Re: ISPACs

"Justin W. Newton" <justin@erols.com> Fri, 06 December 1996 03:49 UTC

Received: from cnri by ietf.org id aa17918; 5 Dec 96 22:49 EST
Received: from nico.aarnet.edu.au by CNRI.Reston.VA.US id aa04409; 5 Dec 96 22:49 EST
Received: from smtp1.erols.com (smtp1.erols.com [205.252.116.101]) by nico.aarnet.edu.au (8.6.10/8.6.10) with SMTP id MAA23143 for <cidrd@iepg.org>; Fri, 6 Dec 1996 12:02:55 +1100
Received: from justin.erols.com (justin.erols.com [205.252.116.48]) by smtp1.erols.com (8.7.5/8.7.3) with SMTP id UAA29865; Thu, 5 Dec 1996 20:02:26 -0500 (EST)
Message-Id: <3.0b36.32.19961205200720.0139c8d4@justin.erols.com>
X-Sender: justin@justin.erols.com
X-Mailer: Windows Eudora Pro Version 3.0b36 (32)
Date: Thu, 05 Dec 1996 20:07:25 -0500
To: Tony Li <tli@jnx.com>
From: "Justin W. Newton" <justin@erols.com>
Subject: Re: ISPACs
Cc: cidrd@iepg.org
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"

At 01:59 PM 12/5/96 -0800, Tony Li wrote:
>
>   I am very capable of cooperation if there is a large benefit, what I
see in
>   your proposal, however, is a large risk with little benefit.  Maybe I am
>   missing something, but what do I, as a small ISP gain by doing this?  My
>   risks are farily apparent (at least some of them are).
>
>So in case it still isn't obvious: you get access to a larger address
>block, which is aggregated as a group.  You're not dependent on address
>space from an upstream provider.  You can provide address space to your
>customers which decreases their need to renumber (depending on the
>proliferation of the ISPAC).

Right, so instead of my customers having to renumber if I want to leave my
upstream provider, my customers have to renumber if I want to leave the
ISPAC I am in.  To use your words, "just like today".

>
>   1) I am dependant on competitors of mine of whom I may be somewhat unsure
>   as to their technical ability, and there is always the possibility that
>   they will act in an unscrupulous manner and /intentionally/ hose over my
>   announcements.  
>
>If you're that worried, then you may wish to pursue the common interconnect
>model rather than the mutual exchange of transit.  Depending on exact
>details, it would mean that you would only have to trust the folks running
>the interconnect.

Ok, so am I buying bandwidth from the people who run the interconnect?  I
see 2 possibilities here, as follows...

1) I am buying bandwidth from the person running an interconnect, in which
case they become my provider, "just like today".

2) Other members of the ISPAC's data cross my network to get to the ISPAC
interconnect, and my traffic transits their network to get to the
interconnect, effectively allowing my competitors to use my potentially
better connectivity, and I become dependant on their network to make
certain that my users connectivity isn't affected.  I believe that I
covered the reasons that people may be uncomfortable with that in detail
before.

>
>   The only defense I have against this is strong legal
>   reprecussions for failure to properly handle routing announcements and
>   sharing of traffic.
>
>Just like today...

Yes, and today I do not enter into agreements with my direct competitors
where They have the ability to destroy connectivity for all of my customers
to the entire internet, unless I am buying transi from them, and even then
it is a deal between my comapny and a single company who I selected, not
some whacked consortium where I may or may not have say as to who has the
ability to affect my traffic.  Yes, BGP peers /could/ do the same thing to
me, but its a lot easier for me to turn off a peering session than
instantaneously renumber my network.

>
>   2) If those strong legal terms /are/ in the contracts, I am suddenly open
>   to lawsuit from my competitors (one may substitute the word competitor
with
>   peer if they so choose).  
>
>Just like today... but without the contract.

Yes, but such a lawsuit would likely be considered frivolous.  Unless I was
maliciously attacking my competitor via ping floods or something of the
like, it would be quite difficult for them to launch a lawsuit that would
be considered other than frivilous in a US court.  If, however, their
traffic crosses my backbone, its is a lot easier for them to have /real
cause/ to be suing me.  What happens when they don't feel that I have
adequate generator power?  What happens when my routers have problems (and
this is something that has happened on every complex backbone I have ever
seen)?  Right now, this only impacts my customers, with whome I have an
agreement as to what damages they can collect in such an arrangement.  If
you (collectively as a group) can come up with a legal agreement which
both: 1) Adequately protects me from my competitor trying to screw me over
when I join an ISPAC of which they are a member, and 2) Protects me from
being sued by my competitor if I am making a good faith effort to deliver
their packets, you will have come a long way in making your proposal a
reality.  Until such a document exists we are wasting our time on this
proposal, noone will use it.  Tony, why don't you go propose this on
inet-access where all of the people who would actually become members of
these ISPACs actually are, instead of over here on cidrd where most of the
members already receive space directly from the regional registry of their
choice?

>
>   If a customer feels that I am providing
>   inadequate service they will likely choose to go to another ISP, which is
>   bad.  If a competitor feels that I am damaging their business through a
>   technical relationship that we have they are likely to take me to court.
>   Guess how expensive that can be?
>
>So based on this, I assume that you have no contracts or agreements with
>anyone.  Moreover, I have to assume that you're not in the US, where you
>can be sued for giving someone a cup of coffee.  ;-)

I have contracts and agreements with a lot of people, my upstream provider,
my bgp peers, my customers, etc etc.  In the case of my upstream, they lose
significant revenue if I leave, in the case of my BGP peers, I can turn off
the peering sessions very quickly which limits the amoiunt of damage they
can do to me.  And yes, I can get sued over a cup of coffee, but this does
not mean that I wish to make my exposure as great as possible, quite the
contrary, I try to limit my legal liability as much as possible.  


Justin Newton
Network Architect
Erol's Internet Services