15:04:02 <tmorin> #startmeeting bgpvpn
15:04:03 <openstack> Meeting started Tue Jan  5 15:04:02 2016 UTC and is due to finish in 60 minutes.  The chair is tmorin. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:04:04 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:04:07 <openstack> The meeting name has been set to 'bgpvpn'
15:04:13 <tmorin> hi doude
15:04:22 <tmorin> hi pcarver
15:04:25 <pcarver> hi
15:04:27 <doude> tmorin: hi
15:04:33 <tmorin> hi timirnich
15:07:08 <matrohon> anybody has some announcements?
15:07:33 <tmorin> nothing big on my side
15:08:41 <tmorin> #topic kilo backport
15:09:00 <tmorin> some work has been done since last meeting
15:09:13 <tmorin> afaics the kilo backport is ready
15:09:33 <enikher> nice to here
15:09:34 <tmorin> maybe some more testing needs to be done, but I'd say we could tag a release
15:09:54 <tmorin> enikher: you did most of the heavy work ! :)
15:10:24 <tmorin> the missing part until last merges, was the bagpipe driver backport
15:10:24 <enikher> yes but the last bits where added by you so thanks as well
15:11:40 <tmorin> ok, so we just need to see if we need/want more testing
15:11:46 <tmorin> next topic ?
15:12:04 <tmorin> good transition!...
15:12:11 <tmorin> #topic tempest testing
15:12:32 <tmorin> thanks to enikher, we now have a start of tempest tests
15:13:25 <tmorin> what is needed now are (1) zuul/jenkins config to use this and (2) more tests
15:13:41 <tmorin> 1 is already in progress:  https://review.openstack.org/#/c/258522/
15:14:21 <tmorin> #help is welcome on 2: adding more tests so that the gate will actually start doing significant functional tests
15:14:30 <tmorin> any comments ?
15:14:33 <tmorin> enikher ?
15:14:50 <tmorin> since you are the one working on the additional zuul/jenkins config
15:14:52 <tmorin> ?
15:15:07 <enikher> sorry I was afk
15:15:16 <enikher> back and reading
15:16:03 <enikher> At the moment I am troubled with other work.
15:16:12 <enikher> so yes help is welcome
15:16:31 <enikher> I think I will make progress at the end of this week
15:17:13 <enikher> I think one thing we should talk about is the neutron client used in tempest
15:17:29 <enikher> it is normaly not the same as used by python-neutronclient
15:17:53 <enikher> which is bad I think because we have to implement things twice and...
15:18:04 <enikher> we are not testing the original code
15:18:12 <enikher> of python-neutronclient
15:18:29 <tmorin> you mean we are not testing our python-neutronclient extension
15:18:33 <tmorin> yes, this is not ideal
15:18:34 <enikher> yes
15:18:36 <enikher> exactly
15:18:41 <matrohon> enikher : why? it has it own client?
15:18:48 <enikher> yes
15:18:56 <enikher> tempest is very stupid there
15:19:21 <tmorin> we need to decide if we need/want to follow this practice
15:19:38 <enikher> I have to check how much impact to tempest it has when we use the real neutronclient
15:20:12 <enikher> from my side a big -1 for this practice :-)
15:21:00 <tmorin> could we have our tempest plugin automatically play each scenario, once with a new built-in client, and once with our python-neutronclient extension ?
15:21:04 <matrohon> enikher, so there is no tempest test for clients?
15:21:05 <tmorin> would that be doable ?
15:22:35 <matrohon> enikher, I see some tempst jobs dedicated to test neutronclient : gate-tempest-dsvm-neutron-src-python-neutronclient
15:23:19 <tmorin> my take is that we should use the bgpvpn neutronclient extension in tempest test, until someone comes up with a practical drawback of doing so
15:25:16 <tmorin> next topic ?
15:25:24 <tmorin> #topic juno backport
15:25:36 <tmorin> any news on the juno backport, doude ?
15:26:00 <doude> it's on going. I'm able to run it on my lab
15:26:12 <tmorin> good :)
15:26:27 <matrohon> doude : with a driver?
15:26:39 <doude> I started the work from a fork of the Kilo branch and I add to add 3 patches to be able to run it on Juno
15:26:50 <doude> actually, only tested with contrail
15:27:00 <matrohon> doude : ok
15:27:07 <doude> I need to rebase last commut of kilo on it
15:27:14 <doude> the submit reviews
15:27:18 <doude> then submit reviews
15:27:23 <tmorin> ok
15:27:56 <tmorin> very good
15:28:15 <tmorin> how are you handling the python-neutronclient aspects ?
15:29:02 <tmorin> not having any support ?  writing a standalone client inheriting from python-neutronclient ?
15:29:15 <doude> I don't
15:30:07 <doude> python-neutronclient does not follow openstack release
15:30:48 <doude> what do you mean by "handling the python-neutronclient aspects" ?
15:31:17 <tmorin> people deploying openstack juno are likely to also deploy python-neutronclient of the same generation
15:31:32 <tmorin> which won't have the extension framework and won't load our extension
15:31:33 <doude> tmorin: not necessary
15:31:52 <tmorin> using a recent python-neutronclient with an older openstack is not always possible
15:32:12 <doude> actually, it's what I'm doing
15:32:28 <tmorin> we tried that for kilo (use a liberty python-neutronclient), but there were conflicting dependencies
15:32:38 <doude> I run the master branch of python-neutronclient with the OS juno release
15:32:40 <tmorin> ok, it means that in the juno case it works
15:32:53 <doude> conflicting dependencies ?
