18:00:55 #startmeeting trove 18:00:55 Meeting started Wed Mar 16 18:00:55 2016 UTC and is due to finish in 60 minutes. The chair is cp16net. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:56 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:00:58 The meeting name has been set to 'trove' 18:01:06 o/ 18:01:06 howdy everyone 18:01:14 cp16net, welcome back 18:01:15 howdy cp16net — welcome back! 18:01:20 thanks :) 18:01:28 ./ 18:01:37 hi 18:01:38 o/ 18:01:45 o/ 18:02:19 #topic action items from last week 18:02:32 #info [amrith] figure out who has to send FFE acceptances 18:03:33 Yup, I did ... 18:03:39 #done 18:03:47 but it appears that FFE's don't extend to client. we'll talk abotu taht later. 18:03:58 yeah ok 18:04:08 #topic Trove pulse update 18:04:19 #link http://bit.ly/1VQyg00 18:05:00 queue growth is negative, not since 9/9/2015 ;) 18:05:14 thats amazing 18:05:19 :) 18:05:28 i'm sure it wont last... 18:05:46 o/ 18:06:02 but over all we've merged many of the patches needed for mitaka 18:06:16 which it looks like attributed to this 18:07:33 anyone have other comments? 18:08:07 ok moving on 18:08:12 #topic Announcements 18:08:29 #info mitaka release schedule 18:08:31 #link http://releases.openstack.org/mitaka/schedule.html 18:08:47 this week is RC-1 cut 18:10:03 that means that newton will be opening up soon 18:11:39 PTL self nominations are happening now and i sent an email out saying that i've decided to not run again for the newton cycle. 18:11:50 do newton specs have to wait or can they be pushed up in advance? 18:12:06 they can be pushed up with the trove-spec repo 18:12:15 just add the newton folder 18:12:24 we need to create a directory for newton specs 18:12:28 yup 18:13:29 cp16net / pmackinn: we should be able to allow folks to target specs to future releases even if they're not open for coding yet. 18:13:37 amrith: has announced his candiacy which i think will be great 18:13:47 SlickNik: i agree 18:13:49 thx cp16net 18:13:57 thanks alot cp16net for all the hard work and helping us with a smooth release. Quite alot of work was done in the release 18:14:05 cp16net / amrith: I saw the PTL candidacy emails! 18:14:19 since the trove-spec repo is not a release repo you can push specs to any of the future releases 18:14:26 cp16net: Thanks for all the awesome work in the Mitaka cycle. 18:14:28 this will not affect anything 18:14:35 +1 cp16net thx 18:15:02 amrith: Looking forward to the direction the Newton PTL takes us! 18:15:08 On that note i'd like to thank everyone for their contributions and hard work over the mitaka cycle 18:15:09 I thought there was already a newton spec, but I was mistaken. 18:15:17 cp16net, thanks for all that you've done for trove 18:15:30 thanks SlickNik 18:15:49 and also i'd like to thank everyone for helping me out while i was away taking care of my new born. 18:15:53 +1 to having Amrith as our next PTL. SlickNik said it well 18:16:24 thx johnma ... 18:16:30 speaking of amrith .... 18:16:37 * amrith runs for the hills 18:16:37 #info [amrith] Cross project Quotas library/project 18:16:48 oh, thx 18:16:54 I put these two as informational items 18:16:58 this one and the next one. 18:17:05 wanted to bring it to everyone's attention. 18:17:11 I posted a response on the quote thing on the ML 18:17:25 it appears like we may have a library to use for quotas in the future. 18:17:39 I was hoping that there would be a quotas capability taht would give us more in the short term. 18:17:46 but maybe we'll have to wait for that. 18:18:04 in particular, I was interested in this because quotas (of other projects) become interesting to us when we muck with remote.py. 18:18:11 it is informational. 18:18:22 same with the next one 18:18:27 #info "Take a look at https://review.openstack.org/#/c/292451/" 18:18:40 it is something that flaper87 proposed. 18:18:52 it took a while to figure out what the dashboard is all about. 18:19:17 once I got it to work (and I left some comments for others who want to tryi it), I felt that the dashboard didn't address the things I'd thought we talked about at previous meetings. 18:19:23 oh yeah we created that a long time ago 18:19:25 but wanted to bring the review to everyones attention. 18:19:35 and it generated the link i've used for a while 18:19:45 In particular, one of the shortcomings of gerrit is that there is no way to tag a review. 18:19:55 i.e. tag a bunch of reviews for Mitaka-RC1 or some such thing. 18:20:03 and so we resorted to the tagged by someone. 18:20:13 when the dashboard was made, it was 'tagged by SlickNik' 18:20:23 yeah 18:20:27 for Mitaka I starred the reviews and shared a link. 18:20:36 so I'm less than enamored of this commit. 18:20:50 it would be nice to be able to corrolate the release to reviews 18:20:52 but figured I'd share and let others see what it was. and maybe we'll get more ideas on what the dashboard should contain. 18:21:01 but i think it would arequire query to launchpad as well 18:21:07 yes, I've been meaning to ask sdague a bunch of questions about it. 18:21:27 all that dash does it creates a link 18:21:33 one was about tagging reviews. the other was about picking a sort order in the dashboard, the third was to make the dashboard we create a part of the project 18:21:52 its easier to update that file than mucking with a 1000+ characters on the url 18:21:53 if you go to https://review.openstack.org/#/admin/projects/openstack/trove 18:21:57 there is a dashboards list 18:22:05 and I want to add dashboards there. 18:22:08 I wonder how to do that 18:22:15 So ... 18:22:18 that was old 18:22:25 #action [amrith] do the things he says above ... 18:22:28 they decided not to keep up with those 18:22:40 oh well, that's sad. 18:22:44 amrith: If I recall correctly, to be able to do that you'd have to modify the list of gerrit default dashboards. 18:22:44 at least the last time i asked about it 18:23:20 SlickNik, that may be it. have to figure out what a good way would be to do something like that. 18:23:27 yeah it requires an update to gerrit backend and i think they frowned upon because it might require a bit of work each time 18:23:30 especially tracking reviews for a milestone. 18:23:30 amrith: Talking to the CI folks, there wasn't an easy way to do that per team (i.e. making a change would result in that dashboard being available to everyone). 18:23:37 I think that'd make sense for a number of projects. 18:23:39 it was on the infra team i believe 18:23:56 anyway, not to rathole on that; I'll go find otu and report back to the team. 18:23:59 that's all I had cp16net 18:24:06 But they have done a couple of gerrit upgrades since then, so I wonder if this has changed at all. 18:24:31 k side note on that you can now update your list of links in gerrit 18:24:41 https://review.openstack.org/#/settings/preferences 18:24:48 in the my menu 18:24:54 create a link and add it there 18:25:10 cp16net, yes. but is there a way to make it project wide? 18:25:22 no 18:25:32 other than adding it to a wiki or something for reference 18:25:39 ok moving on 18:25:44 #topic Health check on outstanding blueprints and features for Mitaka 18:25:51 #link https://review.openstack.org/#/q/starredby:amrith+status:open 18:26:19 are we all good with the two mongodb changes listed there? 18:26:30 I'd like those merged into RC1 if at all possible. 18:27:21 i think they are good 18:27:39 i didnt run through another test on then 18:27:42 and if we can get imandhan's couchdb change merged, I think we're in a position to finalize the release notes and get rc1 going. 18:27:43 them* 18:27:52 i've been testing the couchdb patches 18:28:04 thanks