14:01:49 #startmeeting sahara 14:01:50 Meeting started Thu Mar 24 14:01:49 2016 UTC and is due to finish in 60 minutes. The chair is vgridnev. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:52 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:54 The meeting name has been set to 'sahara' 14:02:13 #chair elmiko 14:02:15 Current chairs: elmiko vgridnev 14:02:26 #link https://wiki.openstack.org/wiki/Meetings/SaharaAgenda 14:02:35 anyone else here for the meeting? 14:02:39 hi! 14:02:40 hi 14:02:41 ) 14:02:44 \o/ 14:02:45 let's wait a little 14:03:00 o/ 14:03:30 hello/ 14:04:35 #topic News / updates 14:04:49 any news folks? 14:05:03 i'm researching microversion support for api v2, working towards writing a spec 14:05:05 Mostly just reviews for me recently. There are a couple bugs on my radar though that I hope to tackle soon. 14:05:21 i'd also like to propose making the bandit gate voting 14:05:36 I busy with internal tasks, trying to get in the new PTL role 14:05:45 Nightly jobs are added! And working on adding new features to scenario-framework 14:06:05 still preparing the sahara-tempest-api migration to sahara-tests, I think I finally have all the steps, but figthing with gerrit ACLs 14:06:22 elmiko, I am ok with making bandit voting 14:06:34 vgridnev: cool, i'll make a review today 14:06:43 tosky: what ACL you mean? 14:07:33 esikachev: Access Control List; I need also to temporarily enable the creation of new branches for cores, so that we can push the reviews with the imported code 14:07:38 I am still working on refactoring around use_floating_ips as a background task -- finished testing neutron, starting on nova 14:07:40 I'm mostly busy with internal activities, nothing to share (( 14:07:54 and the ACLs are again hosted on git, so I'm trying to understand the rules to avoid a -2 on the first review :) 14:08:26 ok, thanks 14:08:30 do we need to Community Roadmap topic? 14:09:11 any other news? 14:09:30 #topic API v2 progress 14:09:37 #link https://review.openstack.org/#/c/273316/ 14:09:41 pretty slow this last week 14:09:41 #link https://wiki.openstack.org/wiki/Sahara/api-v2 14:10:02 i am helping to on-board a few more red hatters who will work on v2, and researching some of the bigger features 14:10:23 i don't imagine any reviews coming for v2 for at least another week or so 14:10:40 that's all for me 14:10:49 elmiko: after moving tempest api tests to sahara-tests, we can start planning the v2 versions of the tests :) 14:10:58 tosky: +1 14:11:05 oh, one more thing 14:11:08 elmiko, oh, where is your deprecation rodeo etherpad for summit? 14:11:33 #topic Scenario tests releases 14:11:39 i'd like to create a new sahara.service.api package to refactor the way we handle the interaction between api and conductor 14:11:47 should i make a spec to propose this change? 14:12:04 Might be a good idea 14:12:15 tmckay: https://etherpad.openstack.org/p/sahara-newton-summit 14:12:20 crobertsrh: is that a +1 to spec? 14:12:25 elmiko, would be nice, I think 14:12:27 yeah 14:12:32 k, thanks 14:12:41 thanks elmiko 14:13:05 esikachev, tosky , hands for you on Scenario tests releases 14:13:27 tosky: any news about tempest tests? 14:14:01 esikachev: what I wrote above, I needed to figure out an additional step and I'm working on the patches for permissions on gerrit 14:14:12 I just want to say that I have ideas to create new launchpad group for sahara tests drivers, where can add esickachev and tosky 14:14:14 do you want to wait for the merge of tempest api for a first release or do you want to have a first release now? 14:14:32 i am moved all default template in scenario dir from etc/ 14:14:57 tosky: how many time you need? 14:15:04 for moving 14:15:06 esikachev: I fear all the next week 14:15:42 vgridnev: what did you think about it? 14:15:59 vgridnev: and +1 for launchpad group 14:16:15 tosky, esikachev I think that we can release sahara tests along with mitaka release 14:16:29 since it's not added to requirements 14:16:37 global, I meant 14:16:55 ok 14:17:29 ok 14:17:51 tosky: next topic? 