.
- After reinstalling the TransactionVision UI/Job Server on Solaris, a tomcat process from the previous installation is left running, causing the TransactionVision menus in Business Availability Center to be disabled. To resolve the problem, use "ps | grep Tomcat" to find the problem process, then kill it.
- When upgrading a TransactionVision Analyzer from a previous version, the migration automatically updates the Analyzer configuration files to use the new standard port for the Analyzer, 21100. However, the Analyzer definition in the Business Availability Center is not updated accordingly and must be done manually. Access the Admin > TransactionVision > Administration > Analyzers > Update Analyzer page, and for any affected analyzer update the port to 21100.
- Corrections and additional instructions for importing the Trade Demo Sample Events can be found in the <TVISION_HOME>\samples\readme.txt file which is included with the TransactionVision Analyzer installation package. Please refer to this file prior to importing the sample events.
- The Delete button is mistakenly enabled in the Communication Link Templates page. If a user is editing one of the default communication link templates they can unintentionally delete the edited communication link template. Workaround: Rerun TVisionSetupInfo.bat/.sh located in the <TVISION_HOME>\bin directory to re-create the deleted communication link template.
- TransactionVision data is always saved to the default profile database in HP Business Availability Center.
- If connection to the TransactionVision database is lost, the following message is displayed:
"TransactionVision Error: The connection to the TransactionVision database has been lost. Please ensure that the database is accessible, and then re-login to BAC again."
In some cases, making sure the database is accessible and re-logging into Business Availability Center does not resolve the problem. In these cases, the problem can be resolved by restarting the TransactionVision UI/Job Server using nanny restartService tv_as. For more information about the nanny utility, see Appendix A in the TransactionVision Deployment Guide.
Business Process Insight Server
Previously, when viewing a TransactionVision-driven process from within the Business Process Insight application inside Business Availability Center, if you selected the Health Page, clicked to drill down to a list of process instances, selected an individual instance and selected to view the data for the instance, a link to TransactionVision reports appeared. However, if you attempted to get to the report by clicking on the link, it failed with HTTP 404 errors, or nothing opened.
This issue has been fixed by changing the mechanism of how the Business Process Insight drill down pages link to the TransactionVision reporting pages.
If you start Business Process Insight while TransactionVision's JMS Event Queue services for sending events to Business Process Insight are unavailable, or if the JMS Event Queue services become unavailable once Business Process Insight has started, the OVBPIJmsEventHandler Service stops. To solve this problem, you need to configure the Windows Service for the OVBPIJmsEventHandler Service under Control Panel > Administrative Tools > Services. You should configure the service in the same way as is described for adapter Windows Services in the Integration Training Guide - Business Events, by selecting the Recovery option and setting each of the First, Second and Subsequent Failures to the value of Restart the Service.
If you deploy a Business Process using the Business Process Insight Modeler and you see unexpected images displayed next to the deployed Business Process definition name, you are likely to be experiencing a problem with the version of the Java Virtual Machine that you are using. The problem is specifically with JRE version 1.5_16.
To resolve the problem, use a version of the JRE that is earlier than version 1.5_16, or a later version, if there is one available.
If you intend to modify the Servlet Engine (Tomcat) configuration files, you need to be aware that changes made directly to the Servlet Engine files are overwritten by Business Process Insight configuration files. Refer to the upgrading chapter of the Business Process Insight Deployment Guide for more details.
If you see an error indicating that one or more Business Process Insight Infrastructure settings is missing, make sure that you have specified a value for the Microsoft SQL Server database name:
Instance Database Settings: Database Name.
You need to provide a value for this setting even if you are configuring the settings for an Oracle database. It can be any string, but you cannot leave the field blank; if you do, you receive an error when accessing the Business Process Insight Application pages.
Very occasionally, following a system restart, the Servlet Engine appears not to have started, both within the Business Process Insight Administration Console and as a Windows Service. The Servlet Engine service is actually in the processes of starting, but is taking a long time, possibly due to other applications causing delay; for example, the database can take a long time to start when it is very full.
If you think that you are experiencing this problem:
If you receive an error message indicating that the Business Process Insight Server is unable to contact the RMI Registry on port 44,000, this is likely to be due to a port number clash.
