search

ietf-wg-ccwg / rfc5033bis

IETF drafts
Other
4 stars 6 forks source link

A final piece of text on "how" to do this #130

Closed gorryfair closed 4 months ago

gorryfair commented 4 months ago

This responds to an AD question in review. The text is proposed to shed some light on how to go about evaluating a new CC algorithm, and when to use different approaches. It could be enough, or it could be improved, please think about how best to say this.

zaheduzzaman commented 4 months ago

Thanks for this.. this looks like very good expression of the expectations on how one can think of evaluating the candidate congestion control algorithms without going into lengthy details.

For the sake of clarity, it seems like the word "overload" (in the quoted sentence below) needs some explanation on what exactly we mean here - I am assuming we are referring to typical congestion scenarios on the path due to packet inflow and out flow mismatch in a bottleneck network node.

"For many algorithms, an initial evaluation will consider individual protocol mechanisms in a simulator to analyse their stability and safety across a wide range of conditions, including overload".

gorryfair commented 4 months ago

overload, aka a lot more traffic than the path can handle... rather than typical congestion. Martin may have suggested wording...