16:59:11 <ahf> #startmeeting Network team meeting, 14 February 2022
16:59:11 <MeetBot> Meeting started Mon Feb 14 16:59:11 2022 UTC.  The chair is ahf. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:59:11 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:59:26 <ahf> hello hello
16:59:28 <mikeperry> o/
16:59:31 <Diziet> o/
16:59:31 <ahf> our pad is at https://pad.riseup.net/p/tor-netteam-2022.1-keep
17:00:35 <ahf> i think eta may wanted to join in too today, but was also gonna be afk for some of the day
17:00:36 <jnewsome> o/
17:00:41 <dgoulet> o/
17:00:41 <eta> o/
17:00:51 <ahf> o/
17:00:54 <ahf> very nice
17:01:16 <ahf> how are folks' boards looking: https://gitlab.torproject.org/groups/tpo/core/-/boards ?
17:02:58 * ahf very excited about the prop#332 changes in review
17:03:33 <eta> looks okay
17:04:23 <ahf> arti#20 is cute
17:04:33 <Diziet> haha :-)
17:04:39 <ahf> cool! i don't see anybody screaming about something bad here
17:04:56 <nickm> arg, lost track of time
17:04:57 <nickm> here now
17:05:13 <nickm> all's well
17:05:15 <ahf> dgoulet: any updates on release work on little-t-tor
17:05:19 <ahf> nickm: o/ no worries
17:05:46 <dgoulet> not much for now. There might be another 047 alpha soon considering prop324 might get merged soon
17:05:52 <dgoulet> that is about it
17:06:17 <ahf> very very nice. it looks like there is good progress on that
17:06:41 <Diziet> oh bum, just moved a card when I meant to get its url
17:07:12 <ahf> :-S clonky web ui's
17:07:24 <ahf> okay, doesn't look like there is anything from other teams, but
17:07:33 <ahf> dgoulet: we have tor#40559
17:07:36 * Diziet puts it back
17:07:56 <dgoulet> yup, I have few things in the pipe for health team
17:08:07 <ahf> i guess this is getting in a patch similar to that we usually do -- i don't know what the timeframe is there or what we want to do release wise, but i think you are just on it and i can ignore it, right?
17:08:19 <ahf> ok, perfect. i am gonna strip it from my list of things to watch out for. sounds good!
17:08:19 <dgoulet> yes
17:08:53 <ahf> i see no announcements. i see no discussions
17:09:07 <ahf> this week is the last week to revise any updates on the pad for january as they go out with end of december updates on the forum too
17:09:13 <ahf> mikeperry: wanna talk about s61 ?
17:09:31 <mikeperry> ahf: ah yes I moved my s61 january items out of december; they were accumulating there by mistake
17:09:48 <mikeperry> ahf: (I just did that this morning, noticing feburary was separated heh)
17:10:04 <ahf> ah, no worries. december was a pretty short list
17:10:12 <ahf> because people go afk sort of the middle
17:10:16 <ahf> not much exciting happens after that
17:10:31 <ahf> and we started late in january too
17:10:55 <nickm> qq for mikeperry: is anything blocked on me now for the CC review?
17:11:00 <mikeperry> so the main thing for s61 so far this month is review and fixups for the negotiation: https://gitlab.torproject.org/tpo/core/tor/-/merge_requests/525/
17:11:08 <mikeperry> there are some @'s for dgoulet in there
17:11:51 <mikeperry> nickm: nothing is blocked, but at some point if you could look at the fixups I've done and if they satisfy your comment, mark 'resolved'? that will help ua focus on the remaining things
17:12:19 <nickm> ok
17:12:26 <nickm> probably tomorrow
17:12:29 <nickm> if that's ok
17:12:51 <mikeperry> I am still going thru the remaining things; at this point mostly additional tests, filing tickets, etc
17:13:16 <mikeperry> but paring down the resolved things will help not miss anything
17:13:19 <mikeperry> nickm: sounds good
17:14:34 <mikeperry> jnewsome,hiro: how are things on the onion svc and utilization graphs for shadow?
17:15:19 <hiro> do we have data?
