IBM Support

Dynamically modify trace settings for an IBM Tivoli Monitoring component

Troubleshooting


Problem

You can access the Tivoli Enterprise Monitoring Server, Tivoli Enterprise Portal Server, almost all of the agents, and other IBM Tivoli Monitoring components from this utility.

Resolving The Problem

This method of modifying trace settings on the Tivoli Enterprise Monitoring Server is the most efficient method since it allows you to do so without restarting this component. Settings take effect immediately. Modifications made this way are not persistent.

How to turn tracing on:

In order to use this utility you need to know a local log-on credential for the machine.

This method uses the IBM Tivoli Monitoring Service Console. The Service Console is accessed using a web browser. Access the utility by using the following link:

http://hostname:1920

where hostname is the hostname or IP address of the system where the ITM component is running.

The utility then appears with information about the components that are currently running on this machine. For example:



This screen shows that the Tivoli Enterprise Portal Server (cnp), Windows OS agent (_nt), and the Tivoli Enterprise Monitoring Server (ms) are running on this system.

Select the link below the component for which you want to modify the trace settings. In the previous view if you want to modify tracing for the Tivoli Enterprise Monitoring Server, you select the "IBM Tivoli Monitoring Service Console" link under the Service Point: system. balayne_ms.

When you select one of the links, you will be prompted for a user ID and password to access the system. This is any valid user that has access to the system. After you log in you will see a screen like the following:


Typing ? displays a list of the supported commands:

The command for modifying the trace settings is ras1.

If you type ras1 in the field at the bottom of the screen, you will then see the help for this command.

The set option (ras1 set) turns on the tracing for the unit or component identified in the command, but does not affect existing tracing. The format of the command is similar to the format used to set trace settings in the env (Windows) or ini (Unix) files. The settings contained in the files are persistent and are a complete set of trace definitions. The values set using the service console are added to the existing settings and are not persistent. Example formats include:

(UNIT:kbbacdl Flow State) - turns on the flow and state tracing for the files kbbacdl* in the ITM component.

Error - turns the error trace on for all aspects of the component

(COMP:kdh Detail) - turns the detail tracing on for all of the files identified as part of the kdh component.

Any number of these formats may be combined to affect tracing in several files or components. An example would be ras1 set (UNIT:xxx ALL) (UNIT:yyy Detail). This command will enable full tracing for the xxx class of the component and low-level detailed tracing on the yyy class of the component.

The ras1 list command displays the current trace settings. It is best to do an initial list in order to track what changes you have made to the tracing settings.

The following list describes the types of tracing available:

ALL - Provides all trace levels. Shown as ALL when using the ras1 list command.

Flow - Provides control flow data describing function entry and exit. Shown as Fl when using the ras1 list command.

ERROR - Logs internal error conditions. Shown as ER when using the ras1 list command. The output also shows as EVERYE+EVERYU+ER.

Other settings which provide component specific information are:

Detail - Shown as Det when using the ras1 list command.

INPUT - Shown as IN when using the ras1 list command.

Metrics - Shown as ME when using the ras1 list command.

OUTPUT - Shown as OUT when using the ras1 list command.

State - Shown as ST when using the ras1 list command.

Setting trace to ALL includes every trace point defined for the component. This might result in a large amount of trace. If you have been given a more specific setting, use it. ALL can sometimes be necessary when isolating a problem. It is the equivalent of setting "Error Detail Flow State Input Output Metrics".


The ras1 units command is used to determine the list of UNITs and COMPs available in an ITM component. The first column is the list of available UNIT values, the last column lists the corresponding COMP values.

Turning on (COMP:KDH ALL) will turn ALL level tracing on for all of the files where KDH is listed in the right hand column (highlighted below).

