16:59:16 <ahf> #startmeeting Network team meeting, 25th April 2022
16:59:16 <MeetBot> Meeting started Mon Apr 25 16:59:16 2022 UTC.  The chair is ahf. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:59:16 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:59:18 <ahf> hello hello
16:59:22 <nickm> hi all!
16:59:36 <ahf> pad is at https://pad.riseup.net/p/tor-netteam-2022.1-keep
16:59:47 <ahf> think david will be out today
17:00:09 <Diziet> o/
17:00:17 <mikeperry> o/
17:00:42 <ahf> how are folks' boards doing: https://gitlab.torproject.org/groups/tpo/core/-/boards ?
17:01:10 <ahf> remember to use the Q2 label for tickets that you plan on diving into this quarter 8)
17:01:24 <ahf> mostly important when taking stuff out of Backlog
17:01:57 <juga> o/
17:02:04 <jnewsome> o/
17:02:28 <ahf> o/
17:02:33 <ahf> can't see anything off in the boards
17:02:34 <ahf> bueno!
17:02:49 <ahf> gonna skip the release stuff since david is out, but the plan is to do 0.4.7 stable this week
17:03:18 <ahf> don't see anything incoming from other teams
17:03:36 <ahf> i have some stuff to talk with ACT about because we closed one of their tickets that they may need for one of their sponsors, but will chat with them about that
17:03:55 <ahf> no discussions or announcements
17:03:58 <ahf> mikeperry: wanna do s61?
17:04:07 <mikeperry> ok
17:04:45 <mikeperry> last week, we set the param for cc_alg=2, and started working with comms for the relay upgrade and release comms
17:05:11 <ahf> cool
17:05:34 <mikeperry> basically we have two pieces of comms: we need to get exit relays to upgrade first, so we will be doing a blog post and tor-relays post after we have debs, rpms, and BSD ports for the stable
17:05:58 <mikeperry> then there will be some outreach to help any stragglers upgrade
17:06:25 <mikeperry> after that, we will be put the new stable in Tor Browser stable series (looks like end of May), and do user-facing comms about that
17:07:17 <mikeperry> other than that, last week was mostly just our quarterly report
17:07:44 <mikeperry> I started poking at how to make test vectors, but did not get to writing them yet
17:08:27 <ahf> nice
17:08:47 <ahf> gonna be good to get this out
17:09:27 <mikeperry> yeah, it is a bit delicate because we don't want a situation where our users try this and are disappointed
17:09:56 <mikeperry> which means avoiding user-facing comms, press activity, or a Tor Browser stable until enough exits have upgraded
17:09:59 <ahf> right, the upgrade phase for the exit nodes is important here
17:10:08 * ahf nods
17:10:11 <mikeperry> but that will also give this time to bake in Tor Browser Alpha too, as exits upgrade
17:10:59 <ahf> ya
17:11:17 <mikeperry> juga: most likely we will want the sbws 1.5.0 w/ the CC param around the end of May
17:11:36 <mikeperry> juga: but you need to make sure no bandwidth auths upgrade to 0.4.7 until they have that release
17:11:43 <mikeperry> of sbws
17:11:46 <juga> mikeperry: ok
17:11:47 <mikeperry> (and onbasca)
17:12:09 <mikeperry> hiro: similarly, around then, we will want to upgrade our onionperf instances on metrics to 0.4.7
17:12:37 <mikeperry> but probably not before, or they will be noisy due to some congestion control activity but not all congestion control
17:14:11 <mikeperry> those are the main points here, I believe
17:14:16 <mikeperry> I think geko is out
17:14:31 <ahf> maybe he is watching the 24h moose tv show in sweden right now
17:14:47 <mikeperry> heh, is possible
17:14:48 <ahf> _maybe_
17:15:12 <mikeperry> oh yeah, I believe our s61 report is pretty much done
17:15:25 <mikeperry> thanks for the help there, geko, juga, hiro, gaba
17:15:32 <ahf> very nice
17:15:41 * ahf has been going over his s30 and s101 parts
17:15:53 <juga> mikeperry: thank *you*
17:16:03 <ahf> yes!
17:16:49 <mikeperry> np
17:16:55 <mikeperry> any questions here?
17:16:58 * ahf good
17:17:33 <juga> mikeperry: *all* bwauths shouuld use the scanner CC version when it's available?
17:17:56 <juga> or is it fine if some will update after end of May?
17:18:21 <mikeperry> juga: we will set that bwscanner_cc=1 sometime around end-of-may I am guessing. depending on Exit upgrade rate
17:18:45 <mikeperry> it is fine to upgrade to new sbws before then
17:18:55 <mikeperry> because we have not yet set bwscanner_cc=1
17:19:31 <mikeperry> but the key piece is that without that sbws upgrade, upgrading to 0.4.7 by bandwidth auths will cause random congestion control, depending on measurement exit
17:19:59 <mikeperry> does that make sense?
17:20:04 <juga> yes
17:20:24 <juga> let's see how deployment goes middle May
17:20:54 <mikeperry> ok
17:21:43 <ahf> nice
17:22:01 <ahf> anything else we have for today?
17:22:14 <nickm> i have a quick schedule note for tomorrow
17:22:35 <nickm> I'm doing the regular tor presentation for the MIT network security course tomorrow at 2:30 pm eastern
17:22:39 <Diziet> I will be afk on Monday.  (due to May Day)
17:22:49 <nickm> I'll be afk for that time, and for travel time earlier and later
17:23:03 <ahf> ah nice -- will it get recorded?
17:23:04 <nickm> I can still make the arti meeting at 1700 UTC
17:23:14 <nickm> ahf: I don't know; some years they do, some they don't.
17:23:28 <ahf> oki, please let us know if you end up with a recording
17:23:30 <ahf> but sounds good
17:23:33 <ahf> Diziet: ack!
17:24:00 <ahf> i will also be hungover on monday. sunday is first 1st of may in many years where i can go to the park and sing a weird mixture of labour songs and sailor songs
17:24:10 <ahf> ok, if we don't have anything else, i will close the bot
17:24:11 <Diziet> rotfl
17:24:32 <ahf> #endmeeting