15:01:06 <markus_z> #startmeeting nova-config-options
15:01:07 <openstack> Meeting started Mon Jan 11 15:01:06 2016 UTC and is due to finish in 60 minutes.  The chair is markus_z. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:10 <openstack> The meeting name has been set to 'nova_config_options'
15:01:40 <markus_z> sfinucan: hey ho, let's see who got the short notice message
15:02:01 <ryansb> markus_z: uh, this is the zaqar meeting time/channel
15:02:12 <ryansb> I don't see nova-config on the os-meeting-cal
15:02:23 <ccarmack> o/
15:02:25 <markus_z> ryansb: I didn't see that in the ical
15:02:51 <ryansb> oh >.> don't mind me, I'm off by a week
15:03:08 <ryansb> carry on
15:03:33 <markus_z> ryansb: cool :) did I interrupt maybe another meeting, so speak up :)
15:03:59 <ryansb> no, doesn't seem like it (from the cal) sorry for the interruption
15:04:08 <markus_z> no worries
15:04:45 <markus_z> ah, well, let's just do it
15:04:50 <markus_z> #topic current state
15:05:09 <markus_z> so it's ccarmack and sfinucan
15:05:13 <ccarmack> hi
15:05:24 <markus_z> hey ho
15:05:30 <markus_z> #link http://lists.openstack.org/pipermail/openstack-dev/2016-January/083598.html
15:05:45 <markus_z> I just wanted to have a short chat about the work on nova's config options
15:05:56 <markus_z> and again, thanks for your support their
15:06:24 <markus_z> do you maybe have some open questions until this point?
15:06:43 <ccarmack> markus_z: I had a question about testing...
15:07:00 <markus_z> ccarmack: ok, go on
15:07:07 <ccarmack> Do we need to write new tests if need to verify the options are not broken
15:07:29 <ccarmack> I'm not sure the current tests exercise all the options
15:07:47 <markus_z> ccarmack: I'd say "no".
15:08:10 <markus_z> Are you worried that they don't get read?
15:08:42 <ccarmack> I'm more worried about breaking on option, like changing the name or the default
15:09:21 <ccarmack> I guess careful inspection is what we can do
15:09:24 <markus_z> the name change should already be catched by unit tests
15:10:04 <sfinucan> ccarmack: If you don't actually change the default or name you should be fine, right?
15:10:26 <sfinucan> I mean, you could but a code reviewer _should_ pick it up
15:10:28 <ccarmack> sfinucan: yes, just need to be careful
15:10:54 <markus_z> I will have a look at that in the reviews
15:11:00 <ccarmack> ok
15:11:04 <sfinucan> ccarmack: True. I've kept changes related to renamed groups etc. in different commits
15:11:17 <markus_z> I don't think we need additional testing here.
15:11:33 <markus_z> renaming groups is a different thing. That needs deprecation and reno files
15:12:06 <markus_z> Did that already happen and I overlooked it?
15:12:23 <sfinucan> markus_z: Not from me, no
15:12:41 <sfinucan> I have some in review sans reno, but I'll go fix that soon :)
15:12:54 <markus_z> ok, cool
15:13:05 <sfinucan> markus_z: I have a question, so?
15:13:11 <markus_z> sfinucan: sure
15:13:35 <sfinucan> Is there anyway to go about getting these merged faster?
15:14:06 <markus_z> This review should be the last one which helps us with the merge conflicts https://review.openstack.org/#/c/260015
15:14:22 <markus_z> That should speed things up
15:14:30 <markus_z> I guess I'm the bottleneck right now
15:14:39 <markus_z> As long as I don't +1 it the cores won't look at it
15:15:08 <sfinucan> markus_z: Ah - I saw that but didn't know how critical it was. I'll rebase my changes onto it
15:15:13 <sfinucan> ...or wait for it to merge
15:15:33 <markus_z> sfinucan: yeah, wait for the merge. I'm fighting with XENCI right now. Should be done by today.
15:15:54 <sfinucan> OK. Even once that bottleneck's cleared though, I'm finding it hard to get eyes on. None of us are cores so we can't force things :(
15:15:59 <markus_z> It also helps if you already have the enhanced help texts in a followup patch
15:16:11 <sfinucan> markus_z: Oh yeah?
