Re: technical question

Brian O'Keefe <bok@nsmdserv.cnd.hp.com> Fri, 17 November 1995 17:31 UTC

Message-Id: <9511171609.AA04817@nsmdserv.cnd.hp.com>
To: Bert Wijnen <wijnen@vnet.ibm.com>
Cc: snmpv2@tis.com
Subject: Re: technical question
In-Reply-To: Your message of Fri, 17 Nov 1995 15:23:26 +0700. <199511171359.IAA20102@relay.tis.com>
Date: Fri, 17 Nov 1995 09:09:04 -0700
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Brian O'Keefe <bok@nsmdserv.cnd.hp.com>

Bert,

I believe that what you would like to do is the intention
of the revised snmpv2-mib... that is the old rfc1213 objects 
that proved to be overboard are reclassified as "obsolete".

Despite the name of the mib "snmpv2-mib", I see a mib as a 
mib as a mib and should be able to be implemented using any
version of the snmp protocol that includes the necessary
pre-requisites (for example, you need snmpv2 to implement
counter64 objects). 

just my 2 cents,
bok

> If I want to provide a bi-lingual SNMPv1/SNMPv2C agent,
> is it then acceptable/advisable to do away with the old snmp Group
> of RFC1213 and just implement the snmpGroup as defined in the
> snmpv2-mib document?
> 
> I personally think that that is OK and that that should be the
> recommended implementation. I think we went overboard with the
> old snmp group that counted too many things.
> 
> I'd say that even a SNMPv1 agent could use the new snmpv2-mib
> instead of the system and snmp groups out of RFC1213.
> 
> Thanks, Bert Wijnen (IBM Reserach).
> 

 bok