The Business Process Insight RMI Registry uses port 44,000, and if another application or service is temporarily using this port, then the RMI Registry cannot start. To resolve this problem, you can:
Select the Port Numbers tab from the Business Process Insight Server Administration Console to change the RMI Registry port number from 44,000. If you change the value for the port number, you need to make sure that any other configuration affected by this change is also modified, for example, the Business Process Insight Infrastructure settings.
The following scenario results in a Process definition that cannot be progressed:
One of the Process definitions is now linked to the newly created Data definition and no changes are made to the progression rules. In this case, the Business Process Insight Modeler should detect that the progression rules are referencing a different Data definition, but it does not and as a result, no process instances are created.
To resolve this problem, use the Export and Import options within the Business Process Insight Modeler to make a copy of a Process definition. Use the option to rename the definitions (not overwrite them).
See "Modeler Management Tasks" in the Using Business Process Insight guide for details of making copies of your Process definitions.
When entering property values for subscriptions within the Business Process Insight Modeler, be aware that the Oracle Server does not distinguish between a zero-length string, and a string with a value that has not been assigned (null string). This means that if you want to test a zero length string within a subscription, you are advised to use a single blank space (" "), and not an empty string ("").
The HP Business Process Insight Accelerator for SAP Applications uses the openadaptor framework. If you install the Accelerator on a different machine to the one where the Business Process Insight Server is installed, the Accelerator installs only the subset of the openadaptor files that it needs to operate.
If you want to extend the capabilities of the Accelerator and use a feature of the openadaptor framework that are not provided as part of the Accelerator installation, you need to copy the complete openadaptor framework from the Business Process Insight machine to the machine where the Accelerator is installed.
The openadaptor framework is located at the following location on the Business Process Insight Server machine:
bpi-install-dir\nonOV\openadaptor
If you see an error message similar to the following when importing a Process definition into the Business Process Insight Modeler, it is possible that you are attempting to import a Process definition with a name that contains characters that are locale-specific:
The file proc-defn-name.zip is not a valid import file. Please see the log file for more information.
where proc-defn-name is the filename, based on the Process definition name.
There are restrictions in the use of locale-specific character sets, whereby a .zip file name can contain characters only from the character set of the system locale where the Business Process Insight Modeler is running. The content of the file can contain any characters, but the file name is limited by the character set of the Windows system that is reading or writing the file. In the case of the zip archive files for exporting and importing processes, the zip archive file name is based on the Process definition name. Therefore, if you want to export a Process definition, you have to consider where you will be importing the file. If you plan to move the file across locales, you are advised to use only ASCII characters for the Process definition name to ensure the file can be read on any configured locale.
This problem can also occur if you change the locale of the Windows system where you have created the Process definitions.
If you modify the locale for a system where the Business Process Insight Server is installed, you might find that you cannot start the Business Process Insight Administration Console when you reboot your system. The Administration Console also returns an error indicating that it cannot find the main class and that the program will exit.
This can happen in cases where Business Process Insight is installed into a directory, which is named using locale-specific characters. If the locale for the system is then changed, it is not always possible for the directory and file names to be readable using the character set of the new locale.
If you have changed the locale of the system where Business Process Insight is installed and you are experiencing this problem, you need to:
If you attempt to run the Business Process Insight Installation procedure using a Turkish locale, the installer does not present the installation screens correctly and, as a result, you cannot answer the installation questions. This is a problem related to the InstallAnywhere software and details of the problem can be found at the following URL:
http://support.installshield.com/kb/view.asp?articleid=Q110959
In order to complete the installation successfully, you need to temporarily set the locale to a United States locale. When you have completed the installation, you can reset the locale to Turkish.
If you enter non-ASCII characters for the Oracle database data directory during the Business Process Insight installation, the installation fails with a Database Setup error. The error indicates that Oracle is unable to create the database instance or a database table.
If the Business Process Insight installer reports a Database Setup Error, and you have used non-ASCII characters for the name of the database data directory, you need to reinstall (overinstall) Business Process Insight into the same location, this time using a database data directory name that contains only ASCII characters.
If you are using Microsoft SQL Server 2005 and you are aware that you are having problems with database checkpoint I/O requests, you are experiencing a known problem with SQL Server 2005. The problem has been fixed by Microsoft and has the following Microsoft defect number: 50000708.
Microsoft provided a solution for this problem in their Service Pack 2 for SQL Server 2005. You need to make sure that you have installed Service Pack 2. Microsoft also recommends that you use the -k3 startup parameter to limit the I/O requests generated by the checkpoint process. Refer the defect text provided by Microsoft, or the Microsoft documentation for details of using startup parameters.
