Trap destinations

[Alex Finogenov]

Hi,

I would like to know what the standard or common practice suggests regarding
trap destinations.

W. Stallings in his "SNMP, SNMPv2, SNMPv3 and RMON 1 and 2", 3rd edition,
says that an agent defines the SNMP communities, i.e. it defines the
managers that may access its MIB. RMON 2 is enhanced by defining Trap
Destination Table where managers can define IP addresses of the stations
that can receive traps for a given community, but I couldn't find anywhere
how trap destination IP addresses should be defined in SNMP. This tells me
that without RMON 2 a manager cannot define trap destinations and they are
defined by the agent.

Do you agree? What are the industry practice, and what do the standards say
about this?

Thank you in advance,
Alex