IBM Tivoli Monitoring > Version 6.3 Fix Pack 2 > Installation Guides > Installation Guide > Tivoli Data Warehouse solutions: common procedures

IBM Tivoli Monitoring, Version 6.3 Fix Pack 2


Configure the Summarization and Pruning Agent (JDBC connection)

Use this procedure to configure the Summarization and Pruning Agent to connect to a Tivoli Data Warehouse database created on any of the supported database platforms and operating systems.

The JDBC driver JAR files for your database platform must be located on the computer where you installed the Summarization and Pruning Agent. Use a Type 4 JDBC driver. Do not use the Type 2 driver.


Table 1 shows where to obtain the driver files for each database platform.


Where to obtain the JDBC driver files for the Summarization and Pruning Agent

Database platform JDBC driver files
IBM DB2 for Linux, Unix and Windows

Use the DB2 for Linux, UNIX, and Windows JDBC Universal Driver (Type 4 driver). The DB2 for Linux, UNIX, and Windows driver files are located with the Tivoli Data Warehouse server installation. The Type 4 driver file names and locations are as follows:

db2installdir/java/db2jcc.jar
db2installdir/java/db2jcc_license_cu.jar where db2installdir is the directory where DB2 for Linux, UNIX, and Windows was installed. The default DB2 for Linux, UNIX, and Windows Version 9 installation directory is as follows:

  • On Windows: C:\Program Files\IBM\SQLLIB

  • On AIX: /usr/opt/db2_09_01

  • On Linux and Solaris: /opt/IBM/db2/V9.1

Microsoft SQL Server

Use the latest Microsoft SQL Server JDBC driver, which supports SQL Server 2008, 2005, and 2000, to connect to a Tivoli Data Warehouse on either SQL Server 2000 or SQL Server 2005. (The SQL Server 2005 JDBC Driver works with a Tivoli Data Warehouse on SQL Server 2000.) Go to the Microsoft Web page at:
http://www.microsoft.com and search for JDBC driver.

Download and install the driver to the computer where you installed the Summarization and Pruning Agent. Follow the instructions on the Microsoft download page for installing the driver. The SQL Server JAR file name and location after installation is as follows:<mssqlinstalldir>/sqljdbc_1.1/enu/sqljdbc4.jar.

Oracle

Obtain the Oracle JDBC Type 4 driver from the following Web site:
http://www.oracle.com/technology/software/tech/java/sqlj_jdbc/index.html.

The Oracle JDBC driver JAR file name and location after installation is as follows:oracleinstalldir/jdbc/lib/ojdbc14.jar

The ojdbc14.jar file supports JRE 1.5 or higher, the required Java Runtime Environment for IBM Tivoli Monitoring.


To configure the Summarization and Pruning Agent.

  1. Log on to the computer where the Summarization and Pruning Agent is installed and begin the configuration:

    1. Open the Manage Tivoli Enterprise Monitoring Services window:

      • Click Start → Programs → IBM Tivoli MonitoringManage Tivoli Monitoring Services.

      • Run...

        ...where install_dir is the installation directory for IBM Tivoli Monitoring. The default installation directory is /opt/IBM/ITM.

    2. Right-click Summarization and Pruning Agent.

    3. Click Configure Using Defaults. Click Configure. If you are reconfiguring, click Reconfigure.

  2. Review the settings for the connection between the Summarization and Pruning Agent and the hub Tivoli Enterprise Monitoring server. These settings were specified when the Summarization and Pruning Agent was installed.

    • Perform the following steps:

      1. On the Warehouse Summarization and Pruning Agent: Agent Advanced Configuration window, verify the communications protocol of the hub monitoring server in the Protocol drop down list. Click OK.

      2. On the next window, verify the host name and port number of the hub monitoring server. Click OK.

      For information about the different protocols available to the hub monitoring server on Windows, and associated default values, see Windows: Installing the hub monitoring server.

    • Verify the following information on the TEMS Connection tab:

      • The hostname of the hub monitoring server in the TEMS Hostname field. (If the field is not active, clear the No TEMS check box.)

      • The communications protocol that the hub monitoring server uses in the Protocol drop down list.

        • If you select IP.UDP, IP6.UDP, IP.PIPE, IP6.PIPE, IP.SPIPE, or IP6.SPIPE, enter the port number of the monitoring server in the Port Number field.

