16:03:46 #startmeeting tacker 16:03:47 Meeting started Thu Sep 3 16:03:46 2015 UTC and is due to finish in 60 minutes. The chair is sridhar_ram. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:03:48 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:03:51 The meeting name has been set to 'tacker' 16:03:58 #chair bobh 16:03:59 Current chairs: bobh sridhar_ram 16:04:05 s3wong here ? 16:04:18 #topic Announcements 16:04:18 hi 16:04:37 Agenda #link https://wiki.openstack.org/wiki/Meetings/Tacker#Meeting_Sep_3.2C_2015 16:05:16 if any last minute items .. please speak up now 16:05:33 looks good 16:05:51 Switch over to master is now complete..! 16:05:58 See #link http://lists.openstack.org/pipermail/openstack-dev/2015-September/073447.html 16:06:23 Follow latest install instruction for master based tacker - #link https://wiki.openstack.org/wiki/Tacker/Installation 16:07:01 As discussed in the midcycle .. Tacker dev process is getting documented in this review #link https://review.openstack.org/219012 16:07:28 Please review when you get a chance ...! 16:08:02 sridhar_ram: what happens to the patchsets in flight which are being tested with kilo branch 16:08:53 sripriya: unfortunately .. they need to be retest w/ latest master based tacker 16:09:09 sridhar_ram: ok 16:09:21 sripriya: there is no other way around .. hopefully this is one-time pain that we never need to visit again 16:09:37 sridhar_ram: agree 16:10:13 Folks - don't trip yourself using your existing tree .. it is worth a switch to a clean master based devstack 16:10:34 lets move on.. 16:10:47 hello 16:10:50 s3wong: hi, morning... good timing 16:11:02 #topic Experiment Features 16:11:15 #topic Experimental Features 16:12:46 We are entering a phase we will be trying new areas to expand to the intended scope of Tacker .. 16:12:56 ... to be the ETSI MANO NFVO + VNFM .. 16:13:51 obviously the areas in the NFVO space are still evolving fast - both in the standards side and in the dependent components .. 16:14:27 One thought is to allow these new features enough space to incubate within Tacker 16:14:49 The proposal I've is to follow the origin neutron's experimental feature model 16:15:10 but not much work there 16:15:31 neutron advanced services like fwaas, vpnaas, etc were once experimental before they graduate into mainstream features 16:15:52 vkss: sorry, I didn't get it... what you mean ? 16:16:39 I'm proposing the following two new experimental features - (a) SFC (b) NFVO... 16:16:52 the advanced services there are very basic 16:16:59 ok 16:17:17 sridhar_ram: Is there a link that describes the neutron model? 16:17:21 we can allow these feature to evolve as experimental features .. which means we don't guarantee backward compatibility 16:18:40 bobh: I don't remember they are explicitly written down .. I'll try look if such a thing exists 16:19:22 vkss: yeah, they kinda look basic now.. but once it was outside the original basic l2 + l3 boundary of neutron... 16:19:24 sridhar_ram, how u re proposing to start with? 16:19:25 bobh, sridhar_ram: it was done via documentation (Neutron official APIs) 16:20:12 once they were split into their own repo, Neutron --- being integrated release --- do NOT put FWaaS, LBaaSv2 and VPNaaS as part of the integrated release list of libraries 16:21:03 I'm open for ideas .. I'm of the opinion we need to create these new features within tacker repo itself .. just mark in the documentation, like s3wong said, that these are deemed 'experimental' 16:21:28 sridhar_ram: I am NOT advocating creating separate repo. Tacker isn't big enough to do that yet 16:21:34 I am just stating how 16:21:54 'experimental' was handled back in Neutron old days, and how it is much more refined on the new day 16:22:03 i have work on fwaas also can help with vpnaas 16:22:18 s3wong: agree, we still need the code base to evolve in the correct way ... 16:22:34 but we need to work on details how to intergrate with tacker 16:23:25 s3wong: so these would be additional API calls that are documented as experimental 16:23:43 vkss: sorry if I got this wrong, it is not about Tacker integrating w/ fwaas or vpnaas ... it is just about the development model for new bleeding edge features\ 16:24:00 bobh: well, technically we don't have any release process and release document, so it is up to how we want to define 'experimental' 16:24:34 sridhar_ram, ok will talk to u after meeting 16:24:35 s3wong: right - just trying to avoid the inevitable "this used to work and now it doesn't" 16:24:44 bobh: and since we are also technically not an OpenStack big tent project, release management doesn't really care to give us a fixed guideline :-) 16:25:01 bobh: s3wong: I think documentation is a fine way to communicate this.. 16:25:05 s3wong: :-) There are benefits to being outside the tent 16:25:30 s3wong: agree, however we need to operator we are indeed a proper project under the governance tent :) 16:25:38 s/operator/operate/ 16:26:37 first feature we can try this approach could be what Tim Rozet has done for SFC ... 