Re: [EXTERNAL] Re: Embedding IP information in an IPv6 address (OMNI)

Philip Homburg <> Thu, 15 October 2020 09:52 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 5BAAC3A13B9; Thu, 15 Oct 2020 02:52:01 -0700 (PDT)
X-Quarantine-ID: <zOefzZ2FtWWd>
X-Virus-Scanned: amavisd-new at
X-Amavis-Alert: BAD HEADER SECTION, Duplicate header field: "Cc"
X-Spam-Flag: NO
X-Spam-Score: -1.624
X-Spam-Status: No, score=-1.624 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, KHOP_HELO_FCRDNS=0.274, SPF_HELO_NONE=0.001, SPF_NONE=0.001] autolearn=no autolearn_force=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id zOefzZ2FtWWd; Thu, 15 Oct 2020 02:51:58 -0700 (PDT)
Received: from ( [IPv6:2001:888:1044:10:2a0:c9ff:fe9f:17a9]) (using TLSv1.2 with cipher ECDHE-RSA-CHACHA20-POLY1305 (256/256 bits)) (No client certificate requested) by (Postfix) with ESMTPS id B7DD13A13B5; Thu, 15 Oct 2020 02:51:56 -0700 (PDT)
Received: from (localhost [::ffff:]) by with esmtp (TLS version=TLSv1.2 cipher=ECDHE-RSA-CHACHA20-POLY1305) (Smail #157) id m1kSzvJ-0000AXC; Thu, 15 Oct 2020 11:51:49 +0200
Message-Id: <>
Cc: "Templin (US), Fred L" <>
Cc: "" <>
Subject: Re: [EXTERNAL] Re: Embedding IP information in an IPv6 address (OMNI)
From: Philip Homburg <>
References: <> <> <>
In-reply-to: Your message of "Wed, 14 Oct 2020 15:25:24 +0000 ." <>
Date: Thu, 15 Oct 2020 11:51:44 +0200
Archived-At: <>
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: "IPv6 Maintenance Working Group \(6man\)" <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 15 Oct 2020 09:52:01 -0000

>We really do want to define the link-local address format for OMNI interfaces.
>Too many things depend on every IPv6 interface configuring a unique link-local
>address. The address format and the means by which it is assured unique is wha
>we want to specify for OMNI in an "IPv6-over-foo"-specifc document.

A few thoughts:
1) I think a general system for an IPv6-over-IPv6 overlay network is useful.
   I don't know enough about OMNI to know if OMNI is the right answer, but
   some of the concepts in OMNI seem quite useful to me.
2) There is too much code that knows about the link local prefix, that 
   adding a another one is probably not going to fly. 
3) Traditionally (i.e. since the inception of IPv6) IIDs are 64 bits. So it
   makes sense for OMNI to conform to that. Embedding the MNP in the IID 
   means that the MNP can be at most 64 bits long. I don't see any big problem
   with that.
4) However, the main thing standing in the way of using all 118 bits allowed by
   the fe80::/10 prefix seems to be the *BSD hack of putting an interface 
   number in the zero bits of a link local address. I think that the BSD
   communities should remove this hack, and we should not effectively let
   them squat those bits.
5) I don't see any argument why the MNP should be longer than 64 bits. So
   point 3 seems the best way to go forward.