[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [RRG] Are we solving the wrong problem?



Mark,

In broad terms, I agree with your assessment and architectural proposal.

One specific comment re the following:

On 19 Feb 2008, at 16:55, Mark Handley wrote:
The specific reason for layer 4 is that you need to stripe data across
multiple *congestion controlled* paths for the traffic to
self-traffic-engineer. Only then do the bytes to be transfered
actually move away from the congested paths to the uncongested paths
in a stable manner.  I don't believe you can get this unless the
congestion control part of the stack is aware of the different
paths/addresses and load-balancing across them.

A few years ago I waved my hands and claimed that one should take a fresh look at congestion control, basically dividing it into a path- specific and session-specific parts. The former might best be located at the current layer 3, somewhere where shim6 and hip live today, while the other probably belongs closer to the application, e.g. "upper part" of layer 4.

Pekka Nikander, "ACCCoRD: Architectural Considerations on the placement of Congestion Control Routines and Data -- Extended Abstract," in Proceedings of High Speed Networking Workshop, Budabest, Hungary, May 17-18, 2004, Budabest University of Technology and Economics.
http://www.tml.tkk.fi/~pnr/publications/hsn2004.pdf

Unfortunately, I haven't really done any real work in the area. So, I may be grossly mistaken. And there definitely are some interesting "cross-layers" issues involved.

--Pekka Nikander


--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg