14:00:40 #startmeeting watcher 14:00:41 Meeting started Wed Feb 17 14:00:40 2016 UTC and is due to finish in 60 minutes. The chair is acabot. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:42 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:44 The meeting name has been set to 'watcher' 14:00:46 hello 14:01:02 o/ 14:01:03 hi 14:01:06 hi 14:01:15 hi 14:01:21 o/ 14:01:28 \o 14:01:34 o/ 14:01:38 #info agenda for today https://wiki.openstack.org/wiki/Watcher_Meeting_Agenda#02.2F17.2F2016 14:01:47 o/ 14:02:02 \o 14:02:10 #topic Announcements 14:02:34 a couple of annoucements regarding discussions on the mailing list 14:02:44 o/ 14:02:59 #info No Open Core #link http://lists.openstack.org/pipermail/openstack-dev/2016-February/085855.html 14:04:07 the important thing in this thread is : "It should have a fully-functional, production-grade open source implementation. If you need proprietary software or a commercial entity to fully use the functionality of a project or getting serious about it, then it should not be accepted in OpenStack as an official project." 14:04:25 interesting 14:04:34 Agreed 14:04:37 I think that's in line w/ what we discussed 14:04:46 its important to deliver Watcher as a "full" project 14:04:53 acabot: and what is the concern with Watcher 14:04:56 ? 14:04:59 jwcroppe I agree 14:05:09 what we miss so far? 14:05:15 I don't believe we have any concern 14:05:26 its more an info that we should keep that in mind 14:05:45 but we are fully compliant right now 14:05:47 The concern in that thread was Poppy, which tries to generalize an API for CDNs 14:05:56 There are no free/open CDNs out there 14:06:10 Oh that makes sense 14:06:40 So the only way to actually use Poppy (and test it) would be to pay for a closed CDN provider 14:06:55 same applies to scoring module - it will not depend on TAP or any other external software 14:07:16 Agreed 14:07:25 acabot: I agree with this overall - I think we just need to continue to be mindful of this as we make assertions about what Watcher 'requires' to be used 14:07:43 jwcroppe : +1 14:08:04 #info roadmap for Nova scheduler #link http://lists.openstack.org/pipermail/openstack-dev/2016-February/086371.html 14:08:36 its a very long thread but interesting to see how the nova scheduler will evolve during mitaka and newton 14:08:57 and how we can take advantage of these evolutions in Watcher 14:08:59 jwcroppe: yes. The other concern with Poppy is that it relies on Cassandra, which only works with the Oracle JDK. 14:09:03 I will read it as soon as I get off my phone ;-) 14:09:45 jwcroppe: so requiring the use of a closed product is another bad thing to do 14:10:03 #info our talks are open to vote until tomorrow, so if you didnt took the time to give a +3, please do it 14:10:13 Oh I see even though it is open source 14:10:53 any other announcement ? 14:11:26 link to talk to vote for please? (if someone has it handy) 14:11:26 edleafe: agree 14:11:36 I am working on getting us to intel for the next mid cycle 14:11:37 jwcroppe : is tpeoples off this week ? 14:11:56 #link https://www.openstack.org/summit/austin-2016/vote-for-speakers/Presentation/7108#link https://www.openstack.org/summit/austin-2016/vote-for-speakers/Presentation/7113 14:11:58 sballe: yes, but if they could convince the Cassandra people to support OpenJDK, then it would be fine 14:12:16 edleafe: makes sense 14:12:36 #topic Review Action Items 14:13:36 we have a lot of specs open 14:13:55 acabot: on my list to do today 14:14:02 Same here 14:15:01 #action sballe jwcroppe review https://review.openstack.org/#/c/273556/ https://review.openstack.org/#/c/276586/ https://review.openstack.org/#/c/274688/ https://review.openstack.org/#/c/266446/ https://review.openstack.org/#/c/278516/​​ 14:15:06 my action items are still waiting - struggling to setup devstack env. I'm putting all my time there... 