14:18:11 #topic Newton summit topics 14:18:11 #topic Newton summit topics 14:18:11 #topic Newton summit topics 14:18:18 #undo 14:18:19 Removing item from minutes: 14:18:21 #undo 14:18:22 Removing item from minutes: 14:18:30 oh, elmiko, do we want to list deprecation topics ahead of time, or just show up with them? on that etherpad, or another? 14:18:33 #link https://etherpad.openstack.org/p/sahara-newton-summit 14:19:02 I still not sure about my attending to summit 14:19:04 tmckay: i'm ok either way, probably best to make a list there if you think it might get forgotten? 14:19:26 hope you can make vgridnev, would be fun to meet face-to-face =) 14:19:39 Yes, our new PTL should certainly be there 14:20:05 I will try to attend 14:21:14 vgridnev, ++, what would be great 14:21:20 I have some ideas to put in there, I think that if will not attend to summit SergeyLukjanov will driver this discussions 14:21:21 sorry, what => that 14:21:56 I will put this topics in ether pad shortly 14:22:56 i added a bunch of stuff there, not sure what we will end up talking about but i tried to capture topics from past summits 14:23:56 #topic Open discussion 14:24:56 Might be worth mentioning that I think we're planning to make the integration tests voting in sahara-dashboard 14:25:10 vgridnev, should we discuss hdp 2.0.6 plugin deprecation and SIE, scenario tests, and doc changes? 14:25:56 so, I have some ideas that I wanna bring to sahara, so. We have a bunch of blueprints that are in unclear state, do we want to have some kind of sahara-drivers-meeting to make cleanup of that? 14:26:05 crobertsrh: +1 14:26:12 crobertsrh, agreed 14:26:20 vgridnev: that sounds good to me 14:26:33 vgridnev, ++, I was just thinking about bp cleanup the other day 14:26:37 that is probably a great idea, +1 14:26:58 vgridnev, also, there are wiki pages associated with sahara that are probably way out of date or not relevant anymore 14:27:09 good idea, the list of blueprints include many which are implemented already, or totally useless now 14:27:13 like, the sequence diagrams for EDP :) 14:28:18 tmckay, about HDP 2.0.6: I think that we can remove hdp206 from sahara-tests only in branches where it deprecated or removed 14:28:38 vgridnev: tmckay https://review.openstack.org/#/c/297173/ 14:28:42 #link https://review.openstack.org/#/c/297173/ 14:28:58 about SIE I think that we can do removal too, but I think that we can super short spec to discuss that 14:29:41 vgridnev: also when it's deprecated? I thought we simply removed HDP 2.0.6 from the default list in Mitaka, but it's still working 14:29:55 vgridnev, ++ on SIE. to clarify, then, you are saying that we should keep scenario tests for stable/liberty, stable/mitaka, etc? 14:30:01 agreed, if that is the case 14:30:44 esikachev, could please say the current state of HDP 206 on CI? 14:30:45 esikachev, thanks 14:31:00 vgridnev: deprecated from all releases 14:31:10 tosky, yes, CR this morning to remove in newton 14:31:25 tosky, https://review.openstack.org/#/c/297001/ 14:31:37 #link https://review.openstack.org/#/c/297001/ 14:32:08 esikachev: do you mean non voting or still running? 