> We just submitted a new draft "Experience and Evaluation of the Distributed
> Node Consensus
> Protocol".
[...]
> Htmlized: https://tools.ietf.org/html/draft-jin-homenet-dncp-experience-00
Thanks, that's interesting.
TL;DR version: on lossless links with 0.5ms delay, with 80 nodes, DNCP
converges within 15s in a high-diameter topology, within 18s in a worst-
case dense topology, and within 2.6s in a best-case topology. Which is
pretty cool.
Would you consider simulating lossy links? I'd be particularly interested
in topologies where the lossy links are redundant, e.g.
B
/ \
A...C
\ /
D
which is something that's likely to happen in home networks, and where
DNCP is likely to work fine. If possible, you should assume lower loss
rates for unicast than for multicast.
Thanks again,
-- Juliusz
_______________________________________________
homenet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/homenet
