21:00:21 #startmeeting swift 21:00:22 Meeting started Wed Jun 12 21:00:21 2019 UTC and is due to finish in 60 minutes. The chair is timburke. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:00:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:00:25 The meeting name has been set to 'swift' 21:00:38 who's here for the swift meeting? 21:00:45 o/ 21:00:57 o/ 21:03:01 i know rledisez and mattoliverau won't be coming 21:03:11 i'm here 21:03:17 lot of good it'll do 21:03:19 agenda's at 21:03:22 #link https://wiki.openstack.org/wiki/Meetings/Swift 21:03:51 #topic Train cycle goals 21:04:20 the TC finalized some goals for the cycle 21:04:47 this is mostly just an FYI; there shouldn't be too much work we need to do for them 21:05:00 #link https://governance.openstack.org/tc/goals/train/ipv6-support-and-testing.html 21:05:34 we've already got an ipv6 dsvm job; i think it's just a matter of ensuring that it's compliant with what they're looking for 21:05:35 py3 or bust 21:05:46 #link https://governance.openstack.org/tc/goals/train/pdf-doc-generation.html 21:05:51 py3 ipv6 opendev is the future! 21:06:48 it sounds like the docs team has already done a lot of work to make it easy for us; should just be a matter of tweaking our docs job 21:06:58 and finally 21:07:01 #link https://governance.openstack.org/tc/goals/train/python3-updates.html 21:07:17 which explicitly says that someone else will be proposing the patches ;-) 21:07:24 thanks coreycb! 21:07:50 good news there is that we already target py37 21:08:49 who has questions about the goals? 21:10:26 who can spare some time to make sure we're compliant with what the TC wants and push up some patches if we're lacking? 21:11:41 whoooooo.. @timburke throwing down the guantlet 21:11:44 is it hurry? anyway, I should go to look at the decision because of storlets... 21:12:30 no big hurry, i think. i can track them for now 21:13:18 i just also know that there's a bunch of stuff we want to get done *for swift* and if we can get this squared away sooner, then i won't need to think about it any more :-) 21:13:35 next up... 21:13:41 #topic Shanghai PTG 21:14:13 diablo_rojo sent out http://lists.openstack.org/pipermail/openstack-discuss/2019-June/007079.html with a rough breakdown of the schedule 21:15:04 basically, Monday - Tuesday for summit, Monday - Tuesday morning for forum, Tuesday afternoon - Friday for PTG 21:15:48 and the ptg page was up to date, https://www.openstack.org/ptg/ 21:16:21 project onboarding is getting rolled into the PTG portion, which i think is a good move 21:16:35 kota_: easier flight for you this go 'round? Have you been to Shanghai before? 21:16:40 #link https://www.openstack.org/ptg/ 21:16:52 (i'm actually kinda excited about using the PTG as a kind of extended onboarding session) 21:17:02 thanks, kota_ 21:17:33 clayg: this is my first time to Shanghai but I believe it's easier than U.S. and EU 21:17:56 it sounds like space is going to be tighter than at prior PTGs 21:18:12 #link https://app.eventxtra.link/registrations/6640a923-98d7-44c7-a623-1e2c9132b402?locale=en/?aff=Shanghaih 21:18:26 that is registration. I don't know what's VIP access tho 21:18:38 I'm glad you're excited about that timburke :) 21:18:42 I am too 21:18:50 so i wanted to remind people of it so i could start trying to figure out what kind of head count we expect 21:19:17 VIP includes a VIP party if I recall correctly? I guess that's a typical thing in China. 21:19:56 if you already know you will or won't go, that's great! if not, please start thinking about that :-) 21:20:16 OIC. thnks diablo_rojo_phon 21:20:24 again, no particular deadline yet, but i wanted to put it out there sooner rather than later 21:20:53 does anyone know for sure one way or the other yet? i'll be going 21:21:28 No problem kota_ :) 21:21:42 We'd like to make it with Romain, but I don't know that we have official approval yet. I'll see if we can get an answer before Romain is off for a few weeks :) 21:22:14 again, that's fine. good to know that you're hoping to go :-) 21:23:50 all right, let's move on to updates 21:23:58 #topic storyboard migration 21:24:32 last week, i asked that you guys play around in the storyboard sandbox, form some opinions 21:24:36 how'd that go? 21:25:25 my Firefox didn't like it 21:25:32 and then I forgot 21:25:38 i got a cert issue with the browser, is that expected? 21:25:46 after raging on the social media for 10 seconds 21:25:56 mine is chrome. 21:26:15 yeah we know the cert is not happy. 21:26:18 It's fine. 21:26:26 i see 21:26:26 Yeah, it's a self-signed cert. They used to do it in 2001. 