[PCN] Marked flow termination with "single-marking"

[Michael Menth <menth at informatik.uni-wuerzburg.de>]

Hi,

we recently finished a paper considering marked flow termination (MFT) 
based on marked AR-overload (excess marking based on admissible rates) 
using marking conversion. Our results show that MFT does not work with a 
small number of flows per ingress-egress aggregate and that it does not 
work with multipath routing. Thus, MFT looses its core benefits compared 
to measured rate termination (MRT). We are happy to receive feedback to 
improve the paper:
http://www3.informatik.uni-wuerzburg.de/staff/menth/Publications/papers/Menth08-Sub-10.pdf

Our conclusion is that excess marking based on the supportable rate is 
needed to cope with a small number of flows per IEA (1-4 flows) or with 
multipath routing. At least we do not know any mechanism that achieves it.

Kind regards,

    Michael

-- 
Dr. Michael Menth, Assistant Professor
University of Wuerzburg, Institute of Computer Science
Am Hubland, D-97074 Wuerzburg, Germany, room B206
phone: (+49)-931/888-6644, fax: (+49)-931/888-6632
mailto:menth at informatik.uni-wuerzburg.de
http://www3.informatik.uni-wuerzburg.de/research/ngn

_______________________________________________
PCN mailing list
PCN at ietf.org
https://www.ietf.org/mailman/listinfo/pcn