[tsvwg] Re: [tsvwg] https://datatracker.ietf.org/doc/html/draft-ietf-tsvwg-nqb-23#name-limiting-packet-bursts-from

[Greg White <[e-mail]>]

To some degree it is a made-up number, but not an implausible one. The keyword is "minimum".  PON networks can have minimum link RTTs around 1ms (as can Low Latency DOCSIS).  So, a CPE device connected via a wired Eth link to an ONU could certainly have a minimum RTT of 2.5ms talking to devices on the public Internet (most likely the minimum would be seen when talking to an edge node).  I agree that this is an aggressive number, but some link technologies (notably Wi-Fi, but also DOCSIS) have historically been designed under the view that 2-5 ms of link discontinuity is perfectly fine, so providing an aggressive target isn't a bad thing IMO. 

-Greg


On 5/31/24, 10:28 AM, "Sebastian Moeller" <moeller0=[e-mail] <mailto:[e-mail]>> wrote:


Dear list,


the current NQB draft contains the following text:


This document does not set any quantified requirements for links to limit burst delay, primarily because link technologies are outside the remit of Diffserv specifications. However, it would not seem necessary to limit bursts lower than roughly 10% of the minimum base RTT expected in the typical deployment scenario (e.g., 250 us burst duration for links within the public Internet). This observation aligns with a similar one in Section 5.5 of [RFC9331].


If 250µs equals 10% then the full minimum base RTT within the public Internet would be (250*10)/1000 = 2.5 milliseconds.
Speaking from an end user perspective, that is not the minimum RTT to my ISP, let alone the 'internet'.
I now wonder, is my intuitive interpretation of what "minimum base RTT within the public Internet" means incorrect or is this simply a made up number?


Regards
Sebastian


P.S.: Essentially the same argument was made in RFC9331, but that dsoes not magically make this claim correct.