14:32:43 tosky: i mean, removed from all ci scripts 14:32:49 about blueprints, there are two options I think: additional topic in agenda, or how it's done in horizon, for example, separate meeting 14:33:08 esikachev: but it should run still for liberty and mitaka (kilo will be out of support soon) 14:33:19 I think that the separate meeting will be overkill for us 14:33:51 tosky: in liberty actual version is ambari 2.2 14:34:04 esikachev: no, hdp is still working 14:34:10 vgridnev: ^ 14:34:27 hdp 2.0.6 is even enabled by default there 14:34:35 vgridnev, ++ I think an agenda item will be enough, and normal channel IRC discussion 14:34:39 and even if it's disabled by default in mitaka, devstack enables it 14:34:49 and it should still work 14:35:05 so I propose to readd hdp 2.0.6 tests to libery and mitaka branches 14:35:21 I would say that we should keep hdp206 for liberty, if it possible 14:35:22 I agree with tosky. For everything up through mitaka, it should remain being tested 14:36:22 we don't want to break it with a backport to stable/mitaka accidentally ... 14:37:04 tmckay, it's deprecated for mitaka, so cluster can't be created 14:37:12 ah, okay 14:37:22 vgridnev, liberty then 14:37:29 I thought we just put out a warning ... 14:37:46 but if we refuse to launch, okay 14:38:28 when you discuss the deprecation policy again in Austin, please talk about the meaning of "deprected" and the cycle again 14:38:34 deprecated* even 14:38:58 I still feel that deprecating should be like what happens with library: a bit statement printed, but still working 14:39:07 agreed, there is a lot of mess with meaning of this word 14:39:25 we suddendly move from "working" to "not working" without notice (as deprecated is basically "not working") 14:39:41 yes. I think warning is good, exception is a bit of overkill. I go in and remove the exception ;-) 14:40:14 tosky: i thought that is how we interpret deprecated now? 14:40:26 we could even add a config "raise_on_deprecation", default true if we like 14:40:28 i thought we warn about deprecation for a cycle, then remove 14:40:44 I thought as well, but I would need to check the code for hdp in mitaka 14:40:49 if not, we need to seriously look at the deprecation tag policies and make sure we are inline with them 14:40:50 elmiko, some of the plugin deprecations are done with an exception 14:40:56 iirc it was just disabled 14:41:10 elmiko, soft stop for a developer (but for a non-tech-savvy customer, effectively immediate breakage) 14:41:29 maybe a little too harsh 14:41:44 then, i agree with tosky. we need to talk about this again and better establish our methodologies 14:41:54 maybe a config to toggle would be a good compromise. "We really don't want you to use this, but if you must .... set the config" 14:41:59 elmiko: may I point out that you +2-ed this? :) https://review.openstack.org/#/c/292364/ 14:42:07 tosky, lol 14:42:15 tmckay: and you too :P 14:42:16 me too 14:42:24 I am ashamed 14:42:43 lol 14:42:48 well, we need to revisit it then 14:42:57 ok, so just please try to define the policy properly for future deprecations 14:43:01 i'm ok being wrong.... occasionally ;) 14:43:02 (I won't be in Austin) 14:43:03 elmiko, let's make it part of deprecation rodea 14:43:14 we need to better follow this: http://governance.openstack.org/reference/tags/assert_follows-standard-deprecation.html 14:43:45 and probably state that we will deprecate (meaning give warnings), then remove 14:44:14 i realize that tag is mainly about api stuff, but still 14:45:23 added a note to the etherpad ... 14:45:32 uh, I forgot something to mention. it looks like new third-party CI was added to sahara, Mirantis Rally CI, but it looks like it's still failed 14:46:41 rally-team working on it 14:47:45 esikachev, do you know something about estimates when it will working? 14:48:36 they said 'asap' 14:49:09 ok, thanks esikachev 14:50:37 also, I will try to be prepared to the future revisiting blueprints, I will add new topic to meeting with the ether pad to track that 14:51:09 I hope that we can find something to discuss at the summit 14:51:15 Maybe 14:51:27 sounds good 14:54:17 5 mins left, anything else to discuss? 14:55:58 * tmckay cannot think of anything 14:56:14 5 14:56:19 4 14:56:24 3 14:56:28 2 14:56:48 1, thanks for attending to the meeting 14:56:57 #endmeeting