The following is a subset of the results for the Windows agent:

kbbcre1.c, 400, May 29 2007, 12:54:43, 1.1, *
kbbcrn1.c, 400, May 29 2007, 12:54:42, 1.1, *
kdhb1de.c, 400, May 29 2007, 12:59:34, 1.1, KDH
kdh0med.c, 400, May 29 2007, 12:59:24, 1.1, KDH
kdhsrej.c, 400, May 29 2007, 13:00:06, 1.5, KDH
kdhb1fh.c, 400, May 29 2007, 12:59:33, 1.1, KDH
kdhb1oe.c, 400, May 29 2007, 12:59:38, 1.2, KDH
kdhs1ns.c, 400, May 29 2007, 13:00:08, 1.3, KDH
kbbacdl.c, 400, May 29 2007, 12:54:27, 1.2, ACF1
kbbaclc.c, 400, May 29 2007, 12:54:27, 1.4, ACF1
kbbac1i.c, 400, May 29 2007, 12:54:28, 1.11, ACF1
kdhsfcn.c, 400, May 29 2007, 13:00:11, 1.1, KDH
kdhserq.c, 400, May 29 2007, 12:59:53, 1.1, KDH
kdhb1pr.c, 400, May 29 2007, 12:59:39, 1.1, KDH
kdhsgnh.c, 400, May 29 2007, 12:59:49, 1.1, KDH
kdh0uts.c, 400, May 29 2007, 12:59:23, 1.1, KDH
kdhsrsp.c, 400, May 29 2007, 13:00:13, 1.2, KDH
kdhs1rp.c, 400, May 29 2007, 13:00:12, 1.1, KDH
kdhscsv.c, 400, May 29 2007, 12:59:58, 1.9, KDH
kdebbac.c, 400, May 29 2007, 12:56:50, 1.10, KDE



The UNIT value matches any unit that starts with the specified value. For example, (UNIT:kra FLOW) prints the FLOW traces for all files which match kra*.

How to turn tracing back off:

The option for turning the tracing off is ANY. For example you would use the following command to turn off tracing for the kbbcrcd class of the Windows OS agent:

ras1 set (UNIT:kbbcrcd ANY)

Note: In some functions, tracing continues even after you turn it off dynamically. The tracing controls are evaluated at the beginning of a function, and if the control is in a "do forever" type event processing, diagnostic tracing continues until the IBM Tivoli Monitoring system is recycled.

Note: Consider the service console a security risk. Entry is controlled by a user ID and password that provides one barrier. No new functions can be performed, but the trace controls can be set so that a very high volume of tracing is created, and that could theoretically impact the ability to service normal work. The initial display also lists services, which might expose more than is wanted. If this is a problem, the initial console display can be prevented by adding the following to KDC_TRANSPORT or KDC_FAMILIES at the front:



HTTP_CONSOLE:N

However, dynamic tracing cannot be used if problems are experienced.

Basic Services tracing

The service console command bss1 config can be used to turn on tracing for some basic services subsystems. This is most often used for communications tracing. For example, you would use the following command to turn on basic services tracing:

bss1 config KDC_DEBUG=Y

The following command changes the tracing back to the default:

bss1 config KDC_DEBUG=N

These commands are case sensitive, so they must entered as presented above.

IBM Support will suggest what values are needed for diagnosis. Most often the settings are for KDC_DEBUG, KDE_DEBUG and KDH_DEBUG. On rare occaisions values different from Y or N are used.

[{"Product":{"code":"SSTFXA","label":"Tivoli Monitoring"},"Business Unit":{"code":"BU053","label":"Cloud & Data Platform"},"Component":"--","Platform":[{"code":"PF002","label":"AIX"},{"code":"PF010","label":"HP-UX"},{"code":"PF012","label":"IBM i"},{"code":"PF016","label":"Linux"},{"code":"PF027","label":"Solaris"},{"code":"PF033","label":"Windows"}],"Version":"All Versions;6.3.2;6.3.1;6.3.0;6.3;6.2.3;6.2.2;6.2.1","Edition":"","Line of Business":{"code":"LOB45","label":"Automation"}}]

Document Information

Modified date:
17 June 2018

UID

swg21266129