Re: Plenary minutes RIPE27

[Steven Bakker <]

==> On Fri, 19 Sep 1997,
==> "RNMR" == RIPE NCC Meeting Registration wrote:

  RNMR> Please accept my apologies for sending these out so late.

Please accept my apologies for replying so late :-)

Couple of comments:

  RNMR> 9. Technical Reports
  RNMR> - TEN-34 (Steven Bakker)
  RNMR> Impromptu presentation by Steven Bakker.

I believe that's putting it friendly :-)

  RNMR> A map of the current TEN-34 network was shown.
  RNMR> Many countries now connected, network is operational and works.
  RNMR> see http://www.dante.org/ for more information

That should be:

	http://www.dante.org.uk/ten-34
    or
	http://www.dante.net/ten-34

  RNMR> plans to start a new project done by smaller subset of
  RNMR> National Research Networks (NRN's)

  RNMR> Questions:
  RNMR> Peter Galvaby: what are the peering policies?
  RNMR> Steven Bakker: basically no policies, no real peering.
  RNMR> Must be careful to coordinate with European Commission on
  RNMR> peering agreements.
  RNMR> European Commission has restrictions on usage of network.
  RNMR> AT&T Unisource gives them transit as part of IP service they offer.
  RNMR> BT also gives them transit. Many of networks themselves have
  RNMR> bi-lateralarrangements with other networks. In discussion
  RNMR> too peer with EBONE.

Latest news on that is that EBONE withdrew from the discussions.  They are 
no longer interested in direct peerings with TEN-34, since they get the 
routes through their AT&T-Unisource peerings.

  RNMR> This will replace the European network
  RNMR> 2 remaining "customers" in Czech Republic and Hungary.

  RNMR> How many overhead from ATM?
  RNMR> Steven Bakker: about 17%

  RNMR> Marten Terpstra: Will the network actually reach 34Mbps in all countries
  RNMR> before the project will finish (2 year project)
  RNMR> Mike Norris: they reached "10" :-)

Depends on whether you count access speed or backbone speed.  Most of the 
backbone is >20M.  Access speeds vary, but the majority is >20M.  Given 
the cost of international high-speed circuits (see below) and the fact 
that the current bandwidth is not exhausted yet, upgrading for the sake of 
upgrading is not very wise.

  RNMR> Steven explained main problem is to do with prices/availability of
  RNMR> 34Mbps
  RNMR> in each country. In some cases this doesn't make sense since ACONEt is
  RNMR> hooked up to DFN who have to share this with other networks...

I'd like to qualify that statement: the problem is not with price/
availability of _national_ circuits, but _international_ circuits out of
certain (many) countries.  Also, there is a cost distribution model within
TEN-34, so the cost of _backbone_ (i.e. shared) lines is shared amongst all
the participants. I though I did mention this, but given the short
preparation time I had it may not have been a very clear story.

  RNMR> Rob Blokzijl: Too little, too late. Took 2 year organisation to get a
  RNMR> network going for a year..what about a follow-up? Organising too late.
  RNMR> commercial business is moving ahead with the development of bandwith
  RNMR> in Europe before R&D organisations are.. Rob says there are
  RNMR> plans for 622Mbps to come into AMS-IX

My reply to that was that there is currently no pan-European high speed 
network on the scale of TEN-34.  Sure, there are high speed islands all 
over Europe, many NRNs have a high speed national infrastructure.  The 
problem is getting a *pan-European* high-speed network.  In that respect,
I believe TEN-34 is a first.

Cheers,
Steven