15:32:59 <tmorin> you run it, but do you also "pip install" it ?
15:33:30 <doude> I install all cli stuff in an isolated python environment
15:33:59 <doude> all cli stuff = python-neutronclient + networking-bgpvpn (just for the cli extension)
15:34:11 <tmorin> if i recall python-neutronclient from liberty has oslo-something >= y that conflict with kilo (kilo having oslo-something <x, with x <y)
15:34:31 <doude> btw, it's not cery confortable to install all networking-bgpvpn just to benefit the CLI extension :(
15:34:36 <tmorin> yes, but that may not work for all deployment approaches
15:35:30 <tmorin> it wouldn't work for the gate I guess
15:35:41 <tmorin> but well we don't have any juno gate anyway
15:35:43 <tmorin> ok
15:35:58 <tmorin> next topic ?
15:36:06 <doude> ok
15:36:29 <tmorin> #topic fuel
15:36:45 <tmorin> for your information, work has been started on a fuel plugin for networking-bgpvpn
15:37:18 <tmorin> https://review.openstack.org/#/q/project:openstack/fuel-plugin-bgpvpn
15:37:53 <tmorin> work kickstarted by michal and enikher
15:38:06 <tmorin> feel free to have a look/contribute !
15:38:18 <tmorin> anything to add ?  enikher ?
15:39:50 <tmorin> next topic then...
15:40:11 <tmorin> #topic drivers
15:40:32 <tmorin> on bagpipe side...
15:40:49 <tmorin> we merged a bugfix on the bagpipe driver yesterday
15:41:07 <tmorin> before that, in december, a lot of cleanups were done in networking-bagpipe
15:41:18 <tmorin> with the addition of a lot of unit test
15:41:23 <tmorin> testS
15:41:48 <tmorin> and we asked to have networking-bagpipe be part of the big tent
15:42:12 <tmorin> we also did what was needed to have clean stable/kilo and stable/liberty branches
15:42:21 <enikher> tmorin: a lot of things happening here on my side sorry I was again afk. I read through the history. Everything fine with me. I will check if we can use neutronclient in tempest
15:42:40 <tmorin> enikher: ok, np :)
15:43:00 <tmorin> anything on ODL driver side ?
15:43:14 <tmorin> I guess not, given who is IRC-active today
15:43:20 <tmorin> next topic ?
15:43:24 <tmorin> #topic heat
15:43:35 <enikher> :-)
15:43:39 <tmorin> enikher: if you are not afk :)
15:43:49 <enikher> Yes this time I am here
15:43:57 <tmorin> i was just curious how your work on heat was progressing
15:44:04 <enikher> Work was finished before xmas already
15:44:43 <enikher> I am very close to contribute my work but there is yet another internal review process.
15:45:00 <enikher> Heat is not core project of my department so I have to align with others :-)
15:45:13 <tmorin> hmm, interesting
15:45:44 <tmorin> the contribution is welcome, independently of what it takes on your side
15:45:50 <enikher> I have the hope to push it this week but lest see
15:46:06 <tmorin> ok, good!
15:46:36 <enikher> I have nothing to add herer
15:46:51 <tmorin> ok, thanks
15:47:01 <tmorin> #topic misc
15:47:05 <tmorin> anything else to add?
15:47:11 <enikher> port assoc
15:47:16 <tmorin> ah yes
15:47:21 <enikher> and static routes
15:47:22 <tmorin> and static routes
15:47:27 <enikher> yes
15:47:29 <enikher> :-)
15:47:48 <tmorin> on static routes I have to send feedback on a proposal sketched out my Jan
15:47:54 <tmorin> I hope to do it this week
15:48:07 <enikher> that would be nice.
15:48:15 <tmorin> enikher: anything particular on port association that you would have in mind ?
15:48:26 <enikher> yes sorry I am writing
15:48:51 <enikher> we good a go to work on this now. And this time we can contribute directly.
15:49:19 <enikher> The plan is to publish things at the end of this month. Dev work did not realy start yet
15:49:39 <enikher> question is who else is willing to give support here?
15:50:14 <tmorin> I am certainly interested in reviewing and discussing the proposals
15:50:47 <enikher> ok, good to know. I just wanted to include everyone interested
15:51:36 <enikher> that was it from my side
15:52:58 <matrohon> enikher, If there is a will to implement it from your side and it doesn't break current code, the implementations of those enhancements are welcomed :)
15:53:39 <enikher> sure this is the intent
15:53:53 <tmorin> one last very minor thing: #openstack-net-bgppvn is now logged
15:53:55 <tmorin> #link http://eavesdrop.openstack.org/irclogs/%23openstack-net-bgpvpn/index.html
15:54:18 <tmorin> feel free to drop by there between weekly meetings
15:55:53 <tmorin> ok, it seems that we're out of things to discuss
15:55:58 <tmorin> and mostly out of time anyways
15:56:09 <enikher> ok
15:56:13 <tmorin> thanks everyone!
15:56:18 <enikher> thanksQ
15:56:24 <tmorin> bye bye
15:56:29 <tmorin> #endmeeting