mSQL Attack/Peering/OBGP/Optical exchange

Stephen Stuart stuart at tech.org
Fri Jan 31 23:13:26 UTC 2003


> 
> > We don't know anything we could do with 50ms provisioning without
> > making a disaster (c) smd 2001.
> 
> indeed.  but i sure would like one or two day provisioning, as
> opposed to 18 months.

The space where that problem exists is within and and at the edge of
carrier networks. 

I think we would agree that provisioning of capacity directly between
providers under the same modern-day exchange point roof with good
old-fashioned direct cross-connects can happen with timeframes similar
to or better than the providers' ability to next-day-ship an interface
card to take delivery of the upgraded capacity.

Stephen
VP, Eng.
PAIX



More information about the NANOG mailing list