21:02:28 <tonyb> #startmeeting stable
21:02:29 <openstack> Meeting started Mon Jan 30 21:02:28 2017 UTC and is due to finish in 60 minutes.  The chair is tonyb. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:02:30 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:02:32 <openstack> The meeting name has been set to 'stable'
21:03:35 * mtreinish noticed today
21:05:19 <sigmavirus> o/
21:06:07 <tonyb> mtreinish: \o/
21:06:11 <tonyb> sigmavirus: hey there
21:06:30 <tonyb> sigmavirus: sorry to split focus on the glance issue and the meeting
21:06:35 <sigmavirus> tonyb: no worries
21:06:51 <sigmavirus> That has been going on for days (I was distracted with release cpl duties)
21:06:56 <sigmavirus> So it can probably wait a bit longer =P
21:07:25 <tonyb> sigmavirus: Perhaps but it might be a fun way for me to remeber what it is that I do ....
21:07:44 <sigmavirus> hah
21:08:43 <tonyb> I haven
21:09:39 <tonyb> 't written up an agenda for today but I guess it's Status, PTG/Pike planning (if any), Pike glance bug
21:09:56 <sigmavirus> Mitaka glance bug?
21:10:25 <tonyb> sigmavirus: Yeah bad inline editing
21:10:37 <sigmavirus> tonyb: no worries. Just wanted to make sure I wasn't missing something
21:10:45 <tonyb> :)
21:11:22 <mtreinish> sigmavirus: heh, tonyb is getting ready to discuss next month's bugs already :)
21:11:36 <tonyb> So status. liberty needs the nest round of EOLing done, then clean up all the old branches
21:11:44 <sigmavirus> Wise and the gift of foresight. What more could we want?
21:11:55 <tonyb> asside from the mitaka glance bug (and pribably a ceilomioter one) is the gate mostly okay?
21:13:01 <sigmavirus> seems to be
21:13:10 <tonyb> \o/
21:13:40 <mtreinish> http://status.openstack.org/openstack-health/#/?groupKey=build_branch&searchProject=stable*
21:14:01 <mtreinish> I do like the stable hammer branch, although I have no idea what it is
21:14:21 <sigmavirus> mtreinish: someone saw a nail?
21:14:58 <sigmavirus> looks like the puppet project
21:15:19 <mtreinish> yeah
21:15:33 <tonyb> I think it's a puppet release name
21:15:54 <tonyb> mtreinish: I have questions on that view ...
21:16:10 <mtreinish> tonyb: sure
21:16:55 <tonyb> mtreinish: Are the failure rates really percentages? it goes from 0 -> 1 so I want to read 0.1 as 10% but that doesn't match with the text
21:17:25 <tonyb> 0.1% is meh 10% is more interesting ....
21:17:46 <mtreinish> yeah it's 10% not .1%
21:17:52 <mtreinish> that's definitely a bug
21:18:23 <mtreinish> it's missing a * 100 somewhere in the js
21:19:10 * sigmavirus would troll about it displaying fine in his browser, but won't
21:19:25 <tonyb> mtreinish: okay I'll fix that
21:20:11 <tonyb> sigmavirus: it displays fine for me or are you sayign that the 'Job Failure Rate' goes from 0 to 100 in the y- axis for you?
21:20:36 <sigmavirus> tonyb: ignore me
21:20:42 <sigmavirus> I've had too much caffeine today :)
21:20:50 <tonyb> sigmavirus: I have the inverse problem ;P
21:21:43 <tonyb> mtreinish: there used to be a way to remove the passing tests from the total jobs graph, did that go away or am I just not seeing it?
21:22:28 <mtreinish> tonyb: it went away when we stopped using nvd3 (for memory and performance reasons)
21:22:45 <tonyb> mtreinish: Ah okay
21:23:03 <mtreinish> I can bug timothyb89 about adding it back. Because I barely understanding all the code he wrote to render the graphs :)
21:23:32 <tonyb> mtreinish: I can poke him in -qa
21:23:57 <tonyb> mtreinish: Thanks for the link
21:24:00 <timothyb89> mtreinish, tonyb: I will look into adding it back
21:24:05 <mtreinish> timothyb89: cool, thanks
21:24:10 <tonyb> timothyb89: \o/
21:24:29 * tonyb really likes the "lurk in all the meeting rooms" thing :)
21:24:38 <tonyb> so ....
