21:00:50 <timburke__> #startmeeting swift
21:00:51 <openstack> Meeting started Wed Mar 10 21:00:50 2021 UTC and is due to finish in 60 minutes.  The chair is timburke__. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:52 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:54 <openstack> The meeting name has been set to 'swift'
21:01:00 <timburke__> who's here for the swift meeting?
21:01:04 <mattoliverau> o/
21:01:12 <rledisez> hi o/
21:01:41 <seongsoocho> o/
21:02:34 <acoles> o/
21:03:04 <timburke__> as usual, the agenda's at https://wiki.openstack.org/wiki/Meetings/Swift
21:03:10 <timburke__> first up
21:03:16 <timburke__> #topic next PTG
21:03:27 <timburke__> #link http://lists.openstack.org/pipermail/openstack-discuss/2021-March/020778.html
21:03:55 <timburke__> i probably should have brought this up last week, but i've gotten behind on the ML
21:04:15 <timburke__> it's going to be another virtual PTG, April 19-23
21:04:35 <timburke__> register (for free) at https://april2021-ptg.eventbrite.com/
21:05:08 <timburke__> i'll get a doodle poll up for next week to pick meeting times
21:05:21 <acoles> I already have my ticket :)
21:05:29 <timburke__> me too
21:05:54 <timburke__> we'll see if we can get notmyname to drop in again :-)
21:06:17 <mattoliverau> Oh I should go get mine.. I assume my employer will let me go
21:06:36 <timburke__> heh "go"
21:06:56 <timburke__> but yeah, pretty sure your boss will be on board ;-)
21:07:02 <timburke__> next up
21:07:07 <timburke__> #topic elections
21:07:26 <timburke__> (again, probably should have brought this up last week)
21:07:40 <timburke__> the nomination period is over!
21:07:42 <timburke__> #link http://lists.openstack.org/pipermail/openstack-discuss/2021-March/020973.html
21:08:45 <timburke__> looks like there are currently 4 candidates for 5 slots on the TC, and no actual elections will need to be held :-/
21:08:59 <timburke__> #link https://governance.openstack.org/election/
21:10:04 <timburke__> i don't really have much else to say about it, but wanted to keep people informed
21:10:25 <timburke__> so on to Swift work!
21:10:38 <timburke__> #topic CORS and s3api
21:11:16 <timburke__> now that the CORS test suite landed, i'd like to keep pushing on the actual behavioral changes i'd like
21:11:33 <mattoliverau> +1
21:11:52 <timburke__> specifically, https://review.opendev.org/c/openstack/swift/+/710330/ (which has s3api translate more cors-related headers)
21:12:14 <acoles> oh yes, I got distracted from that, I'll try to re-review
21:12:15 <timburke__> and https://review.opendev.org/c/openstack/swift/+/710355/ (which allows preflight OPTIONS requests)
21:12:51 <acoles> I've noticed a the zuul cors test has failed on a few patches :(
21:13:12 <timburke__> thanks! fwiw, it looks like there's at least a couple more changes i ought to make to the gate job given the failure on that last patch, but i'm fairly certain tests will pass with the recheck
21:13:37 <timburke__> acoles, did they seem to be timeouts during setup? that's what i saw on one, anyway...
21:13:55 <acoles> TBH I didn't dig into the failure too much
21:14:06 <timburke__> (though it got a little buried by a post-failure when we tried to copy some non-existent outputs)
21:14:23 <timburke__> i'll be sure to dig in for next week
21:14:31 <acoles> other than they weren't 'real' failures - the tests didn't seem to have run
21:15:11 <acoles> ok, or maybe they ran but the output wasn't posted. IDK
21:15:57 <timburke__> as long as the tests actually run, i expect we won't see a post failure, fwiw
21:16:06 <timburke__> next up
21:16:15 <timburke__> #topic relinker
21:16:32 <timburke__> i know we've got a few different threads we're pulling at now
21:16:52 <timburke__> mattoliverau has a few patches for improved logging
21:17:57 <timburke__> acoles has been working on some refactoring that should make cleanup behave a lot more like relinking (and they both should get a lot more tolerant of things like reapable tombstones)
21:18:06 <mattoliverau> but these will need to be rebased on top of acoles changes once they stablise. But mostly effect the post_partition hook for the moment, so mostly seperate.
21:18:42 <timburke__> do we have a preference for order of landing?
21:18:43 <acoles> today I wrote --policy and --partition option patches
21:19:03 <mattoliverau> oh cool, I'll check em out.