21:26:36 It's on our todo list 21:26:40 I thought in 2019 everyone would just use Let's Encrypt 21:27:12 zaitcev: we'd love the help if that's a thing you could get going. 21:27:21 ;) 21:27:23 😍 21:27:43 diablo_rojo_phon: we're better at the raging on the socme than the helping ;) 21:27:51 word 21:28:18 still you know... it raises ~awareness~ 21:29:10 well, there's further feedback, i guess ... submit a story for it 21:29:14 #link https://storyboard.openstack.org/#!/story/list?status=active&project_group_id=57 21:29:26 clayg: yes it's definitely easier to complain ;) I'm quite good at it myself. 21:29:34 Last time I had to use Let's Encrypt, I forgot that I must provide the fullchain in place of the cert, because they have an intermediary, apparently. Took me hours. The error was less than self-explanatory... 21:30:00 #topic py3 21:30:17 i saw zaitcev did some reviews last night, thanks! 21:30:30 that's all I'm doing anymore, you're too good at coding 21:30:56 I started looking at this memcache thing and already there's 2 reviews to fix it 21:31:45 yeah... sorry, i kept finding new and different ways it was broken 21:31:49 I want to get a full conversion and then package RPMs 21:31:53 then functests 21:32:27 but the good news is, i think we're about ready to land tempurl after all those memcached patches! 21:33:02 at least, y'know, once https://review.opendev.org/#/c/664459/ lands to fix the gate 21:33:02 So, we only have tempauth (might as well collapse the middleware dir in tox.ini), reconstructor, small things like locale, and then general issues like memcached. 21:33:13 er, tempurl 21:33:47 zaitcev, it actually collapses all of common -- https://review.opendev.org/#/c/652928/6/tox.ini 21:33:50 :-) 21:33:51 I'll be mostly on PTO this week though. 21:34:29 and reconstructor should be ready -- https://review.opendev.org/#/c/652819/ just has the lower-constraints failure 21:35:14 who can take on test_locale? 21:35:24 I think we should make Train even if we find on-disk incompatibility. But we need to start testing and find it, if it exists. 21:35:30 I can, I'm sure. 21:35:43 excellent, thanks zaitcev! 21:35:56 ok, gtg 21:36:05 thanks for coming 21:36:59 so, i think we should be able to have unit tests ported by next week. i want to do a release as soon as i can after that 21:37:33 in fact, i think i'll start working on authors/changelog while someone reviews reconstructor :-) 21:38:05 #topic lots of small files 21:38:17 alecuyer, how's it going? 21:39:04 I had to work on another project at OVH so not much news from me. I have started today to work on comments you made (thank you and kota for the reviews!) 21:39:41 and, before writing more tests, it should work with py3. I tried that and it doesn't work, one reason being that I can't get an "old enough" version of grpc (to work with eventlet) 21:40:08 so my goal next week is to have a patch to use HTTP instead. (I have one but it's not complete) 21:40:11 oic. that would be problem 21:41:11 HTTP FTW! 21:41:27 Yes it will work :) 21:41:42 and the gate for losf got an error at lower-constraint, you know timburke? 21:41:55 :-/ i don't quite like that py3 would become a blocker... but i know we were hoping to get it to use HTTP regardless 21:42:10 kota_, yeah, that sounds right :-( 21:42:32 not sure that patch to update the chriptography version would resolve tho. 21:42:36 it's the cryptography problem. i'm trying to get it through the gate on master, then we'll propose a merge to losf 21:42:37 you think I should move on with the tests and address py3 later ? 21:42:45 oh. hmm... 21:43:42 timburke: ok. I'll keep to watch it. probably we want to port it feature/losf sooner as possible 21:43:45 alecuyer, i'm not sure. but i trust your judgment 21:45:44 would we collectively be OK with having it be a py2-only feature at first? 21:45:51 i just don't want you to have to go down a rabbit hole when you could be doing more interesting/better things 21:47:26 kota_, fwiw all the lower-constraints failures i've seen on losf look just like the ones from master 21:47:49 timburke: sounds good 21:48:51 well, py2-only-ness is something for us to think about. it might also be that moving from gRPC to HTTP would be a good way to get one of us who hasn't been looking at losf much involved 21:49:08 alecuyer, thanks for the update! 21:49:14 #topic open discussion 21:49:30 who has something else to bring up? 21:52:00 all right, i think that'll do then 21:52:11 thank you all for coming, and thank you for working on swift! 21:52:15 #endmeeting