[Iccrg] IETF plenary: fast timescales

[Tony Li]

On Aug 11, 2010, at 5:28 AM, ken carlberg wrote:

> and speaking of which, one tool you didn't mention is MPLS (aka, traffic engineering).  This is done in part because one can't over-provision to satisfy the sum capacity of all leaf links.  Up until about a year ago, this was becoming more acute with ever rising P2P traffic, motivating the recent actions from Comcast.  


Ken,

The effect of P2P within Comcast was to create point sources and sinks with effectively infinite capacity.  You can't provision to support these at all.  The correct approach is per-user queuing at the leaf.  

The point of MPLS/TE is that routing only gives large knobs and circuits only come in quantized units, so the ability to direct traffic around paths that are non-shortest paths allows an ISP to avoid localized hot spots.

Regards,
Tony