21:12:05 #startmeeting swift 21:12:06 Meeting started Wed May 18 21:12:05 2022 UTC and is due to finish in 60 minutes. The chair is timburke__. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:12:06 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:12:06 The meeting name has been set to 'swift' 21:12:10 how long has that been going for 21:12:14 lol 21:12:30 who's here for *this week's* swift meeting? 21:12:32 o/ 21:12:57 o/ 21:13:27 such a long meeting :D 21:13:38 as usual, the agenda's at 21:13:40 #link https://wiki.openstack.org/wiki/Meetings/Swift 21:14:09 I put in the first topic 21:14:20 and it looks like mattoliver added something to it :D 21:14:38 #topic tempurl and sha1 signatures 21:14:44 mattoliver, take it away :-) 21:14:49 Thanks 21:15:17 Got pinged in the channel during the week, there are still people interested in getting this landed. 21:15:23 Which makes sense :P 21:16:20 So just wanted to bring it to peoples attention. To get final reviews in and let's land at least the tempurl patch 21:17:10 i'll take a look for sure -- seems like there have been some changes since the last time i pushed anything up 21:17:49 Thanks, yeah, I think I pushed something up at some stage too. Can't remember :P I'm in the same boat. So I'll swing back to it myself. 21:18:19 People are waiting on it, so maybe we need to make it a priority. 21:18:33 That's all I have 21:19:14 then we'll check in on other work in progress :-) 21:19:28 #topic ring v2 21:19:52 thanks for the review and follow-up mattoliver! i haven't had a chance to loop back on it yet, i'm afraid 21:20:40 that's ok. I went through the initial patch at attempted to address all the missing todos in the follow up 21:20:58 mostly a bunch of testing. It isn't extensive but wanted to add covereage 21:21:49 i'll try to get a fresh patch up by next week so we can land it soon-ish 21:22:11 #topic memcache errors 21:22:30 the first patch in the chain has already landed! https://review.opendev.org/c/openstack/swift/+/839448 21:23:15 and i just approved the follow-up to make tempauth not hand out tokens that couldn't be set in memcache https://review.opendev.org/c/openstack/swift/+/840250 21:24:41 there was one other follow-up to get the command name in more log messages, but the plumbing is a little awkward https://review.opendev.org/c/openstack/swift/+/839822 21:25:50 we can live without it, but if anyone has some good ideas for a better way to carry some context around, feel free to push up a patch! 21:26:12 kk 21:26:48 #topic getting rid of logging locks 21:27:16 Wait, I could swear that I saw a memcache error propagation to client in proxy 21:27:25 Neither of the two linked above are it. 21:27:30 Must be dementia 21:28:08 Neither of the three 21:28:27 Oh well, logging locks. 21:29:20 zaitcev, what, like an uncaught error that made its way back out to eventlet, which formatted it for clients or something? i thought that was what catch_errors was for... 21:29:46 or another patch 21:30:01 timburke__, sorry. I must be confused about it. I thought I wanted to review that one and even saved it somewhere. But oh well. 21:30:26 This is what Joe Biden feels every day, I understand now. 21:32:41 funny, i've got this dim recollection as well... like you'd added yourself to the CC list, at least *shrug* 21:33:24 anyway, we've got https://review.opendev.org/c/openstack/swift/+/840232 in prod now, but still haven't turned it on. hopefully i'll actually have some data from prod next week? 21:34:21 i also think maybe mattoliver was going to try it out in our demo/staging environment 21:34:38 yeah, I plan too.. but keep getting distracted :( 21:34:53 Will try and get something done by next meeting (famous last words) :P 21:35:08 no worries! you're getting a lot done :-) 21:35:27 #topic backend ratelimiting 21:35:54 we've got the first couple patches in the chain landed! 21:36:02 there's the refactor https://review.opendev.org/c/openstack/swift/+/834960 21:36:33 and a new option for some burstiness https://review.opendev.org/c/openstack/swift/+/835122 21:37:45 next patch in the chain adds the new middleware itself https://review.opendev.org/c/openstack/swift/+/836046 21:38:31 That last one looks good. I pushed back to make sure it was added to the SAIO pipelines so we'd start testing it more. Maybe that's more of a NIT? 21:39:03 Just don't want to forget about it and not realise until later 21:40:28 there's the risk it could cause some flakey probe tests -- but then, we usually have issues with things going too *slow* in probe tests (and popping timeouts), not too *fast* 21:41:05 and we could also just make sure it's configured to some reasonably high default 21:41:25 it should be defaulted to off. but alleast it'll be there to turn on 21:41:28 but true 21:43:41 all right, that's all i've got 21:43:47 #topic open discussion 21:43:54 anything else we should bring up this week? 21:45:35 i bugged temoto for an eventlet release recently, and now upper-constraints has been updated to use it. this should mean the timeouts in unit tests should go away 21:45:59 Cool 21:46:33 nice 21:46:53 Digging through my dashboard just now during the meeting, I found that the conflict between object watcher and container sharding was not resolved. https://review.opendev.org/c/openstack/swift/+/787656 21:47:03 September 2021 21:47:44 Oh, december. Not as long as I thought... I'll try to find cycles to complete it, so it doesn't hang around. 21:48:29 that'll be great, thanks! sorry, i proposed the thing but then haven't really followed up on it 21:50:36 all right, i think i'll call it 21:50:48 thank you all for coming, and thank you for working on swift! 21:50:52 #endmeeting