Many errors encountered when the InstallAnywhere uninstaller is running, for example, a missing JVM, causes the uninstaller to stop responding, or hang.
If you experience a problem where the Business Process Insight uninstaller stops responding, you need to complete the following tasks according to what you have installed:
- Stop the uninstaller process, which is still running on your system. For example, on a Windows system use the Task Manager.
- If you are removing a Business Process Insight Server, delete the Business Process Insight files under the Business Process Insight installation directory.
- Modify the Zerog Global Registry (an InstallAnywhere file):
If the registry file contains only entries for the product that you are attempting to uninstall, you can rename the file. However, if there are multiple product installations, or multiple instances of one product installation, you need to edit the file as described in the following step.
If the registry file contains entries in addition to those for Business Process Insight, edit the file to remove the following entries:
- The product elements for Business Process Insight. Make sure the element references the directory where the files are located that you want to uninstall.
- The components elements for Business Process Insight. Make sure the component element references the directory where the files are located that you want to uninstall.
- Restrictions when using Autopass.
The following are restrictions that you need to be aware of when using Autopass with Business Process Insight:
- Autopass is a separate product and has its own requirements with respect to local language support. It is possible that versions of Autopass are available in languages that differ from the languages supported by Business Process Insight. In this case, you might see a mixture of US English and local language for each of the products.
- Autopass does not support the use of non-ASCII characters in entry fields; for example, you cannot enter customer details using non-ASCII characters. Refer to the Autopass product information for more details.
Business Availability Center for SOA
- Only scripts that are recorded in HP Virtual User Generator version 9.10 or later are supported for SOA breakdown. To use a script recorded with HP Virtual User Generator version 9.00, detach the script from a profile (if attached), open and save the script in HP Virtual User Generator version 9.10 or later, and reattach the script to the profile. Scripts recorded with HP Virtual User Generator version 8.10 FP3 and earlier are not supported and you must record the script again using HP Virtual User Generator version 9.10 or later.
Business Availability Center for SAP
Business Availability Center for Siebel
- The Siebel Cross Performance report requires Internet Explorer 7.0 to display correctly. If you want to view this report when using Business Availability Center for Siebel, use Internet Explorer 7.0.
- If you attach the Siebel KPI to a CI in a non-Siebel view, the propagation propagates the Siebel KPI with the wrong rule, and the KPI displays a No data status. Workaround: Edit the Siebel KPI and attach the required rule (for example the Worst Child rule).
HP SiteScope
Adapters and Views
- In Source Manager, if you perform an adapter synchronization that does not finish before the adapter's scheduled synchronization, the adapter will no longer synchronize automatically. To solve this issue, disable the problematic adapter and enable it again.
- When editing the SiteScope source adapter in Source Manager, if you change the Include measurements value to None, the previous SiteScope monitor configurations are deleted from the CMDB. This may affect, for example, SLAs in Service Level Management that are based on SiteScope measurements.
- When synchronizing a SiteScope source adapter in Source Manager, if the adapter includes a large number of objects, then adapter performance might be slow, taking several minutes.
Platform
- The recommended setting for Oracle parameter NLS_LENGTH_SEMANTICS is the default setting of BYTE; setting this value to CHAR may affect database performance.
- System health monitors will not function if you change the JNDI port. The JNDI port must not be changed from its default value.
- Changes to infrastructure settings are not included in the Audit Log.
- You cannot create a schema that starts with a digit.
- Downtime events are rescheduled incorrectly after a daylight savings (DST) time change.
Integrations
- More than one HP OM (formerly HP OVO) monitor can be created on the same port using the Copy function, resulting in duplicate OM events. In such cases, delete the copied monitor.
- The maximum size of an expression in the EMS Integration configuration file is 4000 characters.
- An EMS monitor may not find its configuration file when upgrading the SiteScope version. In such cases, manually copy the contents of the configuration file to the EMS monitor configuration in the user interface.
- If you ever used a version of Business Availability Center prior to 5.x, subsequently performed multiple upgrades to higher versions, but never used the Diagnostics integration, you may have a missing property in the Sessions table of the Business Availability Center database that prevents the integration between HP Business Availability 8.00 and Diagnostics 8.00.
