20:30:51 <prometheanfire> #startmeeting requirements
20:30:52 <openstack> Meeting started Wed Apr 17 20:30:51 2019 UTC and is due to finish in 60 minutes.  The chair is prometheanfire. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:30:53 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:30:55 <openstack> The meeting name has been set to 'requirements'
20:31:02 <prometheanfire> #topic rollcall
20:31:07 <prometheanfire> tonyb, prometheanfire, number80, dirk, coolsvap, toabctl, smcginnis, dhellmann ping
20:31:10 <prometheanfire> o/
20:32:13 <smcginnis> o/
20:35:33 <smcginnis> Looking like it's just going to be us.
20:35:51 <prometheanfire> ya, will wait for tony a couple of min
20:37:44 <tonyb[m]> \o
20:37:46 <tonyb[m]> Sorry
20:38:29 <prometheanfire> #topic controversies in the queue
20:38:44 <prometheanfire> #link https://review.openstack.org/652624
20:39:38 <tonyb[m]> I feel like just using u-c for now is the right call
20:39:58 <prometheanfire> I'm fine with that and would rather just do that
20:40:01 <tonyb[m]> removing caps is a lot of work so if we can avoid that we should
20:40:09 <smcginnis> ++
20:40:14 <prometheanfire> good point, it has knock on effects
20:40:28 <prometheanfire> I'd suggest adding a != for known bad versions
20:40:41 <prometheanfire> #link https://review.openstack.org/652629
20:40:51 <tonyb[m]> Yup
20:40:55 <prometheanfire> I'd like some more eyes, I haven't eval'd the code yet but looks 'ok'
20:41:27 <smcginnis> I think protobuf is used a lot elsewhere. I think it should be safe.
20:42:01 <tonyb[m]> Yeah I think it's kinda the new hotness
20:42:15 <tonyb[m]> I think it shoudl be okay in general but I haven't done a real review
20:42:21 <prometheanfire> indeed
20:42:33 <prometheanfire> haven't took the deep look yet either
20:42:48 <smcginnis> protocol buffers have been around for years. Not sure about this specific library, but it is a well established thing.
20:43:03 <prometheanfire> #link https://review.openstack.org/651591
20:43:25 <smcginnis> Ah protobuf lib has at least been around since 2008. Should be "mature".
20:43:41 <prometheanfire> that has issues with tempest full, haven't seen what project is failing but it's not a cross test one it seems
20:43:47 <prometheanfire> or I would expect
20:44:06 <tonyb[m]> What's been holding us back on sqlalchemy
20:44:33 <prometheanfire> haven't seen why tempest keeps failing yet, just calling it out for now
20:44:58 <smcginnis> I would be surprised if this is a sqlalchemy issue: http://logs.openstack.org/91/651591/1/check/tempest-full/9a1bcab/job-output.txt.gz#_2019-04-16_22_48_18_679664
20:45:19 <smcginnis> If it's pulling the wrong data from the DB, that would be Very Bad.
20:45:26 <prometheanfire> VERY
20:45:28 <prometheanfire> BAD
20:45:32 <prometheanfire> ya
20:45:44 <tonyb[m]> But we're a long way back 1.2.18 -> 1.3.2 I assume we had a reason for that
20:46:31 <smcginnis> Original neutron issue that kept it back looks different: http://paste.openstack.org/show/748825/
20:46:35 <prometheanfire> to stay on 1.2.18?
20:46:52 <tonyb[m]> I don't know that could be a sqla issue, it looks like the test updates a description if that didn't land correctly in the DB that would case a mismatch
20:47:00 <prometheanfire> at the time it was that openstack stuff kept failing, now it may be sqla itself
20:47:01 <tonyb[m]> I'm not sayin it *is* but it could be
20:47:13 <smcginnis> Comments at patch set 2: https://review.openstack.org/#/c/649508
20:47:40 <tonyb[m]> smcginnis: Thanks for those links
20:48:02 <prometheanfire> ya, could
20:48:12 <prometheanfire> anything else?
20:48:32 <tonyb[m]> okay probably less of an issue than I thought oon the sqla chnage
20:48:45 <tonyb[m]> just all the -2's from zuul on the pike chnages
20:48:55 <tonyb[m]> I haven't looked at them are they legit?
20:50:28 <prometheanfire> I don't think so, some are merge fails, some are random, consider the count I think we are fine
20:50:35 <smcginnis> IndexError: tuple index out of range   OperationalError: (2014, 'Command Out of Sync')
20:50:43 <smcginnis> Random failures that look unrelated to me.
20:51:27 <smcginnis> True, there were quite a few. Sorry, I probably should have given a heads up before I started approving so many of those.
20:51:42 <smcginnis> Guessing we will just have to do some more rechecks to work the rest through.
20:52:32 <tonyb[m]> okay I can plug away at them today
20:52:47 <prometheanfire> cool
20:53:14 <tonyb[m]> There will be more coming over the next couple of weeks
20:53:20 <prometheanfire> #topic conf/ptg
20:53:25 <tonyb[m]> as we transition pike to EM
20:54:13 <smcginnis> tonyb[m]: I was lenient on approving ones that didn't really have code changes, but now considering the extra work they can cause, maybe we should review the list-changes for those some more and drop the ones that don't actually change anything.
20:54:20 <tonyb[m]> Can we mostly do requirements stuff on Friday after lunch?
20:54:35 <prometheanfire> tonyb[m]: wfm
20:54:56 <tonyb[m]> smcginnis: When I looked there weren't that many that didn't have code changes
20:55:01 <smcginnis> Did we figure out a release team time? May be good to combine them or at least make them back to back.
20:55:13 <smcginnis> tonyb[m]: OK, I'll watch for more.
20:55:26 <tonyb[m]> smcginnis: I *think* releases will be meeting Friday morning
20:55:49 <smcginnis> OK, that's good then. Release in the morning, requirements in the afternoon. Works for me.
20:56:02 <tonyb[m]> That's kinda my thinking ;P
20:56:46 <smcginnis> prometheanfire: Honeymoon plans all set? :)
20:58:03 <prometheanfire> heh, that'll be later in the summer/fall I think
20:58:18 <prometheanfire> open floor time?
20:58:32 <tonyb[m]> +1 to open floor
21:00:12 <prometheanfire> #topic meetings for next two weeks
21:00:27 <prometheanfire> so, I'm out next week, and the week after is conf/ptg
21:00:35 <tonyb[m]> I'll be on a plane this time next week
21:00:38 <prometheanfire> so, no meetings?
21:00:44 <smcginnis> wfm
21:00:52 <tonyb[m]> Yeah I think skip the next 2
21:01:00 <prometheanfire> kk
21:01:04 <prometheanfire> #topic open floor
21:01:05 <tonyb[m]> the one post PTG I'll be in the US so huzzah!
21:01:10 <prometheanfire> :D
21:01:26 <tonyb[m]> mind you it'll be 1630'ish ;p
21:01:30 <smcginnis> You can pick up some American Tim Tams. :P
21:01:39 <prometheanfire> lolol
21:01:40 <tonyb[m]> LOL
21:01:59 <tonyb[m]> That reminds me I need to get some for spotz and OUI
21:04:13 <tonyb[m]> anything else?
21:04:29 <smcginnis> Not from me.
21:04:33 <prometheanfire> #endmeeting