21:00:18 #startmeeting swift 21:00:19 Meeting started Wed Mar 8 21:00:18 2017 UTC and is due to finish in 60 minutes. The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:00:20 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:00:23 The meeting name has been set to 'swift' 21:00:25 o/ 21:00:27 who's here for the swift team meeting? 21:00:29 o/ 21:00:32 o/ 21:00:37 hi 21:00:38 o/ 21:00:43 hi 21:00:45 o/ 21:00:48 good morning 21:00:49 hello 21:00:59 o/ 21:01:04 o/ 21:01:25 welcome everyone 21:02:23 * notmyname is wondering if acoles or clayg are around 21:02:39 just got here 21:02:57 all right, let's get started 21:03:01 agenda is at 21:03:02 #link https://wiki.openstack.org/wiki/Meetings/Swift 21:03:03 * joeljwright thinks clayg might **just might** be late ;) 21:03:13 joeljwright: shocking! 21:03:30 just a couple of things to go over, but i want to leave time for open discussion 21:03:32 well he was still online when I finished yesterday, so assume he has to sleep sometime ;P 21:03:40 fair point 21:03:44 mattoliverau: I haven't seen evidence of that yet 21:03:53 lol, true me either 21:04:03 ok, first up is an FYI 21:04:12 #topic outreachy internships 21:04:34 as you've probably seen over the years, openstack participates in teh outreachy internship program 21:04:55 https://wiki.openstack.org/wiki/Internship_ideas is a place where internship ideas are collected 21:05:11 it's where prospective interns go look to see if there's something that sounds interesting to them 21:05:18 if you've got ideas, please let's put them there 21:05:28 and,better yet, if you're interested in mentoring, do it! 21:05:43 both timburke and myself have done it before 21:06:07 I'd definitely recommend it as somethign that helps the mentor grow/learn as much as the intern 21:06:28 so please look that over and consider if you'd be interested in being a mentor 21:06:39 and if you have ideas on projects, please write them down too 21:06:40 agreed. i feel like i should have set more time aside to keep up on reviews though 21:07:05 if you have an idea, don't feel like you *must* be the mentor. or that you have to do it on your own 21:07:33 you've got the rest of the community to help 21:07:39 timburke: are you crazy man, even when mentoring your a reviewing machine ;) 21:08:07 mattoliverau: those are not allowed anymore 21:08:08 i meant keeping up on karenc's patches... 21:08:21 tdasilva: ha! 21:08:29 :-) 21:08:54 that's the FYI on internships/mentors. if there's no questions, we'll move on.... 21:08:56 you'll have to pry the keyboard out of clayg's cold, mechanical hands 21:09:11 tdasilva: lol 21:09:28 :D 21:09:28 #topic updated foundation logo 21:09:32 breaking news here! 21:09:49 about 30 seconds ago I got the final logo in my inbox 21:09:53 perfect timing 21:09:55 what.. is it correct? 21:10:00 yep 21:10:02 https://www.dropbox.com/sh/7gcsuslg0qfjqe9/AAApjLrQpissqd5W6JsL3gOta?dl=0 21:10:04 \o/ 21:10:05 there you go 21:10:06 no, it's a bat now?! 21:10:09 lol 21:10:15 it's actually a swift now! 21:10:22 oh wow! looks good! 21:10:25 not a swallow. not a martin 21:10:28 a swift! 21:10:58 nice 21:11:04 yeah it does. it's very different than the hand-drawn one we've had (and we can still like that old one and use it). but I like this new one too 21:11:17 its good! 21:11:53 ok, next topic 21:11:56 a bigger one 21:12:05 #topic upcoming boston summit/forum 21:12:15 It looks good! 21:12:31 last week I promised to look in to the boston summit and report back to help answer questions about what it's going to be like, who needs to go, etc 21:12:40 there's good news and bad news on this front 21:12:56 but first let me dump some links in here so you can do further reading later 21:12:58 :S 21:13:02 #link https://wiki.openstack.org/wiki/Forum 21:13:07 #link http://lists.openstack.org/pipermail/openstack-dev/2017-March/113459.html 21:13:15 #link http://lists.openstack.org/pipermail/openstack-dev/2017-March/113115.html 21:13:20 #link https://wiki.openstack.org/wiki/Forum/Boston2017 21:13:26 #link https://etherpad.openstack.org/p/BOS-Swift-brainstorming 21:13:31 /fin 21:13:35 just a few links then :P 21:13:50 ok, i'll try to summarize 21:14:13 the first 2 are more about what the summit will look like. the last 3 are related to planning it 21:14:23 