Workaround (not relevant for Software-as-a-Service customers):
SELECT * FROM SESSIONS
WHERE SESSION_NAME='J2EE_For_Default Client'
SELECT * FROM SESSIONSWHERE SESSION_NAME like 'Unassigned%' AND SESSION_TYPE_ID=5
UPDATE SESSIONS
SET SESSION_NAME='J2EE_For_Default Client',
SESSION_ACTION=1,
SESSION_TYPE_ID=10,
SESSION_AGR_POLICY_ID=12,
SESSION_PRIM_TIME_DIFF=120
WHERE SESSION_ID=<<the new id>>
Universal CMDB
General, Installation, and Upgrade
- When upgrading from previous version, you must export configuration changes made to out-of-the-box content (TQLs, views, CI types, Jython Scripts, discovery jobs, and so forth) of UCMDB packages and reapply the changers after the upgrade is complete. This is because the upgrade process overrides custom changes made to out-of-the-box content. For details, see the HP Universal CMDB Deployment Guide.
- After upgrading to 8.00, the CIs of types Database and J2EE (termed WebSphere in the documentation) Server are removed from all instance views, SLAs, and alerts that include the CIs of those types. For details, see the Impact Modeling section in the "Component and Data Upgrade and Limitations" chapter in the HP Business Availability Center Deployment Guide.
- After upgrading to 8.00, check the log file: <HP Business Availability Center root directory>\mam_lib\_reports\customer_[id]\UpgradeInfoReport.txt to view all the CIs that are deleted.
- If an upgraded instance view contains a host (without any monitor), and the host is connected to a software element (that is connected to a monitor), after the upgrade the Host CI appears in Show hidden CIs list in Modeling Studio. This is relevant if the software element is not of the type Database or J2ee server. Workaround: Remove Host CIs from Show hidden CIs list.
- HP UCMDB and Federated Adapters upgrade notes:
- CMDB RMI adapter: The CmdbRmiAdapter.zip package is overwritten during upgrade. If you modified the CmdbRmiAdapter package follow the below instructions:
- Copy the existing <Business Availability Center root directory>\mam_lib\factory_packages\CmdbRmiAdapter.zip file to a temporary directory before starting the upgrade process.
- After the upgrade process, edit the files in the new CmdbRmiAdapter.zip package to match the changes you had made to the old adapter package. You must make the changes in the new CmdbRmiAdapter.zip and not overwrite with your old package.
- Proprietary adapter configured in Generic DB Adapter default folder: If you created proprietary adapter in the default Generic DB Adapter folder, you must perform the following procedures to make sure you do not lose your existing Generic DB Adapter configuration information during the upgrade process.
Discovery and Dependency Mapping
Modeling and Application
- When working in the Model Editor in Modeling Studio, the Find feature only works for CIs that are visible in the expanded model/view. Workaround: Make sure the model is expanded before using the Find feature.
- The Find feature in the CI Selector (in IT Universe Manager and Modeling Studio) does not work for CIs that are not visible in the collapsed view tree until you expand the relevant branch of the tree. Workaround: Once you expand the relevant branch, the Find feature works, even if the branch is collapsed again. If you navigate away from the view and return to it, you need to expand the branch again for the Find feature to work.
- In Modeling Studio, when the Location Based Monitors perspective is selected as the perspective for a view, the Apply Over Model Hierarchy option is selected by default, which results in the inclusion of business process step CIs in the view. Clear the Apply Over Model Hierarchy check box to remove these CIs from the view.
- Changing an integration TQL can result in data loss. To avoid data loss:
- Before making changes to a TQL, run diff replication. This step is needed to avoid losing removed data when running a full replication (the next step).
- After making changes to a TQL, run full replication. This step is needed to retrieve all the data for a second time, to differentiate between new data (that must be retrieved) and existing data (that does not need to be retrieved).
- When changes are made in View Manager and these changes affect the results of a TQL, federated CIs in the view are not updated. This is because federated TQLs are calculated ad-hoc only and are not updated when a view is recalculated. To update the federated CIs, in IT Universe Manager, click the Rebuild View button in the CI Selector pane. (Note that the recalculation may take a long time.)
- UCMDB 8.00 uses an applet that is run by a JRE. The supported JRE versions are 6u4 and later (6u10 recommended, this version is included on the UCMDB server itself for local network download).
- To optimize the user experience in UCMDB 8.00, choose "Yes" when asked about changing Java configuration. This will enable the UCMDB application to use an additional 192MB of memory, which will help it perform much better. After accepting the change, you will need to reopen the application for the change to take effect.