        • If you select SNA, enter information in the Net Name, LU Name, and LOG Mode fields.

        For information about the different protocols available to the hub monitoring server on Linux or UNIX, and associated default values, see Configure the hub monitoring server.

  3. When you are finished verifying or entering information about the hub monitoring server:

    • Click Yes on the message asking if you want to configure the Summarization and Pruning Agent.

    • Click the Agent Parameters tab.

    A multi-tabbed configuration window is displayed with the Sources tab at the front.
    Figure 1 shows the configuration window for a Summarization and Pruning Agent on Windows (values displayed are for a DB2 for Linux, UNIX, and Windows warehouse database). The configuration window for a Summarization and Pruning Agent on Linux or UNIX is similar.

    Figure 1. Sources pane of Configure Warehouse Summarization and Pruning Agent window

  4. Add the names and directory locations of the JDBC driver JAR files to the JDBC Drivers list box:

    1. Click Add to display the file browser window. Navigate to the location of the driver files on this computer and select the Type 4 driver files for your database platform. See Table 1 for the names and default locations of the driver files to add.

    2. Click OK to close the browser window and add the JDBC driver files to the list.

    If you need to delete an entry from the list, select the entry and click Remove.

  5. The default values for the database platform you selected in the Database Type pane are displayed in the other text fields on the Sources pane. Change the default value displayed in the JDBC URL field if it is not correct. The following table lists the default Tivoli Data Warehouse URLs for the different database platforms:


    Tivoli Data Warehouse URLs

    Database platform Warehouse URL
    IBM DB2 for Linux, UNIX, and Windows jdbc:db2://localhost:60000/WAREHOUS
    Oracle jdbc:oracle:thin:@localhost:1521:WAREHOUS
    Microsoft SQL Server 2000 or SQL Server 2005 jdbc:sqlserver://localhost:1433;databaseName=WAREHOUS

    • If the Tivoli Data Warehouse is installed on a remote computer, specify the host name of the remote computer instead of localhost.

    • Change the port number if it is different.

    • If the name of the Tivoli Data Warehouse database is not WAREHOUS, replace WAREHOUS with the actual name.

  6. Verify the JDBC driver name.

    The following table lists the JDBC Type 4 driver names for each database platform:


    JDBC driver names

    Database platform JDBC driver name
    IBM DB2 for Linux, UNIX, and Windows com.ibm.db2.jcc.DB2Driver
    Oracle oracle.jdbc.driver.OracleDriver
    Microsoft SQL Server

    com.microsoft.sqlserver.jdbc.SQLServerDriver

    This is the name of the 2005 SQL Driver. Do not use the SQL Server 2000 JDBC driver, even if the Tivoli Data Warehouse was created in Microsoft SQL 2000. (The name of the 2000 SQL driver was com.microsoft.jdbc.sqlserver.SQLServerDriver. Note the reversal of the string jdbc.sqlserver.)

  7. If necessary, change the entries in the Warehouse user and Warehouse password fields to match the user name and password that were created for the Tivoli Data Warehouse. The default user name is itmuser and the default password is itmpswd1.

  8. In the TEPS Server Host and TEPS Server Port fields, enter the host name of the computer where the Tivoli Enterprise Portal Server is installed and the port number that it uses to communicate with the Summarization and Pruning Agent.

    The default Tivoli Enterprise Portal Server interface port of 15001 is also used after the Summarization and Pruning Agent's initial connection to the portal server over port 1920. Any firewalls between the two need to allow communications on either 15001 or whichever port is defined for any new Tivoli Enterprise Portal Server interface used per the instructions in Define a Tivoli Enterprise Portal Server interface on Windows.

  9. Click Test connection to ensure you can communicate with the Tivoli Data Warehouse database.

    If Microsoft SQL Server is the database platform, an error message might be displayed after clicking Test connection even if you have changed the JDBC driver to sqljdbc4.jar. The problem is caused by the caching of the old sqljdbc.jar file. If you save the settings with the new sqljdbc4.jar file and retry Test connection, the new jar file is used and the error message will not appear.

