2.11.2 How to use enterprise-specific traps
The agent sends the traps using the snmptrap command. For example, you can configure your agent to send traps by executing the snmptrap command from a shell script.
As a manager, you have two alternatives when monitoring the status of an agent. You can
-
Continuously poll the agent from the manager to get information.
-
Send a trap from the agent to the manager.
Polling creates a lot of traffic on the network and, if an event occurs shortly after polling has taken place, the manager might not find out about an event for an extended period of time. The key benefits of using the snmptrap command are that you can decrease the amount of SNMP traffic on the network and that you can find out about an event sooner.
If you have NNM, you can customize your environment by using the snmptrap command in conjunction with the Event Configuration operation. For example, assume that you have written a script on an agent that executes the snmptrap command when a particular process on the agent goes down. You can then use the Event Configuration operation from the NNM station to take an action when the manager receives that particular trap from the agent.