14:00:07 #startmeeting sahara 14:00:08 Meeting started Thu Dec 17 14:00:07 2015 UTC and is due to finish in 60 minutes. The chair is elmiko. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:10 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:12 The meeting name has been set to 'sahara' 14:00:16 hey folks 14:00:18 hi folks 14:00:21 hey! 14:00:21 hi 14:00:26 hi 14:00:31 hi 14:00:35 o/ 14:00:42 #chair vgridnev 14:00:43 Current chairs: elmiko vgridnev 14:01:04 hi 14:01:09 i'll give folks a minute or two to roll in before we get started 14:02:41 ok then 14:02:55 #topic sahara@sahara-dashboard 14:03:00 ;) 14:03:23 i think croberts might be out 14:03:30 anything to report here vgridnev ? 14:03:45 actually it's going, there are several changes was proposed, also integration tests job was merged to project config 14:03:55 cool 14:04:11 so we need to wait for some changes on horizon side 14:04:41 what changes are we waiting for? 14:04:44 vgridnev: so horizon will be running our intergation tests in their gate? 14:05:16 to make pluggable integration tests, let me search for the link 14:05:24 vgridnev: thanks 14:05:41 NikitaKonovalov, no, will run our own tests 14:06:16 #link https://review.openstack.org/#/c/253608/ 14:06:28 ok, at least we will no when our dashboard breaks after some horizon changes 14:07:17 or that the tests are broken by some horizon core change 14:07:22 but yeah, better than now 14:07:30 should be nice 14:07:31 yep, it's probably bad, but we need to make our integrations tests working first 14:07:38 anything else for this topic? 14:07:48 I think no, let's move on 14:07:56 #topic News/Updates 14:08:02 so, what's everyone been up to? 14:08:50 I'm working with improving ambari plugin right now 14:09:09 I'm still in internal research and testing, so not much to share 14:09:28 Image validation SPI spec is up; I have a working and tested sahara.plugins.images module. Doing cleanup before review @ this point. 14:09:31 a bit of reviews and small enhancement, nothing more for now 14:09:31 I hope I'll get some of the docs finished and finally published to upstream 14:09:39 i'm working on migrating the cdh passwords into the external key manager, but i've hit a snag 14:09:59 also, going to start setting up a prototype for the apiv2 work 14:10:40 working on spec, fixing bugs in scenario framework 14:11:42 ok, moving along 14:11:55 #topic Discuss sahara scenario tests extraction 14:11:57 I'm going to add a spec and patch for a --ironic or --baremetal flag on sie 14:12:05 doh, missed the updates :) 14:12:10 oops, sorry tmckay 14:12:11 I was typing 14:12:14 heh 14:12:15 np 14:12:40 #link https://review.openstack.org/#/c/250419/ 14:12:58 this spec is looking good, is there anything more to discuss on the topic? 14:13:16 spec is almost approved, but I think that we should wait for SergeyLukjanov to approve it 14:13:23 +1 14:13:29 i'm ok with that 14:14:05 I have an implementation question, related to repository extraction, but I guess it can wait (for the record: if in such cases a mechanism like git grafting is used) 14:14:32 you can ask now if you'd like tosky 14:14:44 (or it is just "initial import, dump of the revision xyz from this repository") 14:15:39 I think right now it's mostly done with a dump of the last version from the previous repository (I could check with sahara dashboard for example), but still it would be nice to easily track the old history 14:16:03 tosky, you can create local repo, but this repo will be imported during moving to openstack namespace 14:16:13 so, it sounds like the question is are we going to keep the git history when migrating to the new repository? (is that accurate tosky ?) 14:16:46 elmiko: keep the history is one way (apply git filtering to the old repository and extract the history) 14:17:07 I'm still concerned that this will make it harder to make plugins truly plugins (unless plugins require multiple repos to implement, and plug into a server framework and a test framework,) but I think we're deciding that that's okay for now (and we have a lot of harder work in the server to allow plugins to detach anyway, so it's probably fine.) 14:17:10 the other way is allow to access the history from the new repository, but without importing it; this is git grafting 14:17:15 yea, i guess we can't keep the whole history, but only the scenario parts 14:17:37 for reference to git grafting: https://git.wiki.kernel.org/index.php/GraftPoint 14:17:43 tosky: thanks! 