  10. Select the Reporting check box to configure the Summarization and Pruning agent to create and maintain the dimension tables required by Tivoli Common Reporting and IBM Tivoli Monitoring. Enter values for the following fields and then click OK:


    Reporting integration settings

    Field Description
    Enable reporting integration Indicates whether reporting integration is enabled. If enabled, the time dimension tables are created if they do not already exist. Time dimension tables will also be maintained when this variable is enabled. The default value is No.

    If this variable is set to Yes, and the resource dimension tables do not exist, but the KSY_TRAM_USER and KSY_TRAM_PASSWORD are not set, then a warning message in the trace file and in the Summarization and Pruning agent workspace is returned indicating that the resource dimension tables were not created.

    Reporting user The reporting user that creates the time dimension tables. If the database is not DB2 LUW or DB2 z/OS, the reporting user is set by default to IBM_TRAM. However, as IBM_TRAM does not follow the user naming rules for DB2 LUW (name should not start with IBM) or DB2 z/OS (name should not contain an underscore) a different user name must be used. The user should have the necessary privileges to create an IBM_TRAM schema. The KSY_TRAM_USER value only needs to be specified if KSY_TRAM_ENABLE is set to Yes and the time dimension tables and stored procedure do not already exist.
    Reporting password Password for the reporting user. The KSY_TRAM_PASSWORD value only needs to be specified if KSY_TRAM_ENABLE is set to Yes and the time dimension tables and stored procedure do not already exist.
    Confirm reporting password Confirm the password for the reporting user.
    Time Dimension Granularity (Minutes) The granularity (in minutes) of the data inserted into the time dimension table. The minimum value is 1. The default value is 5. If the value specified is below the minimum value, the value is reset by the Summarization and Pruning agent at start up to the minimum value.
    Time Dimension Initial Amount (Months) The amount of data in months to be loaded into the time dimension table when it is empty or first created. The minimum value is 1. The default value is 24 months. If the value specified is below the minimum value, the value is reset by the Summarization and Pruning agent at start up to the minimum value.

  11. Select the Scheduling check box to specify when you want summarization and pruning to take place. You can schedule it to run on a fixed schedule or on a flexible schedule:

    If you select Fixed, the Summarization and Pruning Agent does not immediately perform any summarization or pruning when it starts. It performs summarization and pruning when it runs. It runs according to the schedule you specify on the Scheduling pane. If you select Flexible, the Summarization and Pruning Agent runs once immediately after it is started and then at the interval you specified except during any blackout times.

  12. Specify shift and vacation settings in the Work Days pane:

    When you enable and configure shifts, IBM Tivoli Monitoring produces three separate summarization reports:

    • Summarization for peak shift hours

    • Summarization for off-peak shift hours

    • Summarization for all hours (peak and off-peak)

    Similarly, when you enable and configure vacations, IBM Tivoli Monitoring produces three separate summarization reports:

    • Summarization for vacation days

    • Summarization for nonvacation days

    • Summarization for all days (vacation and nonvacation)

    To enable shifts, vacations, or both:

    • Select when the beginning of the week starts.

    • To configure shifts:

      1. Select Yes in the Specify shifts drop-down list.

      2. Optionally change the default settings for peak and off peak hours by selecting hours in the Select Peak Hours box.

        Changing the shift information after data has been summarized creates an inconsistency in the data. Data that was previously collected is not summarized again to account for the new shift values.

    • To configure vacation settings:

      1. Select Yes in the Specify vacation days drop-down list to enable vacation days.

      2. Select Yes in the drop-down list if you want to specify weekends as vacation days.

      3. Select Add to add vacation days.

      4. Select the vacation days you want to add from the calendar.

        On UNIX or Linux, right-click, instead of left-click, to select the month and year.

        The days you select are displayed in the list box.

        To delete any days you have previously chosen, select them and click Delete.

        1. Add vacation days in the future. Adding vacation days in the past creates an inconsistency in the data. Data that was previously collected is not summarized again to account for vacation days.

        2. Enable shifts or vacation periods can significantly increase the size of the warehouse database. It will also negatively affect the performance of the Summarization and Pruning Agent.

  13. Select the Log Settings check box to set the intervals for log pruning:

    • Select Prune WAREHOUSELOG, select the number of units for which data should be kept, and the unit of time (day, month or year).

