Hitachi

JP1 Version 12 JP1/Base User's Guide


I.1 Using the SNMP trap converter to convert SNMP traps into events

The following figure shows how SNMP traps are converted into JP1 events and registered in an event database.

Figure I‒1: Overview of SNMP trap conversion to JP1 event registration

[Figure]

To use the SNMP trap converter, you need to create the SNMP trap conversion action definition file (imevtgw.conf) and SNMP trap conversion filter file (snmpfilter.conf). These two files are used to specify the conditions for converting SNMP traps into JP1 events, and the severity levels for JP1 events. The SNMP trap converter begins when NNM starts.

While it is running, the SNMP trap converter obtains SNMP traps that satisfy the conditions specified in the SNMP trap conversion filter file (snmpfilter.conf), and then converts these traps into JP1 events. The acquired information consists of the event message, severity level, enterprise name, enterprise ID, object name, object ID, and source list. If the SNMP trap converter is not running, output SNMP traps are not converted into JP1 events. Trapped SNMP messages can be registered as JP1 events using a maximum of 1,023 bytes. If a message exceeds this maximum, the message is truncated from the 1,024th byte when the message is converted into a JP1 event.

All JP1 events converted from SNMP traps are assigned an event ID of 00003A80. For details on attributes of JP1 events, see I.5 JP1 events for SNMP trap conversion.

The SNMP trap converter checks for syntax errors when reading the action definition file and the filter file for converting SNMP traps (snmpfilter.conf). If a syntax error is found, a message appears.

Organization of this subsection

(1) SNMP trapping in a cluster system

The SNMP trap converter operates only on a physical host. Also, it operates by using the NNM function linked with the starting up and shutting down of NNM. It therefore operates independently of JP1/Base failovers.

By default, JP1 events are registered in the event service on the physical host. To register JP1 events in the event service on a logical host, specify the event server name of the logical hosts for the imevt_server parameter in the action definition file for converting SNMP traps. However, if NNM is used in a non-cluster system, and that system is configured to directly register converted JP1 events to the logical host, the standby node cannot monitor the SNMP traps it receives.

The following figure illustrates how to configure a cluster system (where NNM is used) to directly register JP1 events to a logical host.

Figure I‒2: Configuration example for directly registering JP1 events to a logical host

[Figure]

To use NNM in a cluster system, set up NNM on both the primary and secondary nodes, referring to I.2 Setting the SNMP trap converter. Register NNM and JP1/Base in the same cluster group.

If you are using NNM in a cluster system and JP1/Base in a non-cluster system (that is, these programs are being used on the physical hosts only), you must start JP1/Base on the physical host at both the primary and secondary nodes.

(2) Criteria for converting SNMP traps

The following criteria apply to SNMP trap conversion:

Note that the following SNMP traps used internally in NNM cannot be converted into JP1 events:

For details, see the NNM documentation.

(3) Note