14:59:31 <jimbaker> #startmeeting craton 14:59:33 <openstack> Meeting started Mon Jun 27 14:59:31 2016 UTC and is due to finish in 60 minutes. The chair is jimbaker. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:59:34 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:59:37 <openstack> The meeting name has been set to 'craton' 15:00:48 <jimbaker> #topic agenda setting 15:01:13 <jimbaker> so there may be a problem with openstack wiki, which is where we have the agenda posted 15:01:13 <sigmavirus24> o/ 15:01:13 <Zucan> o/ 15:01:13 <jimbaker> usual place: https://wiki.openstack.org/wiki/Meetings/craton; but also we are using https://etherpad.openstack.org/p/craton-meetings 15:01:13 <sigmavirus24> jimbaker: you mean the fact that new accounts can't be created? 15:01:32 <jimbaker> sigmavirus24, that is a well-known issue 15:01:34 <sigmavirus24> jimbaker: yeah most projects use etherpads 15:01:34 <Zucan> The agenbda is working for me: https://wiki.openstack.org/wiki/Meetings/craton 15:01:40 <Zucan> Remove the semicoplon off the url 15:02:11 <jimbaker> Zucan, sure: https://wiki.openstack.org/wiki/Meetings/craton - always ensure a space after a url :) 15:02:19 <sulo> o/ 15:02:19 <Zucan> heh heh :) 15:03:00 <jimbaker> anyway, i think we need to switch over as well. the wiki is going away and we usually use etherpad for this purpose anyway 15:03:38 <turvey> o/ 15:03:56 <cmspence> o/ 15:04:03 <jimbaker> sigmavirus24, so does https://etherpad.openstack.org/p/craton-meetings work for you? 15:04:27 <sigmavirus24> jimbaker: absolutely :) 15:05:04 <jimbaker> awesome 15:07:06 <jimbaker> every one of those points just makes me feel better about it ;) 15:09:07 <jimbaker> fortunately the only thing we have put on the wiki up until now is agendas. which we now stop doing 15:09:07 <jimbaker> oops i just realized i was looking at two channels (#craton as well) 15:09:07 <jimbaker> so sigmavirus24 posted on #craton: "the wiki is hosted on an ancient operating system with ancient PHP installed and an unsupported version of the wiki software iirc" (i occasionally see bouncer lag so i was assuming the case here) 15:09:08 <jimbaker> sorry about that 15:09:08 <sigmavirus24> oh yeah, I should have posted that in here :/ 15:09:11 * sigmavirus24 is distracted and under-caffeinated 15:09:39 <jimbaker> so any objection to us stopping the use of the wiki? we will need to get a small patch done against the overall meeting site 15:10:23 <sulo> sounds good to me 15:10:31 <jimbaker> seeing no passionate people for antiquated php... 15:10:31 <turvey> silence sounds like consent to me 15:10:35 <jimbaker> indeed 15:10:48 <cmspence> same, especially since most of us were having issues with it 15:10:52 <Zucan> +1 to etherpad only :) 15:11:13 <sarob> sorry im late 15:11:23 <sarob> freenode wouldnt let me on 15:11:24 <jimbaker> #action jimbaker will work with sarob to get the link appropriately patched from old bad wiki to etherpad 15:11:50 <sarob> ill catch up on log 15:12:02 <jimbaker> (that should have had more context - but there's only one wiki page - the meeting wiki) 15:12:52 <jimbaker> sarob, we are just going to start using https://etherpad.openstack.org/p/craton-meetings going forward. just need to do some cleanup there, but seems reasonable place to start 15:13:24 <jimbaker> since everyone is here and talking about wikis is completely boring... i suggest we now talk about meetings :) 15:13:38 <jimbaker> #topic midcycle meetup 15:13:57 <sarob> all caught up 15:14:10 * sarob just needs coffee :/ 15:14:10 <jimbaker> at the openstack summit, we agreed that our midcycle would be at the operators meetup midcycle 15:14:25 <sarob> +1 15:15:03 <jimbaker> that midcycle has been set - it will be hosted by bloomberg in NYC on aug 25 & 26 15:15:25 <jimbaker> in addition, openstack east is happening the preceding two days, also in NYC 15:15:46 <sarob> do you have the etherpad link? 15:15:57 <sarob> im looking and not finding 15:16:00 <sulo> we are planning for craton specific meet during the same ? 