    • Select Prune WAREHOUSEAGGREGLOG, select the number of units for which data should be kept, and the unit of time (day, month or year).

    These table inserts are disabled by default. The self-monitoring workspaces provided by the Summarization and Pruning Agent provide sufficient information to determine if the agent is operating correctly. Tivoli Data Warehouse log tables can grow very large and require regular pruning. This new default configuration decreases the Summarization and Pruning Agent processing time, and decreases database resource utilization and contention.

  14. Specify additional summarization and pruning settings in the Additional Settings pane:

    1. Specify the number of additional threads you want to use for handling summarization and pruning processing. The number of threads should be 2 * N, where N is the number of processors running the Summarization and Pruning Agent. A higher number of threads can be used, depending on your database configuration and hardware.

    2. Specify the maximum rows that can be deleted in a single pruning transaction. Any positive integer is valid. The default value is 1000. There is no value that indicates you want all rows deleted.

      If you increase the number of threads, you might consider increasing this value if your transaction log allows for it. The effective number of rows deleted per transaction is based on this value divided by the number of worker threads.

    3. Indicate a time zone for historical data from the Use timezone offset from drop down list.

      This field indicates which time zone to use when a user specifies a time period in a query for monitoring data.

      • Select Agent to use the time zone (or time zones) where the monitoring agents are located.

      • Select Warehouse to use the time zone where the Summarization and Pruning Agent is located. If the Tivoli Data Warehouse and the Summarization and Pruning Agent are in different time zones, the Warehouse choice indicates the time zone of the Summarization and Pruning Agent, not the warehouse.

      Skip this field if the Summarization and Pruning Agent and the monitoring agents that collect data are all in the same time zone.

    4. Specify the age of the data you want summarized in the Aggregate hourly data older than and Aggregate daily data older than fields. The default value is 1 for hourly data and 0 for daily data.

    5. The Maximum number of node errors to display refers to the node error table in the Summarization and Pruning workspace. It determines the maximum number of rows that workspace is to save and display.

    6. The Maximum number of summarization and pruning runs to display refers to the Summarization and Pruning Run table in the Summarization and Pruning workspace. It determines the maximum number of rows that workspace is to save and display.

      Maximum number of Summarization and Pruning runs to display and Maximum number of node errors to display together determine the number of rows shown in the Summarization and Pruning overall run table and Errors table respectively. There is a minimum value of 10 for each. These equate to keywords KSY_SUMMARIZATION_UNITS and KSY_NODE_ERROR_UNITS in file KSYENV/sy.ini.

    7. The Database Connectivity Cache Time determines how long after a positive check for connectivity that the result will be cached. Longer times may result in inaccurate results in the workspace; however, it saves processing time.

      Database Connectivity Cache Time records the number of minutes to cache the database connectivity for reporting purposes. The minimum value is 5 minutes. This equates to keyword KSY_CACHE_MINS in file KSYENV/sy.ini.

    8. Batch mode determines if data from different managed systems are used in the same database batch; this setting also improves performance.

      Batch mode controls the batching method used by the Summarization and Pruning Agent. A value of Single Managed System (0) means that data should only be batched for the same system, whereas a value of Multiple Managed System (1) means that data from multiple systems can be batched together; this can lead to higher performance at potentially bigger transaction sizes. The default value is Single Managed System (0). This equates to keyword KSY_BATCH_MODE in file KSYENV/sy.ini.

    9. Specify if you want to turn Database compression on or off.

    10. Specify if you want to turn Database Table Partitioning on or off. When partitioning is enabled, the Summarization and Pruning Agent performs the following functions:

      • Creates new summary tables as partitioned tables.

      • Adds partitions to existing partitioned tables (including the detail tables created by the warehouse proxy) based on your configuration.

      • Prunes data from partitioned tables by dropping the appropriate partitions.

      See Specify range partitioned tables for the Summarization and Pruning Agent.

    To change these values, you can either use the Summarization and Pruning configuration window's Additional settings tab or update these parameters directly in file KSYENV/sy.ini.

  15. Save your settings and close the window. Click Save to save your settings. On Windows, click Close to close the configuration window.

    • Click Save and then click Close.

    • Click Save and then click Cancel.


Parent topic:

Tivoli Data Warehouse solutions: common procedures

+

Search Tips   |   Advanced Search