here's what it will look like, as best I can make out 21:14:53 first, this is the event where there will be the formal talks/presentations, the expo hall, the marketing stuff, the big loud colorful parts 21:15:24 that much will be very familiar to anyone who's been to an openstack summit before 21:15:33 but here's the change 21:15:40 there's a new thing called "the forum" 21:15:55 this is the second part of the big summit/ptg split master plan 21:16:14 the ptg was what were called the design sessions at the summits previously 21:16:42 the forum is going to be, basicaly from what I can tell, the fishbowl sessions 21:17:03 these are time-limited 40 minute slots, scheduled, that are to discuss a particular topic 21:17:27 with a projector, and etherpad, and a room full of chairs arranged in circles 21:17:49 so all that sounds "normal", or close enough to what we've done in the past 21:18:12 the trick it, or where the difference comes in, is how these are scheduled and how they will be allocated 21:18:29 previously, each project got a few fishbowl slots and a bunch of design session slots 21:18:30 So they've taken the old design summits, split the meeting rooms and fish bowls and seperated them by about 3 months? 21:18:38 mattoliverau: yep 21:18:41 :( 21:19:06 which I guess makes sense in some cases, ops can give feed back from the "last" OpenStack release mid way through etc 21:19:10 so this time, the whole set of fishbowl sessions will be scheduled across all projects instead of allocating a few per project 21:19:18 mattoliverau: that's the other part 21:19:31 in previous summits, there's been an "ops track", normally on the first day 21:19:38 that's rolled into the forum too 21:20:07 so instead of "swift ops feedback" in a swift track in a fishbowl room, there might be an "openstack ops feedback" for everyone 21:20:18 (note the "might". it's hard to predict) 21:20:52 the way these will be scheduled is interesting, and if we want the forum to be productive for swift, we need to get involved now 21:21:04 using the etherpad linked above.. 21:21:08 https://etherpad.openstack.org/p/BOS-Swift-brainstorming 21:21:20 we need to collect fishbowl session topics 21:21:32 they need to have a clear scope and goal and, ideally, a designated moderator 21:22:04 so it might be something based on upgrades or hardware or replication cycle times or whatever we choose 21:22:31 then, the tc and the user committee will look at the topics that every project has chosen and select and schedule the forum sessions 21:22:57 from the email, it seems that the scheduling will likely start in the next few weeks 21:23:14 or the goal is to do scheduling about a month out from the forum 21:23:21 and the forum is about 9 weeks from now or so 21:23:25 right, yeah so we need to get in on the action there then. 21:23:31 so we need to be participating now 21:23:36 mattoliverau: exactly 21:23:56 notmyname: Did you also see the note about the onboarding meeting? 21:24:11 all this also brings up some more challenging questions. specifically, who should go 21:24:11 I remember them saying at some point that there might also be rooms available for projects to have a 'mid-cycle'. Would this be in addition (not formally organised of course) to these fish bowls? 21:24:50 jungleboyj: ah yes. a time for new contributors to get familiar with your project, btu I don't know full details of what that looks like 21:24:57 NOTE: remembering them saying.. was like 6 moths ago or something 21:25:07 mattoliverau: I haven't seen anything about that 21:25:11 kk 21:25:19 well, actually... 21:25:33 notmyname: Right, I am trying to find out more about that. It is supposed to be a logical extension of upstream university which I am also helping with. 21:25:35 it looks like there may be either some unscheduled space for ad-hoc stuff 21:25:49 or a room that's scheduled like the projector room at the PTG was 21:25:50 jungleboyj: nice 21:26:07 mattoliverau: I saw a note saying that they would also have tables set up for ad-hoc mid-cycle type stuff. 