assignment of pTLA ?

[Bob Fink]

Yiorgos,

At 04:17 AM 3/13/98 +0200, Yiorgos Adamopoulos wrote:
>Hi all,
>
>[ newbie mode on ... ]
>
>Maybe it's the local time ( Fri Mar 13 04:14:56 EET 1998 ;-) but although
>I've been through the 6bone.net contents, I cannot seem able to identify a
>document that defines how to ask for a pTLA (if this is possible).
>
>I'd be glad for any pointers, even flames.

The nominal place to look is:

	http://www.6bone.net/6bone_hookup.html

but I'll be the first to admit (and I wrote it) that it is not up to date,
nor clear enough.  I will attempt a cleanup soon.

Meanwhile, I'm enclosing the recent email on pTLAs that I sent to the 6bone
list.  If you have any questions after reading it, please get in direct
email contact with me so the list doesn't have to read it.


Thanks,

Bob

===
6bone Folk,

I am increasingly finding myself in the dilemma of being asked to assign a
pTLA, and then having to decide what is/is not appropriate.  This email is
my attempt to get exposure of the issues and help in deciding an
appropriate course of action in assigning pTLAs.

Let me start by saying that the creation of pTLAs is getting
morecontentious as some believe that we should not create them so easily.
I'm also sure that all of us believe that we must start forging the 6bone
backbone into a reliable transport for ipv6 use and testing.

Having said that we (at least me) need to have some criteria for assigning
pTLAs.  I'll see if I can characterize the criteria I have been using lately.

1. Must have experience with IPv6 in the 6bone, at least as a leaf site and
preferably as an NLA transit under a pTLA.

2. Must have the ability and intent to provide "production-like" 6bone
backbone service to provide a robust and operationally reliable 6bone
backbone.

3. Must have a potential "user community" that would be served by becoming
a pTLA, e.g., the requester is a major player in a region, country or focus
of interest.

4. Must commit to abide by whatever the 6bone backbone operational rules
and policies are (currently there are no formal ones, but the Alain Durand
draft is a start in trying to define some).


To date, when I've explained the above (admittedly not so formally stated
as above) there is typically one of two results. The first is that the
requester goes away and studies more, becomes a leaf site, or forgets the
6bone, etc., that is basically doesn't persist in asking for a pTLA.  The
second is that the requester comes back with strong statements of why it is
important to them, stating that they still want a pTLA assigned.  In the
latter cases I will assign a pTLA (after all, I'm no absolute authority on
any of this).



So, having said all this, I would like to propose a change to this process.
In particular I would like to publish the request along with the

requester's response to the above criteria, and get feedback from the 6bone
mail list (not just the other pTLAs). Then I would make the final call
based on what I think approximates "rough consensus".  

To this end I am enclosing a request for a pTLA from British Telecom Labs
(BT-LABS) and their response to these questions.  I would like responses to
the list on both the process I propose and the BT-LABS request iteself.


Thanks,

Bob

======================
From: Stuart Prevost <stuart.prevost@bt-sys.bt.co.uk> 
To: "'Bob Fink
LBNL'" <RLFink@lbl.gov> 
Subject: Request for pTLA for BT LABS/UK 
Date:
Mon, 2 Feb 1998 15:53:45 -0000 

Dear Bob,

I am writing to request a pTLA on the 6bone, as you know we have been a 
leaf site for a year now. During this time we have gain valuable 
experience in IPv6, and have developed a large site here in the UK. We 
use Cisco routers which currently connect to NRL using BGP4+, in 
becoming a backbone site we feel that we can gain additional experience 
to the benefit of the IPv6 community. 

BT LABS has also formed links to Telenor R&D in Norway. As part of this 
work both companies have a research interest in IPv6. As part of this 
collaboration we plan to create a native IPv6 link to them using the 
JAMES ATM network. I also understand from Tony Dann who attends the 
IETF meetings this will help in the plans to build a Native IPv6 network 
this year.

If you require any additional information before issuing a pTLA please 
let me know.

Regards

Stuart

===
From: Stuart Prevost <stuart.prevost@bt-sys.bt.co.uk> 
To: "'Bob Fink'"
<rlfink@lbl.gov> 
Cc: Tony Dann <tony.dann@bt-sys.bt.co.uk> 
Subject: RE:
Request for pTLA for BT LABS/UK 
Date: Wed, 4 Feb 1998 10:23:12 -0000 

Dear Bob,

Thanks for your response, we don't mind being a test case in helping to 
resolve the issues of criteria for backbone sites. We agree that the 
6bone backbone should evolve into an appropriate infrastructure for true 
IPv6 evaluation. Therefore the response to your criteria for pTLA 
assignment is as follows.

>>1. Must have experience with IPv6 in the 6bone, at least as a leaf site and 
>preferably as an NLA transit under a pTLA.

BT LABS has been a participating in the 6bone global experiment since 
January 1997. In that time we have acted as a leaf site from NRL and 
most recently as a NLA transit site providing connectivity to the 6bone 
for Telenor R&D.

>>2. Must have the ability and intent to provide "production-like" 6bone 
>>backbone service to provide a robust and operationally reliable 6bone 
>>backbone.

BT LABS is the research arm of BT plc, which already has an ISP business 
division. We therefore have the skills to provide a "production like" 
service. We understand the value of the 6bone experiment and intend to 
actively participate in all initiatives which act to increase 
understanding of future IPv6 service issues.

>>3. Must have a potential "user community" that would be served by becoming 
>>a pTLA, e.g., the requestor is a major player in a region, country or focus 
>>of interest.

BT is a major player in Europe and therefore has a large potential user 
community.

>>4. Must commit to abide by whatever the 6bone backbone operational rules 
>>and policies are (currently there are no formal ones, but the Alain Duran 
>draft is a start in trying to define some).

We fully commit to this and welcome progress with Alain Durand draft.
Hope these answers are appropriate and we have no problem with you 
publishing this request to the list. We hope that by starting this off 
that your task will be made easier when assigning pTLA.

Regards,

Stuart

=======from the 6bone registry

ipv6-site:         BT-LABS

origin:            AS1752
descr:             Martlesham Heath
descr:             Suffolk
location:          52 03 52 N 01 17 16 E 0m
country:           GB
prefix:            3FFE:F01:2::/48
application:       ping gate.ipv6.bt.net
tunnel:            IPv6 in IPv4 gate.ipv6.bt.net -> guar.ipv6.nrl.navy.mil
NRL BGP4+
tunnel:            IPv6 in IPv4 gate.ipv6.bt.net -> gate6.lancs.ac.uk ULANC
STATIC
tunnel:            IPv6 in IPv4 gate.ipv6.bt.net -> mbone-eir.nta.no
TELENOR STATIC
contact:           SP1-6BONE
remarks:           Experimental IPv6 evaluation network
remarks:           DNS operational for forward and reverse zones
remarks:           Primary DNS dns.ipv6.bt.net
remarks:           Reverse (.2.0.0.0.0.1.0.F.E.F.F.3.IP6.INT)
remarks:           ipv6-site is operational since 30-Jan-97
changed:           stuart.prevost@bt-sys.bt.co.uk 19971030
source:            6BONE

-end