<<< Chronological Index >>>    <<< Thread Index >>>

Re: [Fwd: [ripe-ttraffic #23440] tt11, N/A & no GPS signal]


Tom,

I think that there is a slight misunderstanding or loss of knowledge
here.  Let me try to solve this:

> - we are not happy about the service fee that is to be 
> paid by a test box host as announced at RIPE-37:
> 
> >Another change is that you will have to pay for the 
> >hardware of the boxes. A test-box currently
> > costs 2750 Euro. This includes the installation costs, 
> >but does not include shipping, taxes
> > and import/export duties. The RIPE NCC has been asked 
> >to introduce a service fee (of
> > approximately 3000 Euro/Year) for the operation of the 
> >boxes. However, for tax-reasons, it is
> > currently not possible to do this. Therefore: 
> >
> >      There will be NO service fee for the remainder of 
> >2000 and the first months of 2001. 
> 
> We are interested to support the test traffic project 
> and would be happy to keep hosting that box, maybe even 
> buy the hardware - but we are not willing to pay for 
> hosting that service. We provide a professional 
> computing environment and normally charge our customers 
> for housing equipment at our premises - not the other 
> way round ;-) 

> So how can we proceed?

The test-box provides information about delays and packet losses between
the ECRC PoP and all other sites hosting a test-box.  This is information
that we think can be useful for your network operations.  It can also be
used to show to your customers that ECRC offers them a good connection to
the global Internet.

Of course, this information is only useful if there is a reasonable number
of test-boxes in the world _and_ it has been shown that the boxes can
reliably do their measurements.

When we first started to offer this service, back in 1998, this wasn't the
case: there were only 2 test-boxes at the NCC site and we had no
experience running a large number of test-boxes on a 24/7 basis.

For this reason, we suggested to the RIPE NCC membership that RIPE NCC
would install test-boxes at sites that voluntered to host one and pay the
costs from the general RIPE NCC budget (that is, all LIR's who are a
member of the RIPE NCC).  ECRC (in the persons of Stephan Sticht and (if I
remember correctly) Waltraud Eber, I'm also aware that both of them have
left ECRC since then) was amongst the volunteers.

After about 2 years, at the RIPE NCC General meeting in October 1999, the
membership felt that TTM had now proven to be a service that could
reliably be offered to all interested members.  At the same time, they
felt that the costs for both hardware and operations should be shared
amongst the sites hosting the boxes.  The RIPE NCC was asked to come up
with a proposal for offering TTM as a regular service to all interested
parties, using a model where the costs are shared amongst the participants
similar to the model used for LIR services.

A first proposal was presented at RIPE 35 (Jan'00), discussed again at
RIPE36, resulting in the final proposal at RIPE37.  The draft that
eventually became RIPE-214 was updated several times in between and
circulated on the mailing list. The contacts at the sites hosting the
boxes (=Stephan for ECRC) were always informed about our plans and
explicitly asked if they, as the first users of the service, agreed with
them.

Since May or June, the draft contains a paragraph on the existing sites:
As selling existing hardware will cause major problems (in particular
outside the EU), we decided to give away the existing hardware to all
sites currently hosting a box.  At the same time, we proposed that we'd
start charging the service fee to the existing sites in 2001. 

What is not in the document (for "political reasons") is that the NCC has
some freedom to adjust the service fee for sites that are interesting to
the project but that did not budget the service fee for 2001 but have been
on the project since the start.  ECRC is such a site, though we would have
to discuss the details.

In the meantime, we discovered that the NCC, in it's present structure,
cannot charge the service fee at all.  This issue is currently being
addressed by the NCC accounts and lawyers, together with the Dutch tax
authorities, there are several solutions which will all affect the current
proposal one way or another. My guess is that it will be well into 2001
before this issue is resolved.

For the time being, we therefor won't charge any service fee and I don't
think that it make much sense to discuss exceptions before the regular
case has been settled.

So, to conclude, I suggest that, if you are still interested in
performance measurements between your site and the rest of the world, you
work with Fotis to re-install the test-box.  We can then discuss the issue
of the service fee sometime next year and work out an agreement there.

It is probably worth noticing that we will soon offer an upgrade path
(against hardware costs) for the existing clock hardware.  The new clocks
do not require coax cable between antenna and PC but use regular CAT-5
instead. We'd like to get a bit of running experience with this new
hardware first, but I expect to announce this in December.  We're also
planning to start a discussion on how to make the TTM service more useful
for your customers on the mailing list soon.

I hope this answers your question, let me know if you need more
information,

Henk

------------------------------------------------------------------------------
Henk Uijterwaal                    Email: henk.uijterwaal@ripe.net
RIPE Network Coordination Centre     WWW: http://www.ripe.net/home/henk
Singel 258                         Phone: +31.20.535-4414,  Fax -4445
1016 AB Amsterdam                   Home: +31.20.4195305
The Netherlands                   Mobile: +31.6.55861746  
------------------------------------------------------------------------------

A man can take a train and never reach his destination.
                                               (Kerouac, well before RFC2780).









<<< Chronological Index >>>    <<< Thread Index >>>