Torturous routes.

Majdi S. Abbas msa at samurai.sfo.dead-dog.com
Sat Apr 21 00:08:32 UTC 2001


On Fri, Apr 20, 2001 at 02:11:01PM -0700, Roeland Meyer wrote:
> ---Plays around on the InterNAP.
>  4  border3.fe5-3.speakeasy-9.sff.pnap.net (216.52.86.28)  20.485 ms  20.795
> ms  21.185 ms
>  5  core4.fe0-0-bbnet1.sff.pnap.net (216.52.80.6)  22.027 ms  20.601 ms
> 21.605 ms

	The thing about InterNAP's transit-only model is this:

	The (claimed) upside: "no congested peering points!"

	The downside: no peering points whatsoever.

	The advantages of going with peering locally or at least
going with a regional provider who does or at least has a large
number of customers (perhaps in addition to transit from the usual
suspects) is that you gain local reachability you will never see
otherwise; if you're one hop away from someone's transit network,
odds are you're being backhauled into the figurative middle of 
nowhere and your data still has to go to an interconnection point
elsewhere.

	I wouldn't bother talking to either Speakeasy or Qwest; 
Speakeasy looks like they're punting packets the only way they can,
to InterNAP, who forwards them on to Qwest.  You have been bitten
by the Network Without Peer (tm).  If you talk to Speakeasy again,
see if you can sell them transit.  I'd be interested to know why
they chose InterNAP -- DSL is typically inbound traffic, which is
something the ASsimilator has very little control over.  It seems
odd to choose a service designed for hosters, unless they're 
getting a really sweet deal.

	And hey, at least it's staying within the same state.
I have seen lots of situations where traffic going across the
street will cross the country at least once to get there.  Anyone
who's ever lived in a rural market has to deal with much, much
worse on a daily basis.

	--msa




More information about the NANOG mailing list