17:15:35 <hiro> I have been working on the CDFs last week
17:15:48 <jnewsome> mikeperry: being able to add onionservices is merged into tornettools, and I have a config checked into the sim repo for running with onion services. it partly works, though something like 25% of the transfers fail
17:16:09 <jnewsome> yes, there's a run from last week with them enabled
17:16:16 <mikeperry> hiro: oh you needed a sim with full descriptors, right? we lost a lot of artifacts last week.. I know jnewsome tried to re-run a baseline, not sure if it got lost too
17:16:43 <hiro> ok jnewsome could you add a link of that sim to the ticket?
17:16:56 <hiro> so I'll start with that
17:17:07 <jnewsome> i reran after artifactageddon
17:17:10 <jnewsome> hiro: sure
17:18:06 <hiro> thanks!
17:18:29 <jnewsome> hiro: which ticket is it again? seems to have disappeared from my "todo list" in gitlab
17:19:34 <hiro> ok let me check
17:19:56 <ahf> artifactageddon LOL
17:20:41 <hiro> https://gitlab.torproject.org/jnewsome/sponsor-61-sims/-/issues/9
17:21:00 <hiro> ah there is an output in nextcloud
17:21:04 <hiro> I could work with that
17:21:08 <hiro> I had missed that updated sorry
17:21:40 <jnewsome> hiro: thanks. that one was just a small test run, i'll add the full run
17:21:48 <hiro> thanks
17:23:14 <mikeperry> jnewsome: you got the extra 2 sim runners? I still think trying an 0.4.6 onion service run is worth it, to rule out the 0.4.7 vg-lite changes contributing to that failure rate
17:23:54 <jnewsome> mikeperry: yup, just brought them online. sure i can kick that off if you don't have anything else to run. otherwise I was going to try to repro and debug locally
17:24:01 <mikeperry> I probably will just be doing fixups today anyway
17:25:11 <mikeperry> it is a bit hard to juggle a bunch of fixups and simultaneously track dropping/adding patches for the cell queue test. too many patches flying around
17:26:27 <ahf> :-/
17:27:01 <jnewsome> mikeperry: ok; i'm a bit in the middle of debugging some shadow/signal/golang stuff, so I'll go ahead and kick that off to run while I'm still working on that :)
17:27:42 <mikeperry> yeah, great plan. hopefully that 0.4.6 datapoint will help narrow the search
17:27:53 <mikeperry> if 0.4.6 has the issue, then I would flag dgoulet to help look
17:28:06 <mikeperry> if it does not, it's time to dig into vg-lite, which I wrote half of
17:29:21 <mikeperry> geko is out
17:29:32 <mikeperry> juga: anything in sbws land for this week?
17:29:40 <juga> mikeperry: nope
17:29:52 <juga> working on #40125
17:30:16 <juga> so far so good
17:30:25 <mikeperry> juga: ok great
17:30:54 <mikeperry> re that control port slow start thing, I can make that patch as a separate ticket, but it should not be necessary to get something working
17:31:36 <juga> not sure about what is the control port slow start
17:31:55 <juga> anyway, i'll wait until tor branch is merged
17:31:56 <mikeperry> so long as you make room for some logic to discard measurements for the beginning of a transfer, either by bytes or checking for a CIRC_MINOR event
17:32:05 <mikeperry> ok
17:32:08 <juga> ah, now i get it, ok
17:32:24 <juga> i'll try the event, then see
17:32:48 <mikeperry> sbws listens for CIRC_MINOR already, or no?
17:32:53 <juga> no\
17:32:59 <mikeperry> oh hrmm ok
17:33:32 <juga> should it already listen?
17:33:52 <mikeperry> no. I am just musing where to put an event that says "circuit left slow start"
17:34:09 <juga> ok, i'll think about it
17:34:29 <mikeperry> ok
17:35:11 <mikeperry> I think that's all for s61, unless there's any questions?
17:35:19 * juga is good
17:36:39 <ahf> ok
17:36:42 <ahf> anything else?
17:37:08 * dgoulet is good
17:38:07 <nickm> fine over here
17:38:15 <ahf> excellent, let's wrap it up
17:38:18 <ahf> #endmeeting