Administration Guide

Chapter 27. Managing SP system events in a network environment

IBM Parallel System Support Programs for AIX includes a Problem Management subsystem (refer to Chapter 26, Using the Problem Management subsystem) that provides access to Event Management function without the necessity of writing C programs that use the Event Management APIs. The Problem Management subsystem gives you the ability to subscribe to Event Management events and to specify actions for those events. Issuing an SNMP trap in response to an event (which can come either from Event Management or the AIX Error Log) is one of the actions you can specify.

The ability to issue an SNMP trap in response to an event allows you to report problem events occurring in your SP system to a network manager existing on a remote node (a network manager application is not supplied with the SP). The Problem Management subsystem provides an SP SNMP proxy agent, sp_configd, (sometimes referred to as a subagent daemon) that runs on the control workstation and every SP processor node. The SP proxy agent provides the following functions:

1. A Management Information Base (MIB)
2. SNMP GET and GET NEXT command support that allows data in the MIB to be accessed by a network manager application
3. Creation and transmittal of SNMP traps to an installation-defined network manager application when the following events occur on the node on which the SP proxy agent is running:
   • A cold start trap is issued when the agent is activated
   • An enterprise-specific trap is issued when an entry with an "alert=true" attribute is written in the AIX Error Log
   • An enterprise-specific trap is issued when an user-specified event is detected within Event Management services

This chapter provides overview and task information for making the SP part of a network management system. In such a system, for example, the SP presents error event information, in the form of SNMP traps, to a network management application, such as NetView for AIX, which displays and logs the trap, and may notify a problem management application, such as Trouble Ticket for AIX, of problem events occurring on the SP.

In order for the SP to be part of such a network management system, network managers must be notified when selected AIX Error Log entries are written and Event Management events occur. SP-specific configuration data is provided so that management applications can determine which nodes compose the SP system. You decide what Error Log and Event Management events will trigger manager notification.

This chapter first discusses general concepts related to Network Management and then discusses information specific to enabling the reporting of SP system events in a network environment. For further discussion of Simple Network Management Protocol and the Management Information Base, refer to IBM AIX Communications Programming Concepts.