21:00:46 <notmyname> #startmeeting swift
21:00:47 <openstack> Meeting started Wed Dec 12 21:00:46 2018 UTC and is due to finish in 60 minutes.  The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:48 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:50 <openstack> The meeting name has been set to 'swift'
21:00:52 <notmyname> who's here for the swift team meeting?
21:01:02 <notmyname> I suspect we may be rather low attendance today
21:01:04 <tdasilva> hi
21:01:07 <rledisez> o/
21:01:13 <kota_> hi
21:01:29 <notmyname> tim and clay are both taking off of work today
21:02:04 <notmyname> I wonder if mattoliverau is around
21:02:10 <notmyname> #link https://wiki.openstack.org/wiki/Meetings/Swift
21:02:10 <kota_> oic
21:02:31 <notmyname> tdasilva: rledisez: kota_: ok, if it's just us, then let's get started :-)
21:02:40 <notmyname> I vote we assign all the work to people who aren't here ;-)
21:02:43 <tdasilva> notmyname: i think mattoliverau said he would not be around
21:02:50 <notmyname> tdasilva: ah ok
21:03:08 <kota_> mattoliverau left a message yesterday evening he might be absent because of wife's bday... birth day, maybe?
21:03:13 <rledisez> notmyname: yu take the good decisions, that's why you get my vote every time ;)
21:03:14 <notmyname> ah, ok
21:03:19 <notmyname> rledisez: lol
21:03:30 <notmyname> kota_: yeah, bday is a common typed abbreviation for birthday
21:03:44 <notmyname> the only real thing I wanted to bring up this week was a new swift release
21:03:55 <notmyname> I'd been waiting for the s3api patches to land, and they have!
21:04:03 <kota_> yey
21:04:13 <kota_> thx clayg!
21:04:18 <notmyname> yep
21:04:38 <notmyname> and I haven't seen other critical outstanding patches, so I think we should go ahead with a tag
21:04:55 <kota_> +1
21:04:58 <notmyname> I'm currently working on the authors changelog patch. I expect to have that proposed in a few hours (my afternoon today)
21:05:06 <notmyname> I'd appreciate reviews, as always
21:05:27 <notmyname> then as soon as that lands, I'll make the release tag request
21:06:06 <notmyname> timburke has identified several things that should be backported and made patches for them
21:06:34 <notmyname> I'll do backport release tags, too, after I get the 2.20.0 tag sorted
21:07:13 <notmyname> any questions about the release? anything that you know of that needs to land?
21:07:25 <notmyname> rledisez: I notices your ssync patch landed. that's good
21:07:54 <rledisez> yeah, didn't know what happen, i think zuul went crazy lask week
21:08:03 <notmyname> yeah, I think so
21:08:55 <notmyname> if you see something that should be in the release, please update the wiki and ping me on irc about it
21:09:09 <notmyname> zaitcev: how's py3 stuff going?
21:10:07 <zaitcev> notmyname: I was away for a week and it looks like Cyril didn't do anything, so no change. I was done with account, started on container and found that it needed object server to run tests, so I was going to do the object server next.
21:10:52 <zaitcev> We're at some 60 tests of 192 done (192 includes func and probe).
21:10:54 <notmyname> great!
21:11:40 <zaitcev> You think?
21:12:00 <zaitcev> We're in a critical red state here for Stein in RDO
21:12:26 <notmyname> I do. it's progress, and it's good to see both you and cyril contributing
21:12:44 <zaitcev> Actually, sorry - looks like Cyril filed https://review.openstack.org/#/c/624791/, but I didn't receive a notification.
21:12:45 <patchbot> patch 624791 - swift - Python3: fix test_xprofile.py - 1 patch set
21:13:09 <notmyname> earlier this week we had someone from canonical ask about it too. I said we could do a lot more if they would help us with it ;-)
21:13:43 <notmyname> but yeah, it seems both redhat and canonical both have deadlines that are earlier than the openstack tc deadline and the python deadline
21:14:25 <mattoliverau> Yeah sorry, I'm out at a cafe with the wife and kids for a birthday breakfast.
21:14:31 <notmyname> zaitcev: do you (and/or cyril) have a particular plan of attack? you said you're going to do the object server next. do you have a plan anywhere?
21:14:54 <notmyname> mattoliverau: go back to brekky with the family!
21:15:27 <zaitcev> We have a rule that says that "core" components must rely on components shipped with the main system, and RHEL 8 does not have py2 in that set. But OpenStack is a core product and Swift is a part of it. So... well, we can delay the whole Stein while I and Cyril are working on Swift.
21:15:32 <kota_> mattoliverau: o/
21:15:57 <zaitcev> Or I dunno. Maybe there's some bureaucratic and packaging trick that allows us to ship. I already started the ball rolling on that.
21:16:37 <zaitcev> notmyname: Sorry, there's no clear plan. The plan is to use tox.ini as a list of everything that works and never detract from it, keep adding randomly until it's all done.
