Some quick comments based on my first pass review, all as an individual
not a chair:

Right now, your problem statement is accurate, in that a router with no
defined IPv4 addresses has no way to derive a 32-bit system ID without
further user input, either in the form of a new algorithm or manually
defining the ID. However, if you are asserting that effort should be made
to preserve this link between the two, there needs to be better discussion
of why the value shouldn't be arbitrary (e.g. a 32-bit random number) as
long as it is unique in the scope that it needs to be. Yes, there are some
instances that you've discussed in the introduction where it's marginally
more useful to have the router-id correspond to a value present elsewhere
on the device. However, I don't think you've really defined any specific
problem that this causes, since it is unlikely that the only identifying
information will be the router ID. Are there any places in the UI of
standard routers or in actual data sent on the wire where the only
information presented is the router-id, thus requiring there to be a
correlation between that value and something elsewhere on the router? If
so, it's possible that the underlying problem is that some implementations
are assigning too much contextual value to an arbitrary field, and this
document needs to have guidance to implementers that they should stop
doing this. Either way, we need to get beyond hand-waving, and into more
specific discussion of this problem space.

What we have to do is separate convention (things that were done because
the info was there) from legitimate areas where the lack of this
information and correlation causes extra steps in troubleshooting or
locating a given router. That will in turn drive your solution space - if
there are places where that correlation needs to be preserved, then an
algorithm for deriving IDs from IPv6 addresses in a way that is
human-readable might be necessary. If not, then this is more about
suggesting a couple of suitably random data sources on the router to use
as input to generating a router-ID, possibly in conjunction with a method
to detect and correct duplication within the scope where they must be
unique.

"there has been suggestion about avoiding protocol change" - better
phrased as "protocol change may be unnecessary to address this problem" --
and more to the point, you need to explain what protocol change you're
suggesting is unnecessary (that existing 32-bit ID fields should be
increased to 64 or 128 bit fields or new fields added in order to
accommodate an IPv6 address). This is obvious to me, because I had read
your previous drafts suggesting that solution in the routing WGs, but
should be explicitly stated in this draft for those not familiar with that
background context.

As the problem statement is refined, it may also help us work through the
different scenarios so that we are able to say definitively whether there
are any cases where the value of the router ID is important to the
protocol and thus must be a valid IP address (v4 or v6), and whether or
not there are any gaps where the protocol still needs to be updated to
handle IPv6 addresses as IDs on an IPv6-only network. Those would likely
spawn separate documents in the appropriate protocol's WG.




Thanks,

Wes George




On 6/20/14, 4:34 AM, "Fan, Peng" <[email protected]> wrote:

>Hi all,
>
>The following draft is about problems of managing router IDs in an
>IPv6 only network as well as some solution ideas. It is an effort
>based on the feedback during the last IETF meeting. Your review and
>comments are appreciated.
>
>Best regards,
>Peng
>
>---------- Forwarded message ----------
>From: [email protected]
>Date: Fri, 20 Jun 2014 00:50:46 -0700
>Subject: New Version Notification for draft-fan-sunset4-router-id-00.txt
>To: Peng Fan <[email protected]>
>
>
>A new version of I-D, draft-fan-sunset4-router-id-00.txt
>has been successfully submitted by Peng Fan and posted to the
>IETF repository.
>
>Name:          draft-fan-sunset4-router-id
>Revision:      00
>Title:         Managing Router Identifiers during IPv4 Sunset
>Document date: 2014-06-20
>Group:         Individual Submission
>Pages:         3
>URL:
>http://www.ietf.org/internet-drafts/draft-fan-sunset4-router-id-00.txt
>Status:
>https://datatracker.ietf.org/doc/draft-fan-sunset4-router-id/
>Htmlized:       http://tools.ietf.org/html/draft-fan-sunset4-router-id-00
>
>
>Abstract:
>   This document describes problems of managing protocol identifiers
>   when turning off IPv4 and migrating to IPv6 only network, with some
>   potential solutions discussed.
>
>
>
>
>Please note that it may take a couple of minutes from the time of
>submission
>until the htmlized version and diff are available at tools.ietf.org.
>
>The IETF Secretariat
>
>_______________________________________________
>sunset4 mailing list
>[email protected]
>https://www.ietf.org/mailman/listinfo/sunset4


This E-mail and any of its attachments may contain Time Warner Cable 
proprietary information, which is privileged, confidential, or subject to 
copyright belonging to Time Warner Cable. This E-mail is intended solely for 
the use of the individual or entity to which it is addressed. If you are not 
the intended recipient of this E-mail, you are hereby notified that any 
dissemination, distribution, copying, or action taken in relation to the 
contents of and attachments to this E-mail is strictly prohibited and may be 
unlawful. If you have received this E-mail in error, please notify the sender 
immediately and permanently delete the original and any copy of this E-mail and 
any printout.
_______________________________________________
sunset4 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/sunset4

Reply via email to