16:26:52 sridhar_ram: yeah, I believe 'experimental' back in Neutron old days is just documentation (though I could be wrong -- but I never saw anything in code that mark anything experiemental 16:27:00 we need to enable such "relevant" experiments in Tacker .. for us to trive 16:27:43 sridhar_ram, +1 16:28:07 vishwanathj: thanks 16:28:39 any further thoughts on this ... ? 16:29:11 documentation of the capabilites would be good to have 16:29:21 but I know it's still very early 16:29:37 so not much to go on 16:30:05 bobh: yeah, we need to catchup on the docs ... 16:30:14 +1 16:30:36 sripriya: in fact MANO api work needs to end up in docs.openstack.org 16:31:17 sridhar_ram: agree, it is a pending task as part of work items in api spec 16:31:40 bobh: point taken, we can make sure to communicate these features as experimental until we have something in the docs.. 16:31:57 in fact we have a devref in the source tree mentioning them as such 16:32:05 *we can have 16:32:51 anyway I'll send an email to ML for any further comments .. we can continue further thoughts over in that thread 16:33:06 lets move on.. 16:33:29 #topic Tacker-SFC 16:33:38 trozet: ping 16:34:18 I can provide a quick update on this... 16:34:37 some of the Tacker members attended the last OPNFV-SFC meeting 16:35:14 we discussed how we can bring in what Tim has experimented in his github to do Tacker --> ODL-SFC 16:35:36 Tim is willing to contribute that back into tacker repo 16:36:08 I expect a tacker-spec from Tim .. and we should plan to review / guide that effort as one of the first experimental features in Tacker 16:36:32 sounds like a good plan 16:36:49 bobh: thanks 16:37:18 sridhar_ram: i will look out for the spec and help with that 16:37:32 sripriya_: cool, thanks 16:39:04 we need to guide that effort such that the SFC APIs are generic enough (it need not be perfect - remember it is experimental ;-) .. and the implementation proposal should gel well with the rest of the Tacker codebase (perhaps as a proper tacker extension) 16:39:25 BTW, for some understandable (!) reasons, SFC invokes lots of strong opinions... 16:40:20 But I would suggest we bias ourselves to get "something" done... rather than going in circles 16:41:21 once the spec lands ... we should plan to have Tim join this call for further questtions.. 16:41:29 anything else on SFC ? 16:42:18 #topic Liberty Release Plans 16:42:46 team - we need to zero in on a date for Liberty Release ... 16:43:19 I'm looking for inputs from Prasanth (for auto-scaling), Bob (for monitoring) and Sripriya (for MANO api + basic unit-test f/w) 16:44:36 would, say Sept 25th, be a good date for your efforts to wrap up.. 16:44:46 That should work 16:44:55 sridhar_ram: ack 16:45:36 prashantD: how abt you ? 16:45:49 bobh: sripriya_: thanks 16:46:41 alright - lets go w/ Sept 25th .. which means you need to wrap few days before that ;-) 16:47:30 Here we go .. we now have a release page for Tacker #link https://wiki.openstack.org/wiki/Tacker/Releases 16:48:14 any other question on the release ? 16:48:39 lets open the floor for gen discussion... 16:48:53 #topic Bugs / Reviews / Open Discussion 16:49:34 if you've something to discuss on Tacker .. please bring up 16:49:44 Quick question on PyCharm - anyone using it behind an HTTTP proxy and connecting to github? It claims to work when I test the proxy settings but fails to pull from github 16:49:59 if I'm outside of the corporate network it works fine 16:50:38 bobh: heavily use pycharm, but never used its git integration features 16:51:00 sridhar_ram: so you run git from command line and import the project into PyCharm? 16:51:12 bobh: yep 16:51:55 back to tacker .. just an Fyi, tosca-parse is now a separate component available outside heat-translator 16:52:03 #link https://github.com/openstack/tosca-parser 16:52:47 I'm planning to look into this .. we need tosca-nfv parser to switch over at some point 16:53:17 anything else? .. we can wrap up early if not 16:54:05 lets wrap .. thanks folks 16:54:10 #endmeeting