21:16:45 <notmyname> ok
21:17:08 <notmyname> is your internal deadline based on the stein release? that's next april
21:17:19 <zaitcev> There's a lot of unexpected, like I just mentioned. Who knew that container needs object, right?
21:18:01 <zaitcev> yes, it's Stein. But we have an RC in RDO earlier. Also, 60/192 is only 32%. All-unit is 80% in case you're wondering.
21:18:45 <notmyname> I wonder if a more public tracking of ported components would help. eg a trello board or something. I feel like surely we tried that in the past
21:19:18 <zaitcev> In other words, I and Tim have completed about 40% (32/80) in 3 months. We're just barely making it by April, not counting what happens if we screw on-disk compatibility somewhere :-)
21:19:24 <notmyname> if your RC for RDO is before the stein release, then it can't include the stuff in the stein release, right?
21:19:51 <mattoliverau> notmyname: kk, and nice use of Aussie slang ;)
21:19:51 <zaitcev> No, wait. April is 4th month, not 3rd. So there's a bit of a reserve for integrated tests (probe e.g.)
21:19:52 <notmyname> I don't know how rdo does openstack releases
21:19:56 <notmyname> mattoliverau: thanks ;-)
21:20:53 <notmyname> kota_: rledisez: tdasilva: do you have any thoughts on this py3 topic?
21:20:59 <zaitcev> Everyone who's new suggests Trello, but then it turns out that it takes more work to update that Trello than fix modules, unless it's some kind of disaster like the bulk middleware.
21:21:04 <tdasilva> notmyname: you looking for a mapping between tests in tox.ini and components?
21:21:07 <zaitcev> So Trello never happens
21:21:26 <notmyname> zaitcev: eh, just something we've used in the past for other stuff in swift
21:22:07 <tdasilva> the tests list in tox.ini gives me a at a glance idea of where things are and what's not there
21:22:11 <tdasilva> IMO
21:22:17 <notmyname> tdasilva: sortof. more just wondering how to coordinate more people. eg if someone outside of redhat shows up to help with py3, then where do they start that doesn't interfere with the stuff redhat is doing?
21:22:30 <tdasilva> that we don't have
21:23:22 <tdasilva> but we don't have the people either i'm afraid, unless canonical has agreed to help or unless you are thinking of next time they show up you point them to the list....
21:23:37 <notmyname> tdasilva: yeah. more the latter, I'm afraid
21:25:05 <tdasilva> maybe a wiki page off of ideas page?
21:25:24 <zaitcev> tdasilva: if someone shows up, just tell look at py3 at Gerrit, look at tox.ini, then select any module they like out of remaining
21:25:46 <zaitcev> It's just not worth the trouble to set up too much organization at this point.
21:25:50 <notmyname> maybe it's something super simple like some comment lines in tox.ini on how to read it and how to add something to the list
21:26:01 <zaitcev> Well, if 5 oe 20 people show up, maybe then.
21:26:19 <notmyname> zaitcev: ...also, I'd like a pony! ;-)
21:26:21 <zaitcev> For some reason Cyril isn't showing up on IRC
21:26:28 <zaitcev> So I e-mail him
21:26:29 <notmyname> it's super late for him
21:26:31 <tdasilva> zaitcev: he was here last week
21:26:36 <zaitcev> oh, okay
21:26:37 <tdasilva> and he was great!
21:26:43 <zaitcev> Good, good
21:27:01 <zaitcev> Well, he's good in general. He had experience with other OpenStack projects.
21:27:31 <notmyname> zaitcev: what do you think of adding some comments to the list of py3 tests on how to read the list and how to find stuff to add to the list?
21:27:46 <zaitcev> My trip was too busy all the time, I basically didn't do any work at all.
21:27:56 <notmyname> in lieu of setting up more comprehensive projet tracking systems
21:28:00 <zaitcev> notmyname: okay
21:28:09 <notmyname> cool
21:28:13 <tdasilva> zaitcev: so account is all done?
21:28:53 <zaitcev> tdasilva: unit tests, yes. This implies that common DB layer is done too, it's just that we don't have container yet to prove it.
21:29:05 <notmyname> tdasilva: zaitcev: rledisez: kota_: anything else to bring up today?
21:29:10 <tdasilva> zaitcev: ack
21:29:22 <kota_> notmyname: nothing
21:29:51 <notmyname> ok
21:30:08 <notmyname> we will have a meeting again next week. we should have the 2.20.0 release out by then
21:30:28 <notmyname> and then that will be our last meeting for 2018. we'll pick up again on january 3
21:30:38 <notmyname> sorry, jan 2
21:30:46 <kota_> oic. Christmas is close.
21:30:46 <rledisez> notmyname: nothing for me neither
21:31:24 <rledisez> fyi, i'll be off for the meeting of january 3rd
21:31:28 <notmyname> ok
21:31:37 <notmyname> thanks, everyone, for your work on swift
21:31:49 <notmyname> #endmeeting