IBM Tivoli Monitoring > Version 6.3 Fix Pack 2 > Installation Guides > Agent Installation Guides > IBM i Agent Installation Guide > Agent installation and configuration

IBM Tivoli Monitoring, Version 6.3 Fix Pack 2


Configure the monitoring agent


Objective

To configure or reconfigure the network connections between the Monitoring Agent for IBM i OS and the Tivoli Enterprise Monitoring Server (monitoring server).


Background information

You must use the IBM i OS non-programmable terminal interface to configure, start, and stop the Monitoring Agent for IBM i OS. Also use this interface to view the Monitoring Agent for IBM i OS message log.

For more information about using the non-programmable interface, refer to the online help. For more information about command and menu interfaces and working with message logs, refer to the documentation provided with the IBM i OS system.

If your environment includes a firewall between any IBM Tivoli Monitoring components, specify IP.PIPE as your communications protocol during configuration. For more information about firewall support including requirements for firewall configurations that use address translation, refer to the following sections in the IBM Tivoli Monitoring Installation and Setup Guide:


Required authorization role

*USER You need authority to access the agent commands. By default, they all are *PUBLIC *EXCLUDE with some user group profiles given *USE authority as shown in Table 1. Use the GRTOBJAUT command to add authorization for other users. Depending upon what options are specified for the CFGOMA, *ALLOBJ and *IOSYSCFG authorities might be required to successfully configure the Monitoring Agent.

Install the monitoring agent as described in Install the monitoring agent.


When you finish

Start the Monitoring Agent for IBM i OS so you can begin using the monitoring agent to monitor the IBM i OS resources. For information about how to start the Monitoring Agent for IBM i OS, see Start the monitoring agent.


Procedure

  1. From an IBM i OS command line, enter the following command:

      GO OMA

  2. Enter 4 (Configure Tivoli Monitoring: IBM i OS Agent).

    The Config IBM i OS Monitoring Agent (CFGOMA) window is displayed.

  3. Enter your site's values for the configuration parameters using the guidelines in Table 1.


    Configuration parameters

    Parameter Description
    TEMS TCP/IP address The TCP/IP address or host name of the computer where the monitoring server resides, such as 127.0.0.1 or RALEIGH. If you use the IP.PIPE or IP.SPIPE parameters, enter *NONE. If the correct TCP/IP address or host name was previously defined, enter *SAME to retrieve this setting.
    TEMS IP.PIPE address If the monitoring agent must connect to the monitoring server through a firewall, use the IP.PIPE communication protocol. Specify the IP.PIPE address or host name of the computer where the monitoring server resides. If you are not using the IP.PIPE communication protocol, enter *NONE.
    TEMS IP.SPIPE Address You can change the local Secure Socket Layer (SSL) IP.SPIPE location in an enterprise network that is using SSL IP.SPIPE communications. Configuration on the agent and the Tivoli Enterprise Monitoring Server must be completed for SSL communications to function.
    Secondary TEMS IP address The TCP/IP address or host name of the computer where the secondary monitoring server resides. The monitoring agent communicates with the secondary monitoring server if it cannot communicate with the primary monitoring server at startup.
    Secondary TEMS IP.PIPE address The IP.PIPE address or host name of the computer where the secondary monitoring server resides. The monitoring agent communicates with the secondary monitoring server if it cannot communicate with the primary monitoring server at startup.
    Partition name (Required only by sites with firewalls that use address translation.) The name of the partition (up to 32 alphanumeric characters) in which the monitoring agent resides.
    Firewall in use If the monitoring agent must connect to the monitoring server through a firewall, enter *YES. If the monitoring agent does not connect through a firewall, keep the default value, *NO.
    TEMS TCP/IP port address The listening port for the monitoring server. The default number is 1918. If the correct port was previously defined, enter *SAME to retrieve this setting.
    TEMS IP.PIPE port address The listening port for the monitoring server. The default is 1918.
    TEMS IP.SPIPE Port Number Secure Shell port number.
    TCP/IP Server Specifies whether or not the Tivoli Monitoring: IBM i OS Agent is defined as a TCP/IP server. If it is a TCP/IP server then it can be started and stopped using the STRTCPSVR and ENDTCPSVR commands. The agent will also be automatically ended when TCP/IP is ended. If the agent is not defined as a TCP/IP server then you must start it after TCP/IP is started and end it before TCP/IP is ended.
    Action user profile The user authority under which user action must be administered. Keep the default value, QAUTOMON, to grant user system operator authority.

  4. Optional: Customize the data collection intervals by changing the values of the following configuration variables in the QAUTOTMP/KMSPARM[KBBENV] file, which are listed with their default values:

    • KA4_JOB_DATA_INTERVAL=15

    • KA4_IOP_DATA_INTERVAL=30

    • KA4_DISK_DATA_INTERVAL=30

    • KA4_POOL_DATA_INTERVAL=15

    • KA4_COMM_DATA_INTERVAL=60

    Valid values for these configuration variables are 15, 30, 60, 120, and 240. These configuration variables follow the rules of the collection interval parameter of the IBM i OS QPMWKCOL API. Keep the following items in mind:

    • Disk and IOP-related data require a minimum of 30 seconds between collection intervals.

    • Communication-related data requires a minimum of 60 seconds between collection intervals.

    • Collect job-related data as infrequently as possible to minimize the impact on system performance.

    • The IBM i OS collection services performance data collector supports data collection at one-minute intervals, not at two or four-minute intervals. Therefore, when using the API and requesting data at two or four-minute intervals, the data is collected at one-minute intervals but reported back every two or four minutes.


Parent topic:

Agent installation and configuration

+

Search Tips   |   Advanced Search