14:17:53 egafford: it's a good point 14:18:32 egafford: i think we would need much more work on the plugin separation before we could contemplate scenario tests bundled with plugins 14:18:38 It is, but I can see the argument that we should do this now, and maybe we can wire the test repo for plugability later when we get closer to that long-term goal. 14:18:39 basically you don't have the old history, but the repository allow "joining" the old repository (if the user does some steps) and do a unified search 14:19:07 tosky, +1 14:19:24 egafford: imo, when we get closer to having true plugin separation, then we should revisit scenario test bundling as well 14:19:37 as time goes on, the need to search the old repo lessens 14:19:45 tosky: might be worth it, if we can setup the infra to do it 14:19:50 true 14:20:09 tosky: do you know if any other openstack projects have done this? 14:20:17 elmiko: I have to say I don't know the details, but you don't need special infrastructure, it's the way you create the new repository and you organize the old one before the removal/split 14:20:22 I found myself wanting to search before the "savanna" rename a few times, but not many ... 14:20:36 and no, I don't think it was done before, but I may be wrong 14:20:37 tosky: ah, interesting 14:20:52 that's why SergeyLukjanov could have helped with his infra knowledge :) 14:20:54 tosky: would you mind making a comment about that on the review? 14:21:01 elmiko: sure 14:21:09 I will keep the +1 14:21:33 #action tosky to comment on the scenario separation spec about git grafting the history 14:21:42 cool, thanks 14:22:00 any other questions/comments/concerns about the scenario test separation? 14:22:24 it would have been handy in the rename move, I think :) 14:22:32 the grafting, I mean 14:22:39 agreed 14:23:15 ok, rolling along... 14:23:22 #topic API v2 progress 14:23:30 not a ton to report here 14:23:35 the spec could use more reviews 14:23:42 #link https://review.openstack.org/#/c/212172/ 14:24:02 other than that, i am still working on setting up the wiki page for the individual work items to be generated 14:24:22 and i am starting to create a local prototype for the /v2 endpoints 14:24:48 any questions or comments about v2? 14:25:14 v2 is awesome 14:25:19 lol 14:25:30 the best api ever 14:25:43 best until v3! 14:25:44 ;) 14:25:59 ok then... 14:26:01 whoa, v3 ... that's never been done before! 14:26:06 haha 14:26:12 #topic New CLI 14:26:51 was there anything else to discuss on this topic, or should we remove it from the agenda? 14:27:09 let's remove that 14:27:12 It's a pretty great new CLI! 14:27:13 hmm I guess we have already discussed everything few weeks ago :) 14:27:18 yea, i think so 14:27:27 i'll remove it after the meeting 14:27:38 (elmiko: added a comment to the spec related to tests, which I missed before, my fault) 14:27:44 #topic Open discussion 14:27:48 tosky: ack, thank you! 14:27:59 any other business folks? 14:28:03 yes: 14:28:19 * elmiko hands the mic to tosky 14:28:24 please vote on this review on tempest https://review.openstack.org/#/c/255968/ (thanks vgridnev) 14:28:45 general comment, I should be more active upstream starting in January. I've been buried in a hole for a while ;-) 14:28:55 so, you can count on me for patches, reviews, etc 14:28:55 the tests there could move to our repository at some point, but pending some dependencies; in the meantime we need that change to support the testing on different clusters 14:29:03 tmckay: \o/ 14:29:17 tosky: ack, added to the queue 14:29:24 vgridnev already put you all in the review, but please take a look :) 14:30:00 for those of us who are unfamiliar with tempest, is there anything we should know about that review? 14:30:22 nothing. just vote with +1 14:30:31 :d 14:30:33 :D 14:30:38 lol, easy enough... 14:31:00 #link https://review.openstack.org/#/c/255968/ 14:31:05 it simply add support for default templates for the various supported plugins, so that the API tests works in any case, regardless of the set of enabled plugins 14:31:24 no fancy code change, if you check the code, just some additional dictionaries 14:31:24 ah, cool 14:31:36 thanks tosky, and vgridnev 14:32:14 any other topics to discuss? 14:32:32 will we have next 2 meetings? 14:32:48 i imagine not for next week 14:33:03 even the week after ... 14:33:04 and it seems like many openstack projects are taking the following week off as well 14:33:08 many of us will be PTO 14:33:11 yea 14:33:22 first week of January? 14:33:27 i think so 14:33:35 that would be the 7th 14:33:44 yea 14:33:51 but then, for people in Russia, don't you have your holidays that time? 14:33:55 so, next meeting will be January 7th 1800UTC 14:34:10 we have hollidays 1 - 10th of January 14:34:13 good question tosky 14:34:16 heh 14:34:23 so adjourn til the 14th? 14:34:27 i think so 14:34:35 there is always IRC, and openstack-dev, after all 14:34:41 ok, scratch that earlier memo 14:34:55 next meeting will be January 14th, 1400UTC 14:35:03 in openstack-meeting-3 14:35:11 vgridnev schedule edp patch is too old https://review.openstack.org/#/c/182310/ My intergation test patch https://review.openstack.org/#/c/255173/ has rebase issue 14:35:32 what kind of rebase issues? 14:35:40 So can Intergation test patch do not rebase on that? 14:36:07 Rebase error I am working on that issue 14:36:07 * tmckay brb 14:39:05 any other topics? 14:40:15 5 14:40:18 4 14:40:21 3 14:40:23 2 14:40:24 NOOOO 14:40:31 1.5 14:40:40 1 14:40:42 #endmeeting