BBN Peering issues

Lawrence A. Deleski lad at inficad.com
Mon Aug 17 17:01:20 UTC 1998


>   What you're saying above directly contradicts what your VP of Engineering
>told me at the time we were considering moving ftp.cdrom.com to Exodus. At
>that time they had no peering with Sprint and were paying for route-reflector
>service from good.net (which was unacceptable to me). The solution was to buy
>DS3 circuits from Sprint - it was definately a for-pay peering arrangement.
>After more than 6 months, the circuits eventually got turned up, but I was
>concerned that the increased cost of for-pay peering with Sprint would
>ultimately be passed on to the customers (us), and I didn't see this as
>a scalable solution.
>   So, either something changed recently, you weren't informed as to the true
>nature of the Sprint agreement, or you are lying. Which is it?
>
>-DG
>
>David Greenman
>Co-founder/Principal Architect, The FreeBSD Project
>

Since you don't let on as to when you were talking to BV Jagadeesh about
this, I can't really say.  My guess is that was some time ago since the
arrangement you describe isn't familiar to me.  My information is current
and true, to the best of my knowledge.  I did disclaim earlier that I do not
work there any longer.  Whatever the facts are, I'm certainly not lying,
since I'm saying only those things of which I have direct knowledge.

It has always been the policy of Exodus to not buy transit.  That dosen't
mean they have resorted to it in the past.  When I first arrived there they
were getting routes to UUNet via a deal with CRL.  Since they now have a PX
with UUNet, so they no longer have to do that.  I guess the bottom line is
you either get peering with all of the places you want your traffic to go or
you resort to other measures, temporarily, to get the job done.

I do know the true nature of the Sprint agreement, but those details are
private to Exodus and Sprint.  If you want the details either BV or Rob will
have to tell you.

Lawrence A. Deleski
lad at inficad.com



More information about the NANOG mailing list