- After finding the relevant CI in the CMDB Search, the context menu on the left-hand side is synchronized only after selection.
- If you run get result count over a TQL with an invisible multiple data store element, the count for the element will not include the CMDB objects satisfying the query for that element. If you have a view based on such a TQL, you will not be able to run discovery based on this view (via IT Universe Manager > View Browser > Show Discovery Status and Changes of View) if the view's result contains data only from the CMDB in the multiple data store node, and no objects in the view's other nodes.
- Exporting data from UI components such as tables and trees is not supported for Asian languages.
Reports
- FIXED IN 8.01: Running Asset reports on template-based views is not supported.
- The results of scheduled Gold Master and Topology reports (known as System reports in 6.x) that were generated in HP Universal CMDB 6.x cannot be viewed in HP Universal CMDB 8.00.
- It is not possible to search inside reports exported to PDF (since the reports are saved as images).
- It is not possible to export chart reports to Excel and CSV.
Documentation Library
- If you experience a JavaScript error when opening the HP Business Availability Center Documentation Library online help system, disable the "Show Exception Dialog Box" in the Java Console and open the help again.
- When viewing the HP Business Availability Center Documentation Library online Help system in Internet Explorer 8, numbered or alphabetical lists may appear corrupted.
Multilingual Support
- If you use a CJK language in your browser, you must ensure that the Gateway server machine running HP Business Availability Center has East Asian languages installed. On the machine on which the HP Business Availability Center Gateway server is installed, you must select Control Panel > Regional & Language Options > Languages > Install files for East Asian languages.
- In CJK languages, when browser locale and Windows system locale are not set to the same language, navigation pane in Business Availability Center documentation library displays gibberish in Search and Favorites tabs. Workaround: Ensure that browser locale and Windows system locale are both set to the same language.
- The new multilingual user interface (MLU) feature enables you to view HP Business Availability Center in different languages by configuring your browser language. For considerations and limitations when working in non-English locales, see Working in Non-English Locales in the Reference Information guide of the HP Business Availability Center Documentation Library.
- In Dashboard Top View, Non-English characters of a specific language will be displayed only if the operating system is using the same language. For example, to see Japanese characters you should use Japanese OS and locale.
HP Business Availability Center Support Matrixes
This section describes updated matrixes for various environments and components supported by HP Business Availability Center.
√=supported, X=not supported
Supported Environments
Note: System health is based on SiteScope - please see SiteScope supported operating systems.
Database Support Matrixes
For complete details on database support, see the HP Business Availability Center Database Guide.
Hardware Requirements
Deployment Size
Number of Processors
Physical Memory
Standard
1 dual core
Minimum: 2G RAM
Recommended: 4G RAM
Large
2 dual core or 1 quad core
Minimum: 4G RAM
Recommended: 8G RAM
For information on sizing criteria, see the HP Business Availability Center Database Guide.
Software Requirements - Oracle Server
The following table lists updates to the Oracle servers supported for working with Business Availability Center.
Oracle 9.2.0.8 Enterprise Edition, 64 BITExamples of Tested Deployments - Oracle Server
Software Requirements - Microsoft SQL Server
The following table lists updates to the Microsoft SQL Servers supported for working with Business Availability Center:
Microsoft SQL Server 2000 Enterprise Edition, 32 BIT, Service Pack 3
- Only supported service packs should be installed, with latest patches.
- Consult the Microsoft documentation for supported platforms.
- To check which version of MDAC is installed on your machine, download and run the component checker tool from: http://www.microsoft.com/downloads/details.aspx?FamilyID=8f0a8df6-4a21-4b43-bf53-14332ef092c9&displaylang=en.
Examples of Tested Deployments - Microsoft SQL Server
The following table details those deployment environments that were tested by HP.
Business Availability Center Component Support Matrixes
SiteScope Support Matrix
SiteScope/Business Availability Center Compatibility Matrix
All the supported versions support both data logging and System Availability Management Administration.
Data logging is the process of logging data collected by SiteScope to Business Availability Center for the purposes of real-time status, reporting, Service Level Management, and so forth.
System Availability Management Administration refers to configuring SiteScope in System Availability Management Administration.
SiteScope System Health Support Matrix
Note: System Health should be installed only from a major version of SiteScope, and not from a SiteScope service pack.
