21:00:25 <timburke> #startmeeting swift
21:00:27 <openstack> Meeting started Wed Oct 16 21:00:25 2019 UTC and is due to finish in 60 minutes.  The chair is timburke. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:28 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:30 <openstack> The meeting name has been set to 'swift'
21:00:35 <timburke> who's here for the swift meeting?
21:01:08 <kota_> hello
21:01:11 <rledisez> o/
21:02:10 <tdasilva> o/
21:02:11 <timburke> agenda's at https://wiki.openstack.org/wiki/Meetings/Swift
21:02:37 <alecuyer> o/
21:02:40 <timburke> i saw a few interesting things this week, so let's jump right in
21:02:51 <timburke> #topic shanghai
21:03:10 <timburke> it's coming up so quickly! less than three weeks away!
21:03:17 <timburke> etherpad's at https://etherpad.openstack.org/p/swift-ptg-shanghai
21:03:23 <timburke> #link https://etherpad.openstack.org/p/swift-ptg-shanghai
21:03:40 <timburke> and i need to start putting slides together :-)
21:04:25 <timburke> mainly just wanted to do a quick reminder about the etherpad again
21:04:28 <timburke> #topic train release
21:04:45 <timburke> the openstack train release is officially out!
21:05:14 <timburke> #link http://lists.openstack.org/pipermail/openstack-discuss/2019-October/010156.html
21:05:34 <kota_> good
21:06:05 <timburke> a few last-minute py3 bugs were identified in swift, but zaitcev, tdasilva, and clayg were really fast to review them
21:06:17 <timburke> and backports are already landed on stable/train
21:06:43 <timburke> i've got a changelog update at https://review.opendev.org/#/c/688848/
21:06:43 <patchbot> patch 688848 - swift (stable/train) - CHANGELOG for 2.23.1 - 1 patch set
21:06:58 <timburke> and i'll aim to have a 2.23.1 release later this week
21:07:57 <timburke> bug fixes were p 688767, p 688766, p 688769, p 688768, and p 688808
21:07:58 <patchbot> https://review.opendev.org/#/c/688767/ - swift (stable/train) - Fix kms_keymaster under Python 3 (MERGED) - 1 patch set
21:07:59 <patchbot> https://review.opendev.org/#/c/688766/ - swift (stable/train) - py3: Fix swift-drive-audit (MERGED) - 1 patch set
21:08:01 <patchbot> https://review.opendev.org/#/c/688769/ - swift (stable/train) - py3: Fix swift-recon (MERGED) - 1 patch set
21:08:03 <patchbot> https://review.opendev.org/#/c/688768/ - swift (stable/train) - py3: fix swift-dispersion-populate (MERGED) - 1 patch set
21:08:05 <patchbot> https://review.opendev.org/#/c/688808/ - swift (stable/train) - py3: fix swift-account-audit (MERGED) - 1 patch set
21:08:21 <tdasilva> interesting that p 688848 builds a docker image, if it gets promoted in docker hub, i might be table to tag with a release 2.23.1
21:08:21 <patchbot> https://review.opendev.org/#/c/688848/ - swift (stable/train) - CHANGELOG for 2.23.1 - 1 patch set
21:09:01 <timburke> speaking of releases and images...
21:09:13 <timburke> #topic centos8 images are available
21:09:17 <timburke> #link http://lists.openstack.org/pipermail/openstack-discuss/2019-October/010146.html
21:09:35 <tdasilva> timburke: can we also backport the build py3 docker images patch?
21:10:05 <timburke> tdasilva, i think that seems reasonable
21:10:27 <timburke> centos8 is mostly just an fyi. i know we use centos7 for our probe tests, maybe i'll finally sort out how to make probe tests run in the gate on py3
21:10:53 <timburke> more release news...
21:11:02 <timburke> #topic py38
21:11:17 <timburke> if you missed it, python 3.8 is out!
21:11:20 <timburke> #link https://docs.python.org/3/whatsnew/3.8.html
21:11:52 <timburke> i've already started trying it out (since i've still got py3 freshly loaded into my head)
21:13:05 <timburke> a few bugs shook out, like the arg to pickle.load() needing to implement readinto() and the cgi module losing some functions
21:13:44 <timburke> but on the whole, seems pretty tame! not sure when it'll be available for gate jobs, though
21:14:11 <timburke> and on the topic of cpython...
21:14:25 <timburke> #topic dropping support for py2 within openstack
21:14:34 <timburke> #link http://lists.openstack.org/pipermail/openstack-discuss/2019-October/010142.html
21:15:17 <alecuyer> Oh I saw that.. but some of your py3 patches are still pending right?
