15:01:10 #startmeeting TSC weekly meeting 15:01:10 Meeting started Tue Nov 8 15:01:10 2016 UTC. The chair is tallgren. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:10 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:01:10 The meeting name has been set to 'tsc_weekly_meeting' 15:01:17 #info Uli Kleber 15:01:19 #info Bryan Sullivan 15:01:23 #info Morgan Richomme 15:01:26 #chair rpaik ChrisPriceAB ulik 15:01:26 Warning: Nick not in channel: ulik 15:01:26 Current chairs: ChrisPriceAB rpaik tallgren ulik 15:01:35 #info Tapio Tallgren 15:01:37 #info David McBride 15:01:44 #chair dmcbride 15:01:44 Current chairs: ChrisPriceAB dmcbride rpaik tallgren ulik 15:01:52 #info Jack Morgan 15:01:58 #info Stuart Mackie 15:02:03 #info Edgar StPierre 15:02:11 #info Frank Brockners 15:02:16 #chair uli-k 15:02:16 Current chairs: ChrisPriceAB dmcbride rpaik tallgren uli-k ulik 15:02:21 #info Lincoln Lavoie 15:02:27 #info Bin Hu 15:02:47 we now have a quorum 15:03:01 I can help for the minutes 15:03:36 #info Tim Irnich 15:03:42 #chair morgan_orange 15:03:42 Current chairs: ChrisPriceAB dmcbride morgan_orange rpaik tallgren uli-k ulik 15:04:46 #info agenda https://wiki.opnfv.org/display/meetings/TSC 15:04:48 #info Tim Rozet(proxy for Dave Neary) 15:05:00 #topic approval of previous minutes 15:05:04 #info Tapio asks about the state of the discussion on the proposed rescuer project. Bin answers that there are actions to close before a creation review 15:05:17 #info no feedback on the minutes thus approved 15:05:24 #topic agenda bashing 15:05:27 #info rprakash 15:05:35 #info Jonas Bjurel (Stand in for Fatih) 15:06:07 #info no other topics proposed 15:06:17 #topic Plugfest/Hackfest 15:06:25 #当局者迷 15:06:29 #undo 15:06:32 #info Julien 15:06:46 #link https://wiki.opnfv.org/display/EVNT/Plugfest+-+Colorado+Release wiki page for the event 15:07:07 #info Trevor Cooper 15:07:16 #link https://wiki.opnfv.org/display/EVNT/Co-located+Plugfest-Hackfest+Planning+Page schedule/planning page 15:07:21 #info 15 companies registered to Plugfest, 45 people 15:07:36 #info 5 companies are providing equipments 15:07:49 #link https://www.regonline.com/OPNFVPlugfestDec2016 registration page 15:08:46 #info community members are encourage to add topics for the schedule page above 15:09:10 #info participants should collaborate with each other on plugfest work planning 15:11:21 Where are the rules of engagement? 15:11:25 #info ROE is part of the registration process 15:12:53 #info there are three rooms available for meeting/breakouts 15:18:22 #info discussion on what can be shared publicly after the Plugfest/Hackfest within 60 days 15:18:22 #info ROE protects results of the plugfest from disclosure, but results from Hackfest are needed for Danube release 15:20:13 #topic Graduation reviews discussion 15:21:32 #info rpaik has started the email discussion, it is ongoing 15:21:55 #action rpaik to setup a discussion during plugfest/hackfest 15:22:49 #link https://etherpad.opnfv.org/p/barcelona-infra-f2f 15:22:52 #topic configuration file and scenario naming 15:23:19 #link https://gerrit.opnfv.org/gerrit/#/c/23725/ 15:23:49 #info jmorgan1 summarized the discussion that started in Barcelona and noted that the current scenario naming convention could be improved 15:24:31 #link https://gerrit.opnfv.org/gerrit/#/c/23727/ 15:25:35 #info jmorgan1 adds that no objections for major feedback received so far from the community 15:26:43 #info TSC approval needed to get a commitment - change will trigger extra work especially on installer projects. But without this evolution CI evolution would not be possible 15:27:37 I might suggest this needs to be discussed a little more. I'm not sure the community is aware this patch changes the naming to be fish based. 