21:19:16 <acoles> I'd really appreciate this one merging (subject to review of course) https://review.opendev.org/c/openstack/swift/+/778530
21:19:17 <timburke__> nice, i should do some reviews :-)
21:20:11 <acoles> p 778530 is blocking me doing some stuff on the refactor chain that I'd like
21:20:21 <timburke__> ...also, i should dust off the priority reviews wiki
21:21:01 <timburke__> acoles, i'll try to make sure it's got a review by the time you wake up
21:21:16 <acoles> thanks!
21:21:18 <mattoliverau> which the amount of churn als code does on the main functions, it'll be easier to land that and then mine. Because mine will just need to be plugged into the partitals for the post_part hook.
21:21:41 <timburke__> sounds like a plan
21:21:45 <seongsoocho> There are no timestamp in relink log in my env. How to add a timestamp in log?
21:22:14 <acoles> mattoliverau: yours and mine may not conflict, but IIRC some tests were failing on your logging patch?
21:23:09 <mattoliverau> yeah, I'll poke it, I reworked it a bit yesterday and pushed stuff up at the end of my day.. I shoud've run tests locally first. But will check it out today.
21:23:29 <acoles> mattoliverau: sorry, it looks like it may have just been the cors test failing
21:23:45 <mattoliverau> \o/ (well kinda yay)
21:24:21 <timburke__> seongsoocho, that sounds like a good logging enhancement to make. fwiw, it should "just" be a matter of updating format='%(message)s' in relinker.py
21:24:48 <timburke__> would you mind writing up a bug (or even a patch!) for it, so we won't forget about it?
21:25:27 <seongsoocho> timburke__:  Ok I will writing up a bug and also write a patch too!
21:25:33 <timburke__> thanks!
21:25:35 <mattoliverau> nice!
21:26:34 <acoles> mattoliverau: so don't wait on me, if https://review.opendev.org/c/openstack/swift/+/779439 is ready lets try to land it then I'll take care of any conflict (now I looked harder there will be some but nothing too tricky)
21:27:27 <mattoliverau> nah it isn't. Like a said it should mostly be passing some extra params to the post_part hook partial call. Which you've just centralised, so actually makes it eaiser :)
21:28:10 <acoles> and TBH I'd rather we merged smaller changes rather than rush through review of the bigger refactor
21:28:39 <acoles> but yeah, the partial call does get easier :)
21:29:08 <timburke__> ok, one last topic
21:29:12 <timburke__> #topic releases
21:29:29 <mattoliverau> True. We'll put them in "order" on the priorities page then. Doesn't matter which order, but if you don't mind a conflict then happy to get mine in first, I just want to finish the 2nd recon one (needs some tests)
21:30:10 <timburke__> still need to get a swift release out -- anyone think they'll have a chance to proofread the changelog at https://review.opendev.org/c/openstack/swift/+/777708 ? make sure i'm not forgetting anything?
21:30:41 <mattoliverau> I'll take a look today
21:30:44 <timburke__> and separately, is anyone aware of any reason we should hold off on getting a release out? any critical bugs?
21:30:59 <acoles> I'll take a look tomorrow
21:31:13 <mattoliverau> maybe relinker improvements?
21:31:26 <mattoliverau> but so long as all the races are fixed and landed then that's good
21:31:31 <mattoliverau> *enough
21:32:12 <timburke__> i guess my main thought on the relinker is whether we've got anything that's demonstrably worse than it was in 2.26.0
21:32:15 <acoles> timburke__: maybe this https://review.opendev.org/c/openstack/swift/+/779308 ?
21:32:25 <mattoliverau> But who relinks very oftern, that could wait until next release :)
21:33:14 <timburke__> i'll take a look
21:33:24 <timburke__> that's all i've got
21:33:30 <timburke__> #topic open discussion
21:33:42 <timburke__> anything else we ought to discuss this week?
21:35:34 <acoles> just an FYI that mattoliverau and I have some patches to improve sharder & swift-manage-shard-range config handling, and we're proposing to change a couple of the defaults that relate to shrinking
21:35:41 <acoles> https://review.opendev.org/c/openstack/swift/+/774584
21:36:03 <acoles> and https://review.opendev.org/c/openstack/swift/+/778989
21:36:08 <mattoliverau> oh yeah! good call to mention it :)
21:36:38 <acoles> unless you use auto-sharding it should not have any impact
21:37:22 <acoles> or 'swift-manage-shard-ranges compact'
21:41:48 <timburke__> all right
21:42:09 <timburke__> sounds like we've got a pretty good idea of what we want to look at next
21:42:21 <timburke__> thank you all for coming, and thank you for working on swift!
21:42:25 <timburke__> #endmeeting