[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: possible new official mirror in the US

karlheg@bittersweet.inetarena.com (Karl M. Hegbloom)  wrote on 08.02.98 in <87n2g2ks8b.fsf@bittersweet.inetarena.com>:

> >>>>> "Branden" == Branden Robinson <branden@purdue.edu> writes:
>     > I thought someone had made the (absolutely true) point that
>     > geographic proximity is only barely correlated with network
>     > proximity, if at all.
>  I've seen a traceroute to a university in the city I live in from my
>  house that actually routes through another city more than 50 miles
>  from here. :-)

I've seen the same going via the US. That's a little more than 50 miles.  
(That was before link 3 below was "official". It no longer does, but  
similar stuff can easily happen again.)

>  I guess if we re-wired the Internet, and straightened out all the
>  cables, there'd be enough wire left over to more than quadruple our
>  bandwidth.

I don't think it's the wire. Usually, it's the inter-provider agreements.

For a pretty strange example, look at this (it's actually simplified):

Münster                Dortmund

 | 1                     |
Nacamar               uni-dortmund.de
 | 2                     |

(Nacamar, Telekom, and DFN being the large providers involved in this  
scenario. And I have accounts with both westfalen.de and muenster.de, for  

1 and 2 is actual ethernet cable in one building.

And it's only used for news (news.uni-muenster.de <-> news.westfalen.de).

westfalen.de has IP addresses from the DFN. So has uni-muenster.de. But  
for political reasons, we have to connect via a link to Dortmund.

IP from muenster.de to westfalen.de goes to uni-muenster.de, then uni- 
dortmund.de, prima.de, and then westfalen.de.

All thanks to politics.

Isn't that great?

(Hi, Joey. For some strange reason, I still don't know what came out of  
the meeting with the Ruhries, and who is actually responsible for the DFN  
link in Dortmund, after all that talk. Another heap of politics, there ...  

MfG Kai

TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org . 
Trouble?  e-mail to templin@bucknell.edu .

Reply to: