clarification request: V2->V1 Trap conversion
Dave Shield <[email protected]>
| Newsgroups | gmane.ietf.snmpv3 |
|---|---|
| Message-ID | <[email protected]> |
A quick question about an issue that's come up on the Net-SNMP coding list.
When converting an SNMPv2 notification PDU into an SNMPv1 trap, the
original varbind list will include varbinds for sysUpTime, snmpTrapOID
and snmpTrapEnterprise. Section 3.2 RFC 2576 describes how to use these
to determine the appropriate values for the SNMPv1 trap header fields,
and then says:
(6) The SNMPv1 variable-bindings SHALL be the SNMPv2 variable-
bindings. [... mumble mumble Counter64 mumble ....]
This variable-bindings list would appear to still include the sysUpTime,
snmpTrapOID and snmpTrapEnterprise varbinds. (Or at least, I can't
immediately spot anywhere that says these varbinds should be removed).
The effect is that this information effectively ends up being duplicated.
Is this correct ?
The Net-SNMP master agent treats incoming AgentX notifications as if
they were SNMPv2 notifications, so applies the same mappings when
sending to SNMPv1 destinations. This means that subagent-generated
notifications end up with the same duplication of information.
I realise this isn't strictly an SNMPv3-related issue, but does
this approach strike people as:
a) reasonable
b) valid
?
Dave
--
Dave Shield [email protected]
Dept. of Computer Science,
Liverpool University, "He who brings [computers] on to his premises
PO Box 147, should be absolutely liable ... for any mischief
Liverpool, L69 7ZF that ensues." Haddock v. Computer 1578/32/W1