14:15:44 tkaczynski : ok could you please use the openstack-watcher irc channel for support ? 14:16:08 acabot ok 14:16:10 or openstack-dev irc channel 14:16:17 as it is logged, its better if we need to give the same explanations later 14:16:54 we also need specs for high priority BPs 14:17:21 #action tkaczynski submit a drafted spec for https://blueprints.launchpad.net/watcher/+spec/scoring-module 14:17:55 jwcroppe: tpeoples had an action to submit a spec for https://blueprints.launchpad.net/watcher/+spec/cluster-model-objects-wrapper, any update on this ? 14:18:15 He is out today - I'll check with him tomorrow (and let's make an action for him) 14:18:37 #action tpeoples submit a spec for https://blueprints.launchpad.net/watcher/+spec/cluster-model-objects-wrapper 14:19:09 #action acabot submit a drafted spec for https://blueprints.launchpad.net/watcher/+spec/select-destinations-filter 14:19:41 we received an email from alexstav regarding https://blueprints.launchpad.net/watcher/+spec/watcher-overload-underload 14:20:15 he will submit a spec for this BP as we had to explain the contribution process 14:20:30 #action alexstav submit a spec for https://blueprints.launchpad.net/watcher/+spec/watcher-overload-underload 14:21:19 on Watcher side 14:21:42 #info Delete related actions when deleting an action plan has been merged 14:22:12 there are bug fixes in review 14:22:43 gzhai3 : could you do some code reviewing this week ? 14:22:49 sure 14:22:53 acabot: watcher-dashboard initial pachset will be merged today 14:23:23 dtardivel : ok tpeoples had an action to review it 14:23:50 jwcroppe : can we merge the initial commit today ? 14:24:20 I think so 14:24:42 watcher-dashboard plugin will be now automatically deployed within horizon, in devstack. 14:24:50 #action dtardivel merge watcher-dashboard initial commit today 14:25:40 #topic Blueprint/Bug Review and Discussion 14:26:12 new watcher-dashboard HTML documentation will be also set online soon 14:27:03 #info regarding the SSL support we discussed last week, there are 3 open bugs (2 on devstack, 1 on watcher) 14:28:04 vtech: did you already start implementing ZHAW strategy ? 14:28:25 yes, by reading these bugs, you will be able to apply workaround on your devstack env, waiting for official fixes 14:28:40 we did not start the watcher implementation yet 14:28:50 we had a variant designed for openstack before 14:29:04 we have looked a little at the watcher supports and the example strategy 14:29:17 but we have not specifically started the implementation yet 14:29:25 it is going up the prio list 14:29:28 seanmurphy : ok thx 14:29:33 we should have something for next week 14:29:44 i will discuss the process with you/jed outside this chat 14:29:59 seanmurphy : I think we will be able to merge the spec before end of week so you can start implementing 14:30:07 ok 14:30:11 thx 14:30:16 pwd 14:30:38 cool 14:31:15 edleafe : any update from the nova team regarding the BP https://review.openstack.org/#/c/276840/ 14:31:32 no, not anything yet 14:31:38 ok 14:31:54 seanmurphy: can you have discussion on openstack-watcher IRC channel please. It will benefits everyone 14:32:05 for sure 14:32:36 acabot: it probably won't see much action until Newton 14:32:47 edleafe : yes I suppose 14:33:25 a quick look at bugs targeted for Mitaka https://launchpad.net/watcher/+milestone/mitaka-3 14:34:00 edleafe: it'd be interesting to get bauza's take on 14:34:20 many of them are unassigned but I correctly understand the release schedule, we can fix them after mitaka-3 milestone 14:35:16 #topic Open Discussion 14:36:28 I had an action after the mid-cycle to share the actions on the mailing list but I didn't find any other project doing it... 14:36:45 do you think we should do it anyway ? 