21:24:48 <tonyb> #topic PTG/Pike planning
21:25:05 <tonyb> we have a shared space with release/requirements in ATL
21:25:21 <tonyb> do we want to schedule a working bee or any key discussion topics?>
21:25:23 * sigmavirus won't be at the PTG
21:25:31 <tonyb> sigmavirus: :(
21:25:58 <mtreinish> tonyb: are we going to have the obligatory make support windows longer discussion? :p
21:26:08 <tonyb> mtreinish: I was hoping to avoid it
21:26:22 <sigmavirus> I'll send someone to instigate it, just for you mtreinish
21:26:49 <tonyb> sigmavirus: prometheanfire is a reasonable proxy (assumeing he'll be theer)
21:27:00 <mtreinish> sigmavirus: LOL
21:27:22 <sigmavirus> tonyb: Matt doesn't listen to me
21:27:45 <tonyb> sigmavirus: He will if you're aking hime to make trouble ... he likes that sort of thing ;P
21:28:02 <sigmavirus> tonyb: not if it's me asking him to cause trouble
21:28:09 <tonyb> sigmavirus: Well if you think of anything you'd like discuessed and if you can attend remotely we can do that thing
21:28:19 <tonyb> sigmavirus: ;P
21:28:30 <sigmavirus> If remote attendance works this year, I'll try to attend remotely, but I have nothing in particular
21:28:39 <tonyb> sigmavirus: okay.
21:28:49 <sigmavirus> I'd love if I could find another glance stable core who has time to review things but I seem to be the only person
21:28:58 <sigmavirus> And really taht's fine because so few people are backporting anything anywa
21:28:58 <tonyb> sigmavirus: I wonder if it's worth trying to nail down the "do we need meetings" thing
21:29:05 <sigmavirus> tonyb: I like meetings
21:29:09 <sigmavirus> even if they're only short
21:29:17 <sigmavirus> The security team has capped meetings to 30 min
21:29:18 <tonyb> sigmavirus: okay.
21:29:26 <sigmavirus> And I'm happy to run meetings if that's desirable
21:29:28 <tonyb> sigmavirus: I wonder if we shoudl do the same
21:29:31 <sigmavirus> I run very tight meetings
21:30:12 <tonyb> I overlap moer with the US so it's easier to have a quick chate with y'all but Eurpoe is harder and it's been good to speak to ttx and apevec from time-to-time
21:30:29 * tonyb makes a note to look at the security meeting logs
21:30:39 * tonyb waffles a bit (read lot)
21:31:13 <tonyb> #topic mitaka glance bug
21:31:31 <sigmavirus> tonyb: at least, if you ever need a sub, I can help =P
21:31:39 <tonyb> sigmavirus: :)
21:32:21 <sigmavirus> So yeah, the glance bug I'm pretty sure is related to setuptools 34.*
21:32:24 <tonyb> sigmavirus: You had 3 possible solutions to the bug new enough pip, chnage the requirements ordering ... what was thr 3rd?
21:32:43 * tonyb really would like to be able to cap setuptools but alas :(
21:32:46 <sigmavirus> I don't think the third is at all preferable but changing upper constraints for setuptools
21:33:08 <sigmavirus> That said, I'm not sure where best to add the macro to get pip's version
21:33:12 <tonyb> sigmavirus: Ahh we don't (shoudln't) have setuptools in u-c
21:33:20 <sigmavirus> And someone on my team is seeing similar issues with knew-enough pip
21:33:28 <sigmavirus> So, I'm not sure option 1 is that much of an option
21:33:43 <tonyb> sigmavirus: okay so how does option 2 help?
21:34:13 <sigmavirus> tonyb: so, pip *should* be ordering what is installed and in what order to avoid problems where a library needs a dependency that doesn't exist
21:34:30 <sigmavirus> If we give pip an explicit ordering, it might help
21:34:39 <sigmavirus> I'm not certain it *will* but I think it's worth trying
21:34:40 <mtreinish> sigmavirus: would opt 3, changing u-c work? because it depends on pip (which includes setuptools)
21:34:57 <mtreinish> oh, I'm slow tonyb realized that already
21:35:05 <sigmavirus> mtreinish: so pip doesn't force setuptools to a specific version
21:35:11 <sigmavirus> it's happy to use something that's remotely new enough
21:35:46 <mtreinish> sigmavirus: well it uses whatever is installed when it creates the venv right?
