[37] in SIPB IPv6
Re: IPv6 coordination
daemon@ATHENA.MIT.EDU (Ken Raeburn)
Thu Aug 1 16:34:51 2002
To: Noah Meyerhans <noahm@lcs.MIT.EDU>
Cc: sipbv6@MIT.EDU
From: Ken Raeburn <raeburn@MIT.EDU>
Date: Thu, 01 Aug 2002 16:34:38 -0400
In-Reply-To: <20020801191721.GG7732@locust.lcs.mit.edu> (Noah Meyerhans's
message of "Thu, 1 Aug 2002 15:17:21 -0400")
Noah Meyerhans <noahm@lcs.mit.edu> writes:
> On Thu, Aug 01, 2002 at 02:52:03PM -0400, Ken Raeburn wrote:
>> d) ipv6-accessible system providing one or more web pages describing
>> our services - do we have *any* ipv6 web servers?
> debian.ipv6.lcs.mit.edu mirrors Debian GNU/Linux, the info-mac Mac
> shareware archive, some IPv6 specific Debian stuff, and (in a couple of
> days) the Winsite shareware archive. It is available via IPv6 for http
> and ftp.
If it has a web page describing such, that might qualify as "the
Applicant's IPv6 services", but I suspect it refers more to the giving
out of IPv6 tunnels. Now, if SIPB's web server could be made IPv6
reachable....
Ooh! My mistake -- "www.ipv6.mit.edu" points to limekiller, which is
running tcpforward to www.mit.edu. So we're all set there. It's
reachable via ipv6, and the fact that it's a cheap hack -- uh, I mean,
that it accesses a non-local database for its web content -- shouldn't
matter.
> After seeing some of the other pTLA allocations on the 6bone mailing
> list I'd say that this is no trouble for us, regardless of JIS'
> involvent. As it is, I consider the server I mentioned above to be
> production quality and I consider its IPv6 connectivity to be nearly as
> critical as its IPv4 connectivity.
Sure, we want production quality support for on-campus systems. But
is a pTLA required to provide tunnels to anyone in the network
neighborhood who asks? I'm not saying it would be a bad thing for us
to do, I just think we should make sure before commiting to doing it
with "production quality" support.
>> a) support staff of two (preferably three) persons registered in the
>> ipv6-site object info - Bill's the only one I saw in the registry
>> when I checked; I might be willing to be a second, if I can get up
>> to speed a little more on bgp4+ peering etc
>
> I'd be happy to register as a contact.
That'd make three... Though I'm not sure if it'd be a good idea if we
can't get you access to the v6 router in W20, and I don't think that's
my decision to make.
>> b) common mailbox for support contact; I'm not sure sipbv6 is best
>> for both that and local MIT IPv6 interest
>
> I'd tend to agree, since I don't have any affiliation with sipb. On the
> other hand, if it works and we can get mail, then why change it?
The list currently includes interested parties with Athena accounts,
not that we have to limit it to Athena. Some are maintaining tunnels
on or off campus (though some tunnel owners are not on the list), some
are actually working on our w20 ipv6 router, some may not be using
ipv6 but are just interested in following local discussion. Not all
of them are likely to be interested in email from outsiders about the
w20 router.
I have no idea if it would be a problem. But we could set up an
address like "mit-6bone-contact" that points to sipbv6 for now, and
change it later if needed.
>> 3) be a major provider of Internet service in a region, country, or
>> focus of interest; does "MIT community and groupies" count as a
>> significant "focus of interest"?
>
> Again, having seen some of the other pTLA allocations, I don't think
> this is a problem.
Okay, good.
That still leaves us with the question of the accuracy of the registry
information. I'm trying to look it up now, but the state of 6bone
whois servers seems pretty lame.
I finally found these, querying for MIT-SIPB at whois.6bone.net:
inet6num: 3FFE:1CE1::/48
netname: MIT-SIPB
descr: site delegation for MIT-SIPB site
....
ipv6-site: MIT-SIPB
origin: AS3
descr: Student Information Processing Board
Massachusetts Institute of Technology
Cambridge, MA
prefix: 3FFE:1CE1::/48
prefix: 3FFE:28FF:4::/48
....
inet6num: 3FFE:1CE1::/32
netname: MIT-SIPB
descr: pNLA delegation for MIT-SIPB
....
So if they're picky about meeting requirement (1)(a) for three months,
we may have a problem.
>> So what's the benefit of getting a pTLA over our current arrangement?
> Well, we'd have our own provider-independant address space, right?
Thus side-stepping the whole problem of doing routing better by making
everyone advertise our address block? Sounds like more of our IPv4
situation -- owning a way bigger chunk of the address space than we
need and making everyone advertise it, because we can.
> I've
> heard that MERIT is (or already has) discontinued their formal 6bone
> project, and that some peers are having a really hard time getting any
> help from them when stuff breaks.
Hm, that sounds like a problem...