14:37:20 acabot: hmm, maybe not 'actions' explicitly - but a quick summary of discussions ? 14:37:46 jwcroppe: +1 14:37:46 ...even a quick pointer to our etherpad 14:37:57 jwcroppe : other projects share the etherpad 14:38:34 ok so a link to the etherpad and a quick review of decisions made ? 14:38:46 +1 14:38:58 +1 14:39:11 jwcroppe: sorry, on a team call now 14:39:26 #action acabot share a link to the etherpad and a quick review of decisions made on the ML 14:39:31 i would like to have some idea how it went/discussions/conclusions etc 14:39:43 jwcroppe: but since he was in on the original discussion with alaski, I think it won't be too contentious 14:39:53 seanmurphy : did you look at the etherpad already ? 14:40:08 no - i was not aware it existed tbh 14:40:32 #info mid-cycle etherpad #link https://etherpad.openstack.org/p/mitaka-watcher-midcycle 14:42:01 thanks 14:42:44 we had a discussion this morning with jed56, dtardivel, vmahe & vincentfrancoise about updating the class hierarchy in Watcher strategies to deal with input parameters (threshold) and output parameters (efficacy indicator) 14:43:18 it will impact 2 open BPs https://blueprints.launchpad.net/watcher/+spec/efficacy-indicator & https://blueprints.launchpad.net/watcher/+spec/optimization-threshold 14:43:36 how to refactoring? 14:43:46 acabot: what about scoring module? will it impact it as well? 14:43:50 vmahe is working on an updated class diagram that will be part of the efficacy indicator spec 14:43:59 tkaczynski : not at all 14:45:07 the idea is to create a new class between "base strategy" and "strategy" to hold a goal and a list of efficacy indicators 14:45:23 it will be a lot clearer with a schema ;-) 14:45:55 there should be a base class for all strategies that achieve the same goal 14:46:06 those strategies would also share the same efficacy indicators 14:46:26 so that we can compare them 14:46:36 the hierarchy would be baseStrategy->energyStrategy->dummy 14:46:43 I think is more easy to discuss that around a specification :) 14:46:51 +1 14:47:07 +1 14:47:32 I think acabot just want to let you know that the possible impacts :) 14:47:40 +1 14:47:42 jed56 : +1 right my point is to say that the BP regarding the threshold will be impacted 14:48:23 so gzhai3 please review the efficacy indicator spec as soon as vmahe has submitted the schema 14:48:31 I will also write a new bp saying that the goal will be returned by each strategy or each common parent strategy 14:48:34 acabot: ok 14:49:12 so we will have 3 BPs correlated 14:50:07 but il will be much easier and cleaner to hold input/output parameters of strategies 14:51:37 sballe : any update on the cross projects meeting ? 14:52:31 sballe: ^ 14:52:36 No they mostly discussed how to do user policies across the projects 14:52:49 ok 14:52:57 meeting with congress ? 14:53:16 The idea is to not have user be admin to be able to do more advanced things 14:53:22 No progress on congress yet 14:53:33 Will follow with them this week 14:54:08 I need to figure out who to talk to 14:54:15 ok 14:54:37 any other subject ? 14:56:09 Not from me 14:56:46 sballe: did you get an update for the lab submission through INTEL ? 14:57:23 So it turns out it is for the intel booth 14:57:35 nothing else from me 14:57:37 I submitted it again 14:57:43 :-) ok 14:58:27 But this time only a demo 14:58:27 No lab 14:58:31 acabot: ^^ 14:58:32 #info vmahe submitted a new BP https://blueprints.launchpad.net/watcher/+spec/get-goal-from-strategy 14:59:08 ok thank you 14:59:24 have a good day (night for gzhai3 ;-)) 14:59:26 bye have nice (day,afternoon, .. ) 14:59:30 Bye 14:59:31 by 14:59:34 bye 14:59:36 bye guys 14:59:37 bye ;) 14:59:44 bye 14:59:45 #endmeeting