21:00:30 #startmeeting swift 21:00:31 Meeting started Wed Jul 18 21:00:30 2018 UTC and is due to finish in 60 minutes. The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:00:32 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:00:35 The meeting name has been set to 'swift' 21:00:38 who's here for the swift team meeting? 21:00:47 o/ 21:01:08 hi o/ 21:01:21 o/ 21:01:48 wecome 21:02:21 agenda this week is pretty much the same as last, so it should go quickly, I think 21:02:23 #link https://wiki.openstack.org/wiki/Meetings/Swift 21:02:30 hi 21:02:48 https://review.openstack.org/#/c/427911/ and https://review.openstack.org/#/c/447129 are two patches I'm watching to hope they land soon 21:02:49 patch 427911 - swift - Replace MIME with PUT+POST for EC and Encryption 21:02:50 patch 447129 - swift - Configure diskfile per storage policy 21:03:51 zaitcev: hello 21:04:01 I'd just mentioned https://review.openstack.org/#/c/427911/ 21:04:02 patch 427911 - swift - Replace MIME with PUT+POST for EC and Encryption 21:04:14 how's that patch going? do you need any help? 21:04:26 I need to rebase it, because its (only) dependency was merged, thanks to Matt. 21:04:34 Otherwise it's perfect 21:05:00 cool 21:05:11 that sounds like pretty good news 21:05:32 Alistair didn't like the dot-in-path. I never understood his argumentation against it... Possibly it injured his sense of the proper design. He has an alternative, based on Content-Type. 21:06:09 yeah, he and I talked about those options. or maybe just in the etherpad 21:06:16 https://etherpad.openstack.org/p/swift-put-post 21:06:18 That patch looks okay to me. A previous version had a fatal flaw: it wasn't future-proof. He promised to fix that, I need to verify. 21:07:04 ok 21:07:23 https://review.openstack.org/582298 21:07:23 patch 582298 - swift - PUT+POST: Detect older object server by not sendin... 21:07:58 notmyname: you can break our tie as our leader! 21:08:00 There's a lot to get into my head there, I'll try and read up on the patches and give em a test and review 21:08:18 same here 21:08:25 mattoliverau: tdasilva: thanks 21:08:53 Oh come on. It's like 7% complexity level of EC. 21:09:00 Or maybe 4% 21:09:05 Lol 21:09:52 anything else to bring up on that patch functionality? 21:11:04 on https://review.openstack.org/#/c/447129, IIRC clayg said he was going to look at it, but I know he's been busy with other stuff 21:11:04 patch 447129 - swift - Configure diskfile per storage policy 21:11:17 m_kazuhiro: could you review that patch please? 21:11:43 ok 21:11:49 thank you 21:12:27 mattoliverau: you'll be pushing a small update to https://review.openstack.org/#/c/541058/, right? 21:12:28 patch 541058 - swift - Add fallocate_reserve to account and container ser... 21:12:34 and then landing it? 21:13:20 Yup, pushed up one with a comment last night 21:13:27 oh, ocol 21:13:51 but were you going to +2/+A? 21:13:55 Was waiting for Gerrit 21:13:59 ah 21:14:08 Will do it after breakfast :) 21:14:13 kk 21:14:23 looks like a probe test failure. not sure if real or not 21:14:30 well 21:14:54 I noticed we're getting a ton of those probe tests failures recently, especially sharder related. 21:15:00 hmm 21:15:07 mm..clayg pointed me to that failure last week on another patch 21:15:12 At first I thought I induced them, but trunk is the same 21:15:52 ok, so something else to track down 21:16:17 I made a note 21:17:02 as a reminder, I'll propose an authors/changelog patch for python-swiftclient. it needs to be tagged next week 21:17:15 if there's anything you want in swiftclient, please work with someone to get it landed this week 21:17:29 (all the current open stuff that was ready to land has already landed) 21:17:31 Maybe we need a section on priority reviews about bugs and investigations. There are a few sharing ones I'd like to look into when I have the time. 21:17:46 Ie somewhere to put the note 21:17:53 mattoliverau: good idea. I had a couple of bugs listed at the bottom 21:17:56 but yeah, I'll do that 21:18:10 In the times before PUT+POST, I pooh-pooed probe tests. But during PUT+POST there was a time when we had functests passing, but probe tests broke and it totally was my fault. After that I realized that functests aren't good enough. BTW, not in-process. Even real functests. 21:18:10 much better than in my own todo list :-) 21:18:55 tests are great! 21:19:22 anything else to bring up in today's meeting? 21:21:16 all right. let's end the meeting then 21:21:21 thanks for your work on swift! 21:21:27 #endmeeting