21:26:15 for those of you not familiar with how the ptg went, there was a room that could be reserved for a time slot 21:26:25 jungleboyj: mattoliverau: right 21:27:05 so who should go... 21:27:36 cool, so other then particaping in fish bowls that we important to us (as a project and cross openstack) we can also do some mid-cycling then.. albeit unstructured, but thats the way we roll ;) 21:27:43 I do know this: for any swift contributor who is there, we will definitely find a way to get together and talk about swift. that's just what we do when we're in the same city 21:28:03 :) 21:28:14 :-) 21:28:20 mattoliverau: yes, but probably in a limited sense. I'm not sure how much room we'll have. will it be like the lunch room with lots of space? something smaller? hard to know 21:28:55 I cannot promise that we'll have time to talk about stuff like we've done at previous summits 21:29:20 but worst case there is always the hallway track I assume 21:29:46 I do think the ops feedback will be great, but in my experience, when you've got 40 minutes to talk about some ops concern like "rolling upgrades" or "managing dependencies", swift hasn't gotten a lot of attention (it's just not the major pain point) 21:30:36 or said another way, I think it will look more similar to the first two days of the PTG than the last 3 21:30:37 On the bright side, the PTG was much better than expected. Hoping the Summit/Forum goes the same way. 21:30:43 jungleboyj: :-) 21:30:55 endless optimism :) 21:31:08 jungleboyj: +1 21:31:09 joeljwright: I have gotten good at it. 21:31:25 I'm not trying to be gloomy about it. I think that there's some great potential for the forum. it's just very uncertain because it's new 21:31:41 notmyname: ++ 21:32:11 notmyname: my positivity is just expressed differently :) 21:32:13 but *everything* I've heard about it has been much more "solve all-of-openstack problems" instead of "focus on your openstack project's problems" 21:32:44 so here's what we need to do: 21:32:57 first, add our topics that we're interested in to the etherpad 21:33:32 second, it would be good to figure out over the next few weeks who will be going and who won't (just to help inform what we'll be able to get done at the summit) 21:33:51 and finally, don't write it off before it actually happens :-) 21:34:26 so that's my report on the forum. I expect that there will be a lot of questions, if not now, then in the days to come. please feel free to ping me with them 21:34:28 sounds great. So we can use the etherpad for the first 2 things then? 21:34:48 mattoliverau: yes! 21:35:07 ie, ideas and a list of people, when they know they're coming can add their names 21:35:36 done. there's a section for that now 21:35:56 I'd encourage you to read over the info that's out there about the summit/forum. start with the links above 21:36:30 ok, next topic (and we can keep talking about this, too) is.... 21:36:34 #topic open discussion 21:36:36 open! 21:36:58 what's on your mind that needs to be brought up in this meeting? 21:37:41 i hate eventlet -> https://github.com/eventlet/eventlet/issues/210 21:37:50 nothing is every simple 21:38:07 :D 21:38:28 oh! I'll tell you my homework for the next meeting 21:39:07 I say "hate" - I mean I mad at it right now 21:39:09 next week, I plan on having some words written down in a linkable place that talks about priorities in swift, current problems, and what to work on, specifically related to the whole golang/hummingbird thing 21:39:38 notmyname: nice 21:39:43 clayg: :( 21:41:42 mattoliverau: I would have swore I'd done multiprocessing queues and eventlet green imports before - i'm not sure exactly what my problem is right now 21:41:46 not a lot of clamoring to bring up stuff... 21:42:14 my symptoms are ust it locks up and doesn't do what I expect it to - more so than any mutliprocessing libariaries throwing stack on EAGAIN 21:42:36 but like *generally* we don't do monkey patching in deamons - we just import green 21:43:33 clayg: you want someone else to jump into it with you? 21:44:36 idk, I was hoping rledisez would tell me "no need to write a patch that adds multiple workers to the reconstrcutor i already did it -> linky" 21:45:26 clayg: sorry, i’m just starting multiple processes, concurrency the old way… 21:46:08 clayg: there you go, take eventlet out of the equation :P 21:47:14 not a lot of talking here, so continue converations in -swift and close the meeting 21:47:28 thank you everyone for coming. thanks for working on swift! 21:47:31 #endmeeting