21:15:24 <timburke> discussions have started regarding when and how and with what sort of communication projects can start dropping py2 support
21:16:08 <timburke> alecuyer, on 3.8? yeah, https://review.opendev.org/#/c/688591/ landed
21:16:08 <patchbot> patch 688591 - swift - py38: Provide readinto() interface for RingReader (MERGED) - 1 patch set
21:16:15 <timburke> but https://review.opendev.org/#/c/688594/ is still outstanding
21:16:16 <patchbot> patch 688594 - swift - py38: cgi lost some names - 1 patch set
21:16:57 <timburke> i'm kinda tempted to wait a little and see what six does with https://github.com/benjaminp/six/pull/290 / https://github.com/benjaminp/six/issues/159
21:18:34 <timburke> on the py2 front, of course, we already talked about wanting to continue support for a while, in part to continue supporting anyone with xenial's distro python packages
21:20:05 <timburke> i'll continue to push for keeping py2 testing viable, but it's something to be aware of as the ussuri cycle unfolds
21:20:31 <timburke> those are all the general announcements i had
21:20:56 <timburke> does anyone have any questions or comments on them?
21:21:59 <tdasilva> timburke: thanks for keeping us updated
21:22:49 <timburke> 👍
21:23:48 <timburke> i know i got a lot of value out of having notmyname curate and highlight the bits of the mailing list that were actually relevant when i got burned out on trying to keep up a couple years back :-)
21:23:57 <timburke> on to updates!
21:24:05 <timburke> #topic versioning
21:24:39 <timburke> tdasilva, how are you feeling about https://review.opendev.org/#/c/682382/ ?
21:24:40 <patchbot> patch 682382 - swift - WIP: New Object Versioning mode - 13 patch sets
21:25:10 <tdasilva> making progress...
21:25:23 <tdasilva> I need to start adding a whole ton of tests around listings
21:25:49 <tdasilva> and of course rebase the patch on top of the null-character patch
21:27:41 <timburke> nice! i'm working on getting https://review.opendev.org/#/c/673682/ rebased on the new swift versioning -- got like 5 of the 8 tests clayg wrote in test/functional/s3api/test_versioning.py passing at the moment
21:27:41 <patchbot> patch 673682 - swift - s3api: Implement versioning status API - 3 patch sets
21:28:28 <tdasilva> timburke: cool, please feel free to push any new patchsets to p 682382
21:28:29 <patchbot> https://review.opendev.org/#/c/682382/ - swift - WIP: New Object Versioning mode - 13 patch sets
21:28:56 <timburke> will do. i was thinking about adding an is_latest key when doing ?versions listings...
21:29:46 <timburke> tdasilva, anything else people should be aware of, thinking about, or looking at?
21:30:45 <tdasilva> no, but i'm sure there will be churn during code review, so the sooner that starts the better :)
21:31:00 <timburke> #topic lots of small files
21:31:08 <tdasilva> and of course I've been putting off documentation so far, maybe I should start working on that on a separate patch
21:31:23 <timburke> alecuyer, thanks for fixing up the requirements so https://review.opendev.org/#/c/686846/ could land!
21:31:23 <patchbot> patch 686846 - swift (feature/losf) - EC frags are the ones with the # (MERGED) - 2 patch sets
21:31:40 <alecuyer> np, thanks for your help with it and others
21:31:48 <alecuyer> I have pushed this small one here https://review.opendev.org/#/c/659254/
21:31:49 <patchbot> patch 659254 - swift (feature/losf) - In _get_vfile(), handle the case when a referenced... - 2 patch sets
21:32:32 <alecuyer> and I need to work on properly detecting and exposing the case where the leveldb database is corrupted
21:32:58 <alecuyer> (as I mentionned, we currently check this from logs, outside of swift itself)
21:33:27 <alecuyer> and then we should see if that would have to be monitored by the operator or self-heal
21:34:59 <timburke> right... fwiw, needing to fix up swift-drive-audit made me realize we already have some stand-alone tools for parsing kernel logs and such...
21:35:27 <timburke> self-heal would be nice, but i think operator-driven would probably be a good initial release
21:35:31 <alecuyer> If you think there's a good example there to look at , let me know :)
21:35:44 <alecuyer> ok
21:36:40 <timburke> idk -- the fact that we didn't notice that it wasn't py3-ready tells me we don't really use it much, at least in our dev workflows... so might not be the best example ;-)
21:37:07 <alecuyer> right well I'll do something and we can discuss it
21:37:21 <timburke> 👍
21:38:30 <timburke> all right... with mattoliverau not here, i'm thinking we could probably skip talking about sharding...
21:38:55 <timburke> unless someone wants to pick up looking at https://review.opendev.org/#/c/675014/
21:38:56 <patchbot> patch 675014 - swift - Latch shard-stat reporting - 4 patch sets
21:39:31 <timburke> there's at least one significant issue i need to address, but i think the core of the thing is in pretty good shape
21:40:32 <timburke> #topic open discussion
21:40:43 <timburke> anything else anyone would like to bring up?
21:43:36 <timburke> kota_, how was the typhoon? did you and your family have to evacuate?
21:44:03 <kota_> it was safe on my home including itself.
21:44:19 <timburke> good to hear
21:44:28 <kota_> some part of tokyo affected but here not.
21:44:57 <kota_> thx
21:45:17 <timburke> all right, i suppose we ought to let you go get breakfast :-)
21:45:29 <timburke> thank you all for coming, and thank you for working on swift!
21:45:35 <timburke> #endmeeting