15:29:14 More to the point I don't know that we have a backend prototype that renders the scenario.py file to actually be able to say "this wirks" 15:29:19 #info questions from JonasB and frankbrockners asks if this will be scalable going forward esp. with CI evolution 15:29:20 s/i/o 15:30:31 #info frankbrockners reminded that a similar proposal was made for Brahmaputra and later, without success 15:31:37 * ChrisPriceAB realises he was commenting on the wrong patch. (no audio today) 15:31:40 dropping off IRC in a few mins.... 15:33:15 #info a way to ensure the success of CI evolution: maybe indicate that it is a condition to be part of Danube 15:33:45 #info frankbrockners asks if we are forcing people to adopt this proposal for Danube 15:37:32 #info on the previous topic (Plugfest vs Hackfest), I created a new child page "Colorado Hackfest Test Cases", moved the VNF Portability Scope description there, and linked it to the "Co-located Plugfest-Hackfest Planning Page" page. I'll explain the rationale for the change on the list as discussed on the call. 15:39:26 #info jonas brings up concern to cover all architectures in these proposals 15:39:49 jmorgan1: https://gerrit.opnfv.org/gerrit/#/c/23725/2/config/shark.yaml this looks kind of odd to me 15:40:07 #topic Goals & vision for the Danube release 15:40:11 jmorgan1: it looks like a deploy settings file, which I don't think should be in pharos 15:40:36 #action jmorgan1 to collect feedback over then bring proposal back in two weeks for vote 15:41:11 trozet: we only used pharos to push the patch 15:41:22 #info uli-k encourages fbrockners and JonasB to call-in to the INFRA meeting and work on the missing parts. 15:41:42 trozet: it wont be there in the future, these files would be in the securelab git repo i beleive 15:41:57 trozet: please provide inut on what is missing ;) 15:42:15 jmorgan1: i dont get why we need a common scenario file? 15:42:21 jmorgan1: theys hould just live in each installer 15:42:30 jmorgan1: CI shoudl just call the correct file as an argument to the installer 15:42:32 We cannot hide all scenario info in the scenario name 15:43:15 trozet: the idea is to come up with a better strategy to define scenarios as current model doesn't include enough details 15:43:41 jmorgan1: then the strategy should be to come up with a better common naming convention, not standardize the file 15:43:42 If a scenario shall be deployed by more than one installer, the description of the scenario should be stored only once. 15:43:43 trozet: this was the result of our effort to come up with a better strategy 15:44:31 trozet: its not standardized in that each scenario would have its own file and include details (variables) defined in the patch above 15:44:57 trozet: like a better way or more complete way to define what is in the scenario 15:45:09 jmorgan1: so then just fix the naming of the file? 15:45:16 trozet, jmorgan1, let's do that discussion tomorrow in Infra meeting, so we can listen to Morgan 15:45:49 jmorgan1: how does CI knowing the storage is ceph help CI? 15:45:58 jmorgan1: while the scenario names are tuna and salmon lol 15:46:15 jmorgan1: just fix the naming and abandon this. I dont think it provides any value 15:46:24 The info should be in the scenario description file, not in the name. 15:46:31 trozet, jmorgan1, let's do that discussion tomorrow in Infra meeting, so we can listen to Morgan 15:47:24 jmorgan1: the common hardware settings are a good idea, since CI relies on those 15:50:27 #info morgan_orange presents the slides at https://wiki.opnfv.org/download/attachments/2925933/Building%20a%20TSC%20vision%20for%20Danube.pptx?version=1&modificationDate=1478612403468&api=v2 15:52:14 if you are planning to join the release meeting in ten minutes, make sure that you have the correct GTM link: https://www.gotomeeting.com/join/775124285 15:57:54 #info there could be a bottom up approach from WG's, feature projects, etc. 15:58:23 #info IMO we are certainly building reference platforms(s) 15:58:37 #info concern raised on if we are giving up on reference platform 15:58:39 #info I do not understand the assertion that we are not 15:59:45 trozet: thanks for feedback, let's chat more about these proposals 15:59:51 #info suggestion to continue the discussion via mailing list 16:00:06 jmorgan1: let's do it in gerrit and jira 16:00:27 trozet: sure, sounds good. 16:00:37 #endmeeting