Real User Monitor Support Matrix
Note: Real User Monitor 7.01 (probe and engine) works only with Business Availability Center 7.01.
Business Process Monitor Support Matrix
For each BPM version, the latest supported VuGen version is mentioned. All previous versions of VuGen are supported as well.
Business Process Monitor/QuickTest Professional (QTP) Support Matrix
DDM Discovery Probe Support Matrix
Note: Discovery Probe 7.51 works only with Business Availability Center 7.51.Discovery Probe 7.01 works only with Business Availability Center 7.01.
Business Availability Center/Diagnostics Integration Matrix
Business Availability Center/Service Manager & ServiceCenter Integration Matrix
Note: For integration with Service Manager to function properly, additional setup is required in both Business Availability Center and Service Manager. See the Business Availability Center documentation for details.
TransactionVision Agent Compatibility Matrix
Notes:
- All TransactionVision 7.5 agents are compatible with all versions of TransactionVision Analyzer 8.0x.
- Business Availability Center and TransactionVision Analyzer versions must be aligned when possible. For example, Business Availability Center 8.05 must be run with the TransactionVision 8.05 Analyzer and TransactionVision 8.05 UI/Job Server. When version alignment is not possible, Business Availability Center must be used with the latest release of the TransactionVision Analyzer and UI/Job Server. For example, Business Availability Center 8.07 must be run with the TransactionVision 8.05 Analyzer and TransactionVision 8.05 UI/Job Server.
- Many TransactionVision Agents are backward compatible with newer Analyzer versions. For details on backward compatibility, see the TransactionVision Deployment Guide.
Business Process Insight − Business Availability Center Support Matrix
Notes:
- BPI is only available on the Windows platform (that is, the BPI Server only runs under the Windows OS)
- BPI requires either a Microsoft SQL 2005 or Oracle database (it will not work with a Microsoft SQL 2000 database)
- The minor versions of Business Process Insight 8.0x must match with the minor versions of Business Availability Center 8.0x to be compatible. If the versions do not match, see the workaround in the Version Content and Limitations section in this Readme for making Business Availability Center and Business Process Insight compatible.
Business Availability Center/HP Operations Manager Integration Matrix
Business Availability Center for SAP Component Support Matrix
Business Availability Center for Siebel Component Support Matrix
HP Software Support
This web site provides contact information and details about the products, services, and support that HP Software offers. For more information, visit the HP Support web site at: HP Software Support Online.
HP Software support provides customer self-solve capabilities. It provides a fast and efficient way to access interactive technical support tools needed to manage your business. As a valued support customer, you can benefit by being able to:
- Search for knowledge documents of interest
- Submit and track progress on support cases
- Submit enhancement requests online
- Download software patches
- Manage a support contract
- Look up HP support contacts
- Review information about available services
- Enter discussions with other software customers
- Research and register for software training
To access the Self-solve knowledge base, visit the Self-solve knowledge search home page.
Note: Most of the support areas require that you register as an HP Passport user and sign in. Many also require an active support contract. To find more information about support access levels, go to: Access levels.
To register for an HP Passport ID, go to: HP Passport Registration.
Legal Notices
© Copyright 2005 - 2011 Hewlett-Packard Development Company, L.P.
Confidential computer software. Valid license from HP required for possession, use or copying. Consistent with FAR 12.211 and 12.212, Commercial Computer Software, Computer Software Documentation, and Technical Data for Commercial Items are licensed to the U.S. Government under vendor's standard commercial license.
The only warranties for HP products and services are set forth in the express warranty statements accompanying such products and services. Nothing herein should be construed as constituting an additional warranty. HP shall not be liable for technical or editorial errors or omissions contained herein.
The information contained herein is subject to change without notice.
For information about third-party license agreements, see the Third_Party_Agreements directory on the product installation media.
Trademark Notices
Adobe® and Acrobat® are trademarks of Adobe Systems Incorporated.
Intel®, Pentium®, and Intel® Xeon™ are trademarks of Intel Corporation in the U.S. and other countries.
Java is a registered trademark of Oracle and/or its affiliates.
Microsoft®, Windows®, Windows NT®, and Windows® XP are U.S registered trademarks of Microsoft Corporation.
Oracle® is a registered US trademark of Oracle Corporation, Redwood City, California.
Unix® is a registered trademark of The Open Group.
If you have any comments or suggestions regarding this document, please send them by e-mail to SW-Doc@hp.com.