[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: inet address mib issue
- To: mibs@ops.ietf.org
- Subject: RE: inet address mib issue
- From: "Wijnen, Bert (Bert)" <bwijnen@lucent.com>
- Date: Wed, 10 May 2000 18:07:51 +0200
- Delivery-date: Wed, 10 May 2000 09:09:23 -0700
- Envelope-to: mibs-data@psg.com
Does the discussion we had about this topic mean that
maybe we should add an IpAddressPrefix TC that allows
to specify a length?
Bert
> ----------
> From: Peder Chr. Nørgaard[SMTP:pcn@tbit.dk]
> Sent: Thursday, May 04, 2000 8:27 AM
> To: Andrew Smith
> Cc: 'Michael Thatcher'; mibs@ops.ietf.org
> Subject: RE: inet address mib issue
>
> On Wed, 3 May 2000, Andrew Smith wrote:
>
> > Mike,
> >
> > I had the same issue with the Diffserv MIB that I'm hacking right now -
> I've
> > added an additional InetAddressType variable, as the immediately
> preceding
> > OID to the matching InetAddress variable, as required by the Textual
> > Convention instructions.
> >
> > For Address/Mask pairs though, Fred Baker pointed out that Masks are
> almost
> > universally contiguous and that a mask length was a more appropriate
> > solution. I don't know the details of your problem but the same answer
> may
> > help you too.
>
> I would second this; I have written and implemented several private MIBs
> containing IPv4/IPv6 address structures (not using the InetAddress yet,
> but that will change with the issuing of this MIB.)
>
> For masks I have always found it much easier to simply use mask lengths,
> both for IPv4 and IPv6. A side effect: it saves space in the PDUs,
> especially when the mask/mask length is part of an index.
>
> I think the only place in the standards where non-contiguous masks are
> allowed is in the old IPv4 subnet specifications - and I now of noone who
> actually uses those, nor can I see any reason to.
>
> Prefix masks, in my opinion, is a relic. Of course, standard MIBs, living
> a long life, must necessesarily contains relics; but there is usually no
> reason to keep them when building new MIBs.
>
> best regards
>
>
> >
> > Andrew
> >
> > > -----Original Message-----
> > > From: Michael Thatcher [mailto:thatcher@redback.com]
> > > Sent: Wednesday, May 03, 2000 2:05 PM
> > > To: mibs@ops.ietf.org
> > > Subject: inet address mib issue
> > >
> > >
> > > the description states that each object of InetAddress requires a
> > > corresponding object of InetAddressType. This seems a little too
> > > restrictive when I want to define an address and a mask for
> > > the address.
> > > Even though the mask will always have the same type as the
> > > address, I'm
> > > still required to have another InetAddressType object for the mask.
> > >
> > > miket
> > >
> > >
> >
>
> --
> Peder Chr. Nørgaard Senior System Developer, M. Sc.
> Ericsson Telebit A/S tel: +45 86 75 68 49
> Fabrikvej 11 fax: +45 86 75 68 81
> DK-8260 Viby J mob: +45 21 28 66 58
> Denmark e-mail: pcn@tbit.dk
>
>