15:16:27 <jimbaker> sulo, correct - i suggest we do take advantage of the two meetings for our midcycle 15:16:59 <sulo> ok. can we post this on openstack-dev to attract more interested parties ? 15:17:00 <jimbaker> many operators are planning to do the same. more importantly this gives us a good meeting location 15:17:05 <jimbaker> sulo, absolutely 15:18:13 <jimbaker> here's the link to the operators midcycle confirmation - http://lists.openstack.org/pipermail/openstack-operators/2016-June/010788.html 15:18:34 <jimbaker> and the corresponding openstack east: http://www.openstackeast.com/ 15:18:57 <sulo> jimbaker: also, will we need to get a place/room allocated for us ? .. not sure how that works ? 15:19:42 <jimbaker> sulo, openstack events that have any dev focus will have some breakout space. but we should confirm specifics 15:19:53 <sulo> ok 15:21:41 <jimbaker> so here's what we can focus on: the first two days will be focused on craton dev midcycle correction; the two days at the operators can then focus on taking that plan to operators for fine tuning 15:22:43 <jimbaker> obviously it will be a mix during the 4 days. but this should have enough time to help us work on completing phase 2 of work, and move on to the next phase 15:23:28 <jimbaker> so any objections here? hopefully this sounds like a good plan in terms of having our meetup 15:24:25 <sulo> +1 15:24:38 <Zucan> +1 15:24:38 <sigmavirus24> Can y'all take good notes for those of us who won't be there? 15:25:11 <jimbaker> sigmavirus24, certainly. we will make certain it's captured in an etherpad, which we have done a decent job to date i believe 15:25:18 <sulo> i think we should record most of the sessions if we can 15:25:52 <jimbaker> sulo, if possible, that sounds good. but etherpad should be the principal doc 15:26:08 <sulo> yeah thats fine 15:26:39 <jimbaker> (i refer all the time to https://etherpad.openstack.org/p/Fleet_Management for example, which was our kickoff meeting for introducing this project) 15:26:57 <jimbaker> #action jimbaker will announce our midcycle meetup on openstack-dev 15:27:39 <sarob> cool 15:28:18 <jimbaker> in terms of trip planning - please make travel plans such that you are available tues morning through fri 5p 15:28:23 <jimbaker> in NYC 15:28:55 <jimbaker> if you want to attend the full meeting 15:29:09 <jimbaker> now to far more interesting stuff! 15:29:21 <jimbaker> #topic retrospective and phase 2 planning 15:29:46 <jimbaker> https://etherpad.openstack.org/p/craton-phase2-planning 15:30:59 <jimbaker> we still need to figure out specific dates; however, the intent of the craton team is to complete audit & remediation workflow support in time for barcelona 15:31:25 <jimbaker> phase 2 focuses on audit workflows, plus extending the inventory support we have done to date 15:31:56 <jimbaker> (the diff with remediation workflows is that scheduling of work is different) 15:32:39 <jimbaker> i believe we should use FLT005 from https://specs.openstack.org/openstack/openstack-user-stories/user-stories/proposed/fleet-management.html as the motivating user story 15:33:13 <jimbaker> specifically the support of the openstack ansible security role 15:34:23 <jimbaker> this requires capture of the security audit data into inventory, possibly extending the support in that role for such capture 15:35:25 <jimbaker> other work items we have identified focus on integration, such as with nova and with tooling like oneops 15:36:23 <jimbaker> sarob, any update on the meeting planned with VitaliyZ? 15:36:40 <sarob> he has time tommorrow 15:36:48 <sarob> what works for y'all? 15:37:28 <jimbaker> sarob, i believe 1500 UTC probably works best - so same time as today 15:37:51 <jimbaker> i know sulo would like to attend 15:38:06 <sulo> yes, this time is good for me 15:38:33 <jimbaker> (context here: this is with respect to oneops integration. we are exploring this as an interesting integration point with an open source CMDB/governance/workflow platform) 15:39:22 <jimbaker> sarob, can you add the meeting to https://etherpad.openstack.org/p/craton-meetings once we have the specifics? 