[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: snmpconf Trap parameters in snmpSend



Two comments: Steve M. thanks for posting these points they are helpful.
The second is for Wes and is below:
> >>>>> On Mon, 16 Jul 2001 11:34:50 -0400, Steve Moulton <moulton@snmp.com> said:
> 
> Steve> 1  Add a different accessor function for sending v1 traps, with
> Steve> the associated generic-trap, specific-trap, and enterprise
> Steve> parameters.  The agent-addr parameter could either require an
> Steve> additional argument, or be intuited from an IP address from the
> Steve> agent.
> 
> Steve> I don't really favor this. 
> 
> Steve> 2 Specify that, when using the Trap operation with snmpSend,
> Steve> that the supplied parameters should follow V2Trap semantics,
> Steve> and automatically converted using the rules of rfc2576, section
> Steve> 3.2.
> 
> Steve> I favor this approach, as it removes the need for adding an
> Steve> accessor function for an exception case.
> 
> FYI, this is functionally what the net-snmp API does (as well as many
> other APIs, I suspect) and we end up supporting both the "I'm sending
> a v1 trap" and "I'm sending a v2 trap" and they both get automagically
> converted (when possible).  However, if I were to pick one (which I
> think SHOULD be done for snmpconf for simplicity purposes) I'd say go
> with specifying v2 traps and having them get automatically converted
> to v1 if the CR/NR in question supports v1 trap generation.
> -- 

By you last sentence you mean what by CR/NR. Is it not true that the
node will be configured to genererate the appropriate trap or inform to
the correct destinations based on the agent configuration? What am I
missing?

> Wes Hardaker
> NAI Labs
> Network Associates
> 

Thanks,
/jon
--

Jon Saperia		     saperia@jdscons.com
			     Phone: 617-744-1079
			     Fax:   617-249-0874
			     http://www.jdscons.com/