Re: [Isms] open issues

["Randy Presuhn" <randy_presuhn at mindspring.com>]

Hi -

> From: "Juergen Schoenwaelder" <j.schoenwaelder at jacobs-university.de>
> To: <Pasi.Eronen at nokia.com>
> Cc: <isms at ietf.org>
> Sent: Thursday, April 24, 2008 12:31 PM
> Subject: Re: [Isms] open issues
...
> I am not so much concerned about the multiplexing on the notification
> receiver side. The real issue is that we need an authenticated
> principal for the access control to work correctly and there might be
> several different principals behind the same notification receiver
> endpoint.
...

With USM, even though the privacy key gives us the other half
of the authentication, we *still* have to trust the notification receiver
to demultiplex things correctly, and that's beyond the scope of the
SNMP RFCs.

So I think the real concern has be making sure that the SNMP
engine in front of the notification receiver application is indeed
the one requested by the subscriber, while still maintaining
sufficient information in the PDU to permit demultiplexing.

The question would still arise about the possibility of a forged
subscription, but that is handled through the design of the
target and notification MIBS in conjunction with VACM, as
long as we don't start coalescing securityNames.

Randy

_______________________________________________
Isms mailing list
Isms at ietf.org
https://www.ietf.org/mailman/listinfo/isms