15:39:51 <sarob> yeah will do 15:39:55 <jimbaker> thanks 15:40:44 <jimbaker> i encourage everyone to look at the planning we are doing in https://etherpad.openstack.org/p/craton-phase2-planning 15:41:56 <jimbaker> i do think it matches what we discussed with stakeholders; prioritization input will be valuable! 15:43:00 <jimbaker> if you have questions, let's address those at the end of today's meeting 15:43:10 <jimbaker> #topic current work 15:43:51 <jimbaker> sulo, anything you want to discuss with the issues you have been working on? 15:44:25 <sulo> i am finishing up the ansible-inventory api endpoint .. then working on adding few missing rest endpoints for project and user creation 15:44:50 <sulo> and somethign interesting that we discussed last week 15:45:17 <sulo> is more work on modes 15:45:29 <sulo> for integrating network devices etc 15:45:43 <sulo> some ideas captured here https://docs.google.com/document/d/1N3f_H3bjZ1a08wkxRLCMt5ANXRGmFAOEUZ_mtI3BYxw/edit 15:46:22 <jimbaker> sulo, what additions if any do we need to our underlying inventory modeling? 15:46:25 <sulo> one interesting bit being modeling in racks 15:46:41 <sulo> so we can do variables resolution on rack 15:51:28 <jimbaker> sulo, do we need to break this out; or can we use the label support for such var resolution? 15:51:48 <sulo> i think we will need to break this out 15:51:55 <sulo> particularly because it nests stuff like ip subnets and policies for per rack basis 15:51:55 <jimbaker> perhaps we should bring back label DAG support? 15:52:02 <sulo> it would be nice to get this done though labels 15:52:06 <jimbaker> VitaliyZ raised the same issue, which is why they came up with the CI approach 15:52:08 <jimbaker> i believe labels are the same thing for us 15:52:08 <sulo> yeah that makes sense 15:52:09 <jimbaker> anyway, i 15:52:25 <jimbaker> am glad to see good progress on network device modeling questions - that will definitely drive good dev work 15:53:22 <sulo> in that doc .. (c) is the property that might be difficult to track though labels .. but we probably need to get a working set going to get a good idea 15:54:13 <jimbaker> (c) looks like a good test case 15:54:13 <sulo> right now though, we are doing that though variables .. just dont have api to get that data speciically 15:54:13 <sulo> so yeah .. its a good test case 15:54:13 <jimbaker> sulo, do we have data for the network devices test bed? 15:54:13 <sulo> yeah i have been playing with Rackspace specific data 15:54:14 <sulo> i will ask for more generic data from hvprash_ team .. or anyone else here who want to pitch in 15:54:14 <sarob> i gots to jump a bit early 15:54:14 <jimbaker> ack, we need to make this available 15:54:14 <jimbaker> sarob, take care 15:54:22 <sarob> patch for openstack project here https://review.openstack.org/#/c/332470/2 15:54:31 <jimbaker> 5 min to go 15:54:36 <jimbaker> #topic questions? 15:55:03 <sarob> im going to set no-op for jobs 15:55:19 <sarob> until we have some more bits in the repos 15:55:23 <sarob> cheers 15:55:33 <jimbaker> sarob, cool, and thanks! 15:55:39 <sulo> sarob: cool. thanks 15:56:06 <jimbaker> sulo, let's discuss more modeling shortly on #craton 15:56:16 <sulo> ok 15:56:52 <jimbaker> the other question i would like to raise with you there is with respect to access contexts (!) a term i just came up with for managing the alternatives we need to work with various devices 15:57:04 <jimbaker> including say the VPN issue we are facing 15:57:12 <jimbaker> in the osic ref cluster 15:57:44 <jimbaker> anyway, i think we can just move over discussion and follow up questions at this point to #craton 15:58:22 <jimbaker> #endmeeting