21:35:51 * mtreinish is still fighting jetlag
21:35:55 <sigmavirus> mtreinish: uh, not quite
21:35:57 <sigmavirus> heh
21:36:10 <sigmavirus> virtualenv has code to download/install latest pip/setuptools by default iirc
21:36:18 <tonyb> sigmavirus: option would be a bit house of cardsish wouldn't it .... I mean if we say $lib1 $lib2 and thr deps for $lib1 chnage then the ordering we check will break wont it?
21:36:21 <sigmavirus> That might not be true for system-packaged versions of virtualenv though
21:36:49 <sigmavirus> tonyb: yeah exactly,
21:36:57 <sigmavirus> tonyb: that said, six has no dependencies, so I figure moving it to the top would be safe enough
21:36:58 <tonyb> sigmavirus: I though venv always used the vendored tools ans tox woudl update them but I'm hazy there
21:37:15 <sigmavirus> tonyb: nah, tox uses what ever virtualenv installs
21:37:44 <tonyb> sigmavirus: okay so much confusion in how all this works for most people including me
21:38:51 <tonyb> sigmavirus: I think that moving six it a reasonable band-aid for now.  I assume it's only hitting mitakak 'cause of somethign else we've cappu in u-c?
21:39:24 <mtreinish> tonyb: oh, is it just an ordering probably in requirements? that's an acceptable fix for the lack of a dep solver
21:39:30 * mtreinish hasn't seen the actual bug
21:39:42 <tonyb> https://bugs.launchpad.net/glance/+bug/1660444
21:39:42 <openstack> Launchpad bug 1660444 in Glance mitaka "Glance stable/mitaka docs cannot be built" [High,Triaged]
21:40:43 <sigmavirus> I don't lightly blame third party devs who do the work in their free time, but I feel like this was borderline irresponsible of setuptools and this has caused serious problems for other (non-OpenStack) people/projects
21:41:50 <tonyb> sigmavirus: Yeah it seems like a strange move and not well tested/communicated but we can't really help that now
21:41:52 <sigmavirus> Actually, setuptools isn't even in our requirements list
21:42:01 <sigmavirus> *our = Glance
21:42:43 <sigmavirus> So moving this around won't really do very much because pip is relying on setuptools and pkg_resources at run time and then six is being pulled out before pkg_resources can import it? (Maybe?)
21:42:44 <sigmavirus> Idk
21:42:48 <sigmavirus> I need to dig deeper into this
21:42:48 <mtreinish> sigmavirus: it wouldn't be, normally it's set via pbr (or something else)
21:42:50 <tonyb> sigmavirus: that's pretty common for OpenStack
21:43:07 <sigmavirus> Yeah, so reordering probably wouldn't even work
21:43:09 <sigmavirus> \o\
21:43:13 <sigmavirus> /o/
21:44:24 <tonyb> sigmavirus: okay we
21:44:37 <tonyb> ll play with it. how much longer are you online for?
21:44:48 <sigmavirus> tonyb: technically I should be getting offline
21:44:54 <sigmavirus> or rather, should already be offline
21:45:04 <tonyb> sigmavirus: okay.
21:45:25 <sigmavirus> I might reach out to Donald Stufft for some help
21:45:31 <tonyb> sigmavirus: I need to take the kids to school and I wont ask you to stay around longer
21:45:37 <sigmavirus> No worries
21:45:46 <sigmavirus> I'll probably be online on my non-work IRC in an hour or two
21:45:56 <sigmavirus> Have to release some non-OpenStack software to the PyPIs
21:46:30 <tonyb> sigmavirus: okay I'll ping you in -stable replay if you online and want to chat I wont be offedned if you ignore me ;P
21:46:50 <sigmavirus> heh, I don't use this nick after I finish work. But I'll hop in with my other one
21:47:13 <tonyb> sigmavirus: Yeah that'd be the one I highlighted ;P
21:47:39 <sigmavirus> anyway, anything else for today?
21:47:54 <tonyb> I was about the # endmeeting ....
21:48:37 <tonyb> Thanks all!
21:48:41 <tonyb> #endmeeting