18:08:14 <SumitNaiksatam> #startmeeting networking_policy
18:08:14 <openstack> Meeting started Thu Jan  5 18:08:14 2017 UTC and is due to finish in 60 minutes.  The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:08:15 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:08:17 <openstack> The meeting name has been set to 'networking_policy'
18:08:24 <SumitNaiksatam> happy 2017 to all!
18:08:39 <tbachman> SumitNaiksatam: likewise!
18:08:52 <SumitNaiksatam> lets starting on the happy note of pending patches
18:08:57 <SumitNaiksatam> #topic Pending Patches
18:09:24 <SumitNaiksatam> #link https://review.openstack.org/#/c/416527/
18:09:35 <SumitNaiksatam> “Set project when doing neutron DB migrations"
18:09:55 <SumitNaiksatam> (i am picking up the patches which are not driver or nfp-specific"
18:09:57 <SumitNaiksatam> )
18:10:05 <SumitNaiksatam> anyone seen this one?
18:10:29 <SumitNaiksatam> as long as all that this is doing is pointing to the neutron conf directories i am good with that
18:10:57 <SumitNaiksatam> it didnt seem to break our integration job, so i think this should be good
18:11:16 <SumitNaiksatam> any objections to the above?
18:11:51 <rkukura> makes sense to me, I guess, but I don’t claim to know the config internals
18:12:03 <SumitNaiksatam> rkukura: same here, hence wanted to check with the team :-)
18:12:11 <SumitNaiksatam> i will +2
18:12:56 <SumitNaiksatam> next, “neutron-lib: use L3 constant from neutron-lib” - #link https://review.openstack.org/#/c/414976/
18:13:28 <SumitNaiksatam> my feeling on this (and the one we will pick after this) is that we need to do the newton refactor first before we can get to this
18:13:32 <rkukura> SumitNaiksatam: I +2/+A’d the config patch. Does that one need back-porting?
18:13:43 <SumitNaiksatam> rkukura: yes, it would
18:13:48 <SumitNaiksatam> rkukura: thanks
18:14:44 <SumitNaiksatam> the integration jobs fail on this presumably because we are still on mitaka, and it doesnt go well with the use of the neutron-lib
18:14:56 <rkukura> I do think we should use neutron_lib, at least as far as it exists in mitaka
18:15:50 <SumitNaiksatam> rkukura: okay, i just checked, it seems that it exists in mitaka
18:15:57 <songole> neutron-lib is available in mitaka itself I suppose
18:16:27 <rkukura> Its evolved since, but I think the changes in this patch only depend on what’s in it in mitaka, or else UTs would file, right?
18:16:29 <SumitNaiksatam> okahy my bad
18:16:35 <SumitNaiksatam> we are already using neutron_lib
18:16:49 <SumitNaiksatam> yeah my bad
18:17:04 <SumitNaiksatam> so need to go back and look at this patch as to why this is failing
18:17:58 <SumitNaiksatam> next one - “Change passing session to context in segments db functions” - #link https://review.openstack.org/#/c/400811/
18:18:31 <rkukura> How do you get a context when you don’t have one?
18:18:41 <SumitNaiksatam> i dont think this is a mitaka compatible change
18:19:01 <SumitNaiksatam> rkukura: create one?
18:19:19 <rkukura> sure
18:19:37 <rkukura> Actually, this patch is just for one function call, not a general change
18:19:38 <SumitNaiksatam> oh you mean, here where the change is made
18:20:38 <SumitNaiksatam> the last i checked, the changed behavior that is being referred to in this patch, is in neutron newton
18:20:40 <rkukura> SumitNaiksatam: So this release_network_segments signature changed post-mitaka?
18:21:20 <SumitNaiksatam> rkukura: thats what i recall, but need to double check
18:21:28 <rkukura> ok
18:22:59 <rkukura> SumitNaiksatam: Any idea what this enginefacade change was, and whether we should be using context instead of session in more places (our own DB layets, etc.)?
18:23:04 <SumitNaiksatam> rkukura: #link https://github.com/openstack/neutron/commit/df7e36c8c8b894ff385e0404b30698f7cb2a2ffb
18:23:41 <SumitNaiksatam> rkukura: i think Anna had sent out an email, and there are some pointers in that email, but i dont recall enough to summarize here
18:24:29 <SumitNaiksatam> rkukura: so confirmed, in mitaka its still the old signature #link https://github.com/openstack/neutron/blob/stable/mitaka/neutron/plugins/ml2/managers.py#L254
18:24:45 <SumitNaiksatam> and that would explain the failed UTs
18:25:38 <SumitNaiksatam> so unfortunately, this patch would have to wait until the newton catch up (which will hopefully happen soon)
18:25:52 <SumitNaiksatam> any other patches we need to discuss?
18:25:54 <rkukura> right - we need this patch when we support newton, but I’m more concerned about whether we should be passing context everywhere we pass session right now
18:26:11 <SumitNaiksatam> rkukura: sure, we would need to evaluate that
18:26:41 <rkukura> #link https://blueprints.launchpad.net/neutron/+spec/enginefacade-switch
18:26:52 <SumitNaiksatam> actually this signature exists in newton as well: #link https://github.com/openstack/neutron/blob/stable/newton/neutron/plugins/ml2/managers.py
18:27:05 <SumitNaiksatam> so this is a Ocata feature
18:27:08 <SumitNaiksatam> rkukura: thanks
18:27:39 <SumitNaiksatam> i think Anna posted this change on the master branches of different projects assuming they are all on Ocata
18:27:55 <rkukura> it looks like it was addressed in OSLO in kilo
18:28:26 <SumitNaiksatam> rkukura: okay
18:28:40 <SumitNaiksatam> songole: anything at your end we need to look at?
18:29:19 <songole> SumitNaiksatam: are you referring to this Enginefacade?
18:29:28 <SumitNaiksatam> songole: no, NFP
18:29:53 <songole> SumitNaiksatam: no outstanding patches
18:29:58 <SumitNaiksatam> songole: oh cool
18:30:18 <SumitNaiksatam> okay, so if no one has anything else, lets wrap a little early today
18:30:33 <SumitNaiksatam> rkukura and I have another meeting to catch
18:30:47 <SumitNaiksatam> thanks all for joining, and happy new year again!
18:30:50 <SumitNaiksatam> bye!
18:30:54 <tbachman> SumitNaiksatam: bye!
18:30:59 <songole> bye
18:31:00 <SumitNaiksatam> #endmeeting