CUSEEME and Europe

[David Farber <farber () central cis upenn edu>]

Date: Sat, 15 Oct 1994 22:10:43 -0400 (EDT)
From: danu () dccs upenn edu (Dan Updegrove)


Dave,


This list has had some interesting chat of late about the threat posed by
CU-SeeMe. (I know you're already on the list.) This message has some
especially interesting data about European long line costs and a
wonderful tag line by the sender.


Regards,
Dan




Forwarded message:
> From: Per Gregers Bilse <bilse () EU net>
> To: Multiple recipients of list <cu-seeme-l () cornell edu>
> Subject: Re: Borre's bandwidth concern ex Nordic sites
> Date: Sat, 15 Oct 1994 14:42:37 -0400
>
> I've been lurking on this list for a couple of weeks, and I'm quite
> happy to see that bandwidth consumption is being of concern to people.
> > From our (EUnet's) perspective, CU-SeeMe actually poses a serious
> threat to the quality of the services we can offer, and we have been
> considering various means of protecting ourselves and our customers.
>
> On Oct 14, 19:31, Mike Stanyer wrote:
> > I would argues from a slightly different perspective. It may well be
> correct
> > that for Nordic academic institutions that there are no plans to
> improve the
> > trans-Atlantic bandwidth. BUT is this true for fee paying access?
>
> Speaking as the person who evaluates, plans, implements, and controls
> the transatlantic bandwidth for -- no holds barred -- Europe's biggest
> commercial ISP, bandwidth is certainly going up and up.  We double
> every six months.  But ...
>
> >  In the UK this is certainly not the case and the commercial internet
> providers
> > are steaming ahead with their parrallel networks and their own funded
> > trans-Atalntic links. This may contain the germ of a bandwidth
> solution... if
> > these commercial internet access providers find that it is worth their
> while,
> > they would certainly find it commercially sensible to fund local reflectors
> > rather than invest ever increasing sums into long haul high bandiwdth comms
> > circuits. The network of reflectors solution seems the correct way to
> go but
> > quite what it will eventually look like is unclear to me.
>
> .. local reflectors and/or whatnot doesn't solve the situation
> where somebody with a Mac behind a 64k (or even 19.2 async) link goes
> to the US (or elsewhere in Europe, for that matter) and turns on what
> can only properly be described as a megabit-blast (which happens a
> couple of times every week).  Even if they had the local bandwidth to
> receive it, I don't think they would be prepared to pay for it.  In
> order to serve these people, we would need to have a separate 1-2Mbit
> transatlantic link, just for their video.  2Mbit transatlantic plus
> _one_ 2Mbit link across Europe costs on the order of USD
> 100000-150000 per month, just for the lines themselves; we'd need to
> have separate 1-2Mbit links across Europe to every single country we
> serve (over two dozen), simply to avoid CU-SeeMe tearing apart
> regular traffic (telnet, ftp, WWW, customer-specific applications,
> etc).  Obviously, this is not an issue open for discussion.  Were we
> to bill people for their usage, it would easily come to USD 50-100
> per hour.
>
> For the benefit of US-based readers of this list who find this too
> incredible, here are some examples of European vs US leased line
> prices:
>
> Route                      Cap          Monthly   Distance    Equiv.
>                         (Mbit/sec)      cost      (miles)     US Cost
>                                       (GBP x1000)           (GBP x1000)
>
> Brussels/Paris                2          22.87       170       3.06
> Brussels/Amsterdam            1          17.74        98       1.54
> Brussels/Luxembourg           1          17.7        117       1.58
> Brussels/London               2          31.53       211       2.12
>
> London/Amsterdam              2          32.12       230       2.17
> London/Frankfurt              2          42.64       400       2.56
> London/Geneva                 2          43.69       457       3.69
> London/Paris                  4          80.83       209       4.24
> London/Stockolm               4          75.44       908       7.48
>
> Paris/Luxemburg              1.5         21.8        180       1.96
> Paris/Geneva                  2          29.3        250       2.21
> Paris/Madrid                  1          29.35       649       2.59
>
> Frankfurt/Amsterdam           2          31.88       228       2.10
> Frankfurt/Geneva              2          34.29       287       2.3
> Frankfurt/Luxemburg          1.5         28.03       110       1.8
>
> Amsterdam/Luxemburg           1          17.51       196       1.73
>
> Geneva/Barcelona              1          34.56       395       2.36
> Geneva/Milan                  2          40.16       147       1.98
>
> Total                                   634.14                62.37
>
> GBP 1 = USD 1.35, give or take 10-20%.
>
> There's no question that CU-SeeMe is a great tool, great fun,
> and a wonderful step forward (although I've never used it myself :-).
> But careless use can have devastating effects on the networks
> used for the transport.  As mentioned, wee see this happening
> once or twice every week.
>
> My personal guess is that unless something is done to ensure more
> "socially responsible" use in the US, the problems we have in Europe
> now will become widespread in the US in 6-12 months time.  Rationale:
> At the current level of CU-SeeMe use European and transatlantic lines
> are ready for meltdown, while in the US (where lines typically are 10
> times bigger) one observes "busy day".  When CU-SeeMe use has reached
> 10 times its current level, one will in the US have to solve the same
> problems we in Europe have to solve over the coming 6-12 months.
>
> Imminent death of the Internet _not_ predicted, BTW. :-)
>
> --
> bilse <bilse () EU net> +31 20 592 5109 (dir: 5110);  fax +31 20 592 5163
>                 ``We used to ! but now we @'' (jensen)







--
Daniel A. Updegrove                           University of Pennsylvania
Associate Vice Provost                        3401 Walnut, Suite 221A
  Information Systems and Computing           Phila, PA 19104-6228
Executive Director                            215 898-2883
  Data Communications & Computing Service     fax 898-9348