IBM Books

Network Utility User's Guide


Chapter 9. General Management Tasks

This chapter gives procedures and commands for important Network Utility operations. It serves as a supplement to some of the concept presentations in previous chapters.


Monitoring Events

This section supplements the background information on event logging and viewing provided in Event Logging (Using talk 2, the Monitor Process) and Monitoring Event Messages. It introduces the commands that control what events are logged, and where they are logged.

Accessing the Event Logging System

You must use the command-line interface to activate event logging. From the Configuration Program, you can configure only general remote logging parameters.

From either the main talk 5 or talk 6 prompt, type event to enter the ELS Console or Config subprocess, respectively. You see essentially the same commands whether you are working under talk 5 or talk 6. Talk 5 ELS commands take effect immediately and are quite useful for turning on messages to debug a particular flow in a running system. From talk 6, you configure the events you want to be logged all the time, so you do not have to activate them each time you reboot the Network Utility.

Commands to Control Event Logging

There are six basic commands for activating and deactivating event logging, two for each of the three possible destinations of log messages:

All of these commands use the same method for specifying which events are to be activated or deactivated. Following the name of the command on the command line, you normally type one of the following (there are other options):

The following list gives a few examples of these commands:

disp sub all
Enables logging to talk 2 of all error and unusual informational messages in all ELS subsystems. This is a good general setting to configure in talk 6.

rem sub dls
Enables remote logging of all error and unusual informational messages in the DLS subsystem. Separately, you need to configure the destination host for remote logging.

disp sub sdlc all
Enables logging to talk 2 of all messages in the SDLC subsystem. You might enable all messages when trying to trace an error situation.

nodisp ev sdlc.008
Disables logging to talk 2 of a particularly chatty SDLC message, which may be interfering with seeing more important messages in the error log.

trap ev dls.475
Enables sending an SNMP trap when a particular DLSw QLLC error event occurs.

For detailed information about these commands, how to configure remote logging, what the logging levels are, and more, refer to "Using the Event Logging System (ELS)" in the MAS Software User's Guide.


Monitoring Memory Utilization

This section describes how Network Utility memory is used, and how you can monitor its status.

Network Utility Memory Usage

A Network Utility ships with either 256 MB or 512 MB of main memory. When you boot the system, it loads operational code from disk into this memory, taking a certain amount of memory space for each load module. Once the operational code is loaded, the system splits up the remaining memory between APPN/TN3270 (if configured) and the routing function. The routing function includes IP, DLSw, TCP, channel gateway; in short, every function except APPN and TN3270 server.

When you configure APPN either from the Configuration Program or the command line, you can specify the amount of memory to be reserved for APPN. In Network Utility, this value is preset to the memory required for a maximum TN3270E server configuration12. This value should be reasonable for non-TN3270 APPN applications as well, so you should not need to change it. If your configuration does not enable APPN, Network Utility ignores the configured value and does not reserve memory for APPN. If your configuration enables APPN, Network Utility allocates the specified amount of memory to APPN and then allocates all remaining memory to the routing function.

You can monitor memory utilization in a running Network Utility either from a command-line console or from an SNMP management station. Either way, you look separately at the status of APPN memory and the status of routing function memory. Once the system is loaded, these memory partitions are fixed and are managed independently.

Monitoring Memory from the Command Line

To monitor routing function memory from the command line:

  1. From the * prompt, type talk 5 and press Enter to reach the + prompt.
  2. Type mem to see summary and detailed statistics on current memory utilization. The output uses the term heap to refer to the memory being used by the routing function.

To monitor APPN/TN3270 memory from the command line:

  1. From the * prompt, type talk 5 and press Enter to reach the + prompt.
  2. From the + prompt, type p appn and press Enter to reach the APPN Console subprocess.
  3. Type mem and press Enter to see summary and detailed statistics on APPN memory utilization. The output breaks down APPN memory into various pieces and shows the state of each piece.

Monitoring Memory using SNMP

Network Utility supports IBM enterprise-specific MIBs that provide access to memory utilization information both for the routing function and for APPN/TN3270.

The Nways Manager products discussed in IBM Nways Manager Products provide full statistical support for both the APPN and routing function memory partitions. For either partition, you can view real-time and historical utilization information. You can set up alarm thresholds for either utilization percentage, so you can be notified when memory utilization reaches a certain level.

You can also configure Network Utility from the command line to send an SNMP trap when available routing function memory falls below a given threshold. From the talk 6 prompt Config>, type the command patch mosheap-lowmark and give the percentage value if you want to change it from the default value of 10%.


Monitoring CPU Utilization

This section describes how to control CPU monitoring, and get reports from talk 5 or direct periodic messages to the talk 2 log.

Accessing Performance Monitoring

From either the main talk 5 or talk 6 prompt, type perf to enter the Performance Monitoring Console or Config subprocess, respectively. From talk 6 and from the Configuration Program, you can enable or disable CPU utilization monitoring and set its operating parameters as part of your Network Utility configuration. From talk 5, you can make the same changes take effect immediately, and you can get reports on the CPU utilization in a running Network Utility.

Monitoring CPU Utilization from the Command Line

Once you are at the PERF Console> prompt, the following commands are available to you:

report
Give a summary of current CPU utilization, high water marks, and historical distribution of values.

enable cpu, disable cpu
Control the overall gathering of CPU utilization information. By default, Network Utility runs with CPU utilization enabled, with negligible impact on system performance. If you are running TN3270 server functions with Network Dispatcher, it is particularly important to leave CPU utilization enabled.

enable t2, disable t2
Control the generation of a periodic ELS message in talk 2 showing current CPU utilization. If you enable this message, you can avoid having to repeatedly type the report command to monitor how CPU utilization is changing.

set, list, clear
Set the time window for statistics gathering. View the current values of all settings. Reset statistics.

All the same commands or parameters are available from talk 6 and the Configuration Program, except for clear and report.

For more information on these commands and examples of their output, see "Configuring and Monitoring Performance" in the MAS Software User's Guide.

Monitoring CPU Utilization using SNMP

Network Utility supports an IBM enterprise-specific MIB that provides access to current and historical CPU utilization information.

The Nways Manager products discussed in IBM Nways Manager Products provide full statistical support for Network Utility CPU utilization. You can view both real-time and historical utilization information. You can set up alarm thresholds for the utilization percentage, so you can be notified when it reaches a certain level.


Footnotes:

12
With the introduction of support for 512 MB, the Configuration Program defaults to assume a target Network Utility will have 512 MB of memory. If you load this configuration onto a Network Utility with 256 MB, it will automatically adjust the memory setting downward to the default value for boxes with 256 MB of memory. You do not need to change the Configuration Program default.


[ Top of Page | Previous Page | Next Page | Table of Contents ]