Testing Eyeball Happiness
Note: This ballot was opened for revision 05 and is now closed.
(Ron Bonica) Yes
(Dan Romascanu) Yes
The Security Considerations section should mention the potential threat of overloading production networks if the conditions described in the last paragraph in Section 1 are not respected and advise care in using the test procedures in operational networks.
(Stewart Bryant) No Objection
(Gonzalo Camarillo) No Objection
(Wesley Eddy) No Objection
In the timing requirements for each metric, it's confusing whether the requirement is 0.1 ms error over 60 seconds, or 1ppm, since both are mentioned.
(Adrian Farrel) No Objection
(Stephen Farrell) No Objection
(1) 2.2 says to do "at least N" runs when Bob has N addresses, but isn't that a bit misleading when alice has M addresses as well given that you're interested in the min and max over (I guess) all combinations of the (M,N) pairs that can ever work, blocking all but one each time? Maybe characterising the number of iterations like that would be better? (2) I wasn't clear if these metrics are also intended to be used e.g. for a config with a middlebox between router1 and router2 that enables one of Alice's IPv6 addresses setup a TCP session with one Bob's IPv4 addresses. I'm not asking for anything in particular, I just wondered and didn't get an answer from the text. Editorial nit-like stuff: - Saying "installing" "Path MTU issues" is awkward at best and possibly misleading - maybe put the word "installing" in each of the bullets but the last, and say "creating path MTU issues" for the last bullet and merge the next paragraph into that bullet? - Section 2.2: the sentence "Different measurement trials revise..." is hard to understand and I think could just be omitted. - 3wHS is a bit cryptic in Figure 2 - if you just add the abbreviation to the text describing figure 2, that'd be ok, as in "three-way handshake (3wHS)", or maybe even make that paragraph be the caption for figure 2. - 2.3.1: Name == "Session Setup Interval" but then you say that "Session setup time" is measured in ms. Better to use the terminology consistently. Similarly the "units" text in 2.3.2 & 2.3.3 also seems wrong. Just saying "Units: milliseconds" would be fine. - Seems odd for the measurement point(s) text in 2.3.2 and 2.3.3 to refer to 2.3.1 but to then repeat the text for Timing. Same thing for 2.3.4 I guess. - I guess I don't understand what a "Descriptive Metric" might be, but I assume the intended readership would. I also don't see how it has "units" of ms if it requires text to be understood? If there's an RFC that defines this term then referencing that would be good.