15:16:33 <sfinucan> I was figuring they'd be more contentious, so was focusing on the easier, even-lower-hanging fruit first :)
15:16:46 <markus_z> yeah, that's the main focus of this task
15:17:11 <markus_z> It should be somehow a single bucket to look at
15:17:36 <sfinucan> This is true
15:17:51 <sfinucan> OK, I'll put some energy into that later today/tomorrow so
15:17:52 <markus_z> mitaka-2 milestone is ahead any many people try to merge their features, that's also why the cores have less time
15:18:11 <sfinucan> markus_z: also true (I'm one of those people)
15:18:30 <sfinucan> markus_z: One more question?
15:18:46 <sfinucan> How broad reaching is this spec going to be?
15:18:46 <markus_z> I have to discuss with our PTL if we will be concerned by the cut-off date or if we can keep on pushing changes as we don't provide a feature which could introduce major bugs
15:18:58 <markus_z> what do you mean by that?
15:19:15 <sfinucan> Are we going to be concerning ourselves with those "items for consideration"?
15:19:41 <sfinucan> i.e. "this patch is a stringopt when it should be an integeropt"
15:19:41 <markus_z> hm, what are "items of consideration"? Did I miss something?
15:19:50 <markus_z> ah, ok, got it
15:20:05 <markus_z> just note it down for later
15:20:15 <markus_z> the main focus is help text for our opts
15:20:21 <ccarmack> sfinucan: for the help text, I see you committed the vnc changes.  I did the spice changes, but have different help text.   https://review.openstack.org/#/c/264271/
15:20:23 <markus_z> good help texts for our opts
15:20:43 <ccarmack> we should have simliar text for the console groups
15:21:07 <markus_z> sounds reasonable, but don't let the perfect be the enemy of the good
15:21:08 <sfinucan> ccarmack: Agreed. I copied mine from somewhere else, though I don't recall where...
15:21:39 <sfinucan> I'm happy to fix - feel free to stick in 'nit:' comments as suits, ccarmack
15:21:49 <ccarmack> sfinucan: maybe when are changes are reviewed, we can match up the help text
15:21:52 <ccarmack> ok, will do
15:22:59 <markus_z> ok, cool, more questions?
15:23:25 <sfinucan> ccarmack: Maybe stick in the "preferred" format in the spec - I'll copy that
15:23:34 <sfinucan> markus_z: Not from me, no
15:23:47 <ccarmack> markus_z: I'm good
15:24:25 <markus_z> ok, I'd like to give you a heads up for this patch: https://review.openstack.org/#/c/260015/4
15:24:39 <markus_z> It introduces a change for your /nova/conf/xxx.py modules
15:25:32 <markus_z> In the previous patches you returned a tuple in "list_opts", like return ("vnc", vnc_opts)
15:26:00 <markus_z> that will change to a dict like return { "vnc": vnc_opts }
15:26:35 <markus_z> the new "nova/conf/opts.py" module will then read it in and everything is fine
15:26:41 <sfinucan> markus_z: Sounds good. I'd suggest also putting that into the spec
15:26:56 <markus_z> you just have to remove a previous reference in the old "nova/opts.py" (or similar)
15:27:38 <markus_z> sfinucan: yeah, a follow up patch there makes sense, it's a somewhat bigger change in that concept
15:28:44 <markus_z> the change also means that all the spread "opts.py" modules will be stripped down with the listing of the config options until they don't have any listing anymore
15:29:02 <markus_z> because everything is done automatically by the one and only "nova/conf/opts.py"
15:29:13 <markus_z> all other opts.py modules will be deleted at the end
15:29:24 <markus_z> does that make sense to you?
15:30:17 <markus_z> #link example of stripping down an opts.py module: https://review.openstack.org/#/c/260015/4/nova/opts.py
15:31:47 <sfinucan> markus_z: Yes, I think so
15:31:51 <sfinucan> If not, I'll ping you :)
15:32:13 <markus_z> :) yepp, please do so
15:32:39 <markus_z> ok, that are the major points I wanted to discuss
15:32:54 <markus_z> Thanks again for helping with that
15:33:06 <markus_z> Any open questions?
15:33:12 <ccarmack> none from me
15:33:58 <markus_z> cool, thanks for your participation :)
15:34:00 <markus_z> #endmeeting