RE: recent slowdown in routing table growth

["Ayyasamy, Senthilkumar (UMKC-Student)" <saq66@umkc.edu>]

>  1) raised awareness: this exponentially-looking growth raised some
> alarms, there was a presentation in IETF51 in London (Aug 
> 2001), etc. --
> (some)  people fixed some of their systems.

> Mainly 1), but I'd also like to get some more concrete references.


http://www.potaroo.net/papers/bgp/bgp-june02.pdf
http://www.packetdesign.net/docs/cengiz.pdf

Major reasons for the dip in growth:
-  reduced number of specific entries due to awareness 
   based on CIDR report.
-  Use of prefix length route-filtering

But,as per geoff's presentation, multi-homing/TE will surely
be a major contributor in future.

 Also, see slide 10-12 in cengiz's presentation on the exact 
break down for the contributing factors in multi-homing.
It is good to hear that misconfigurations/multi-homing by means 
of MOAS is decreasing. Although, MOAS is an operational
technique, people do engineer them purposely/accidently.

One can check geoff huston BGP page for an up-to-date report.
Also, compare the above presentations with the previous IETF 
ptomaine/plenary presentations. 

I think, prefix filtering contributed the most to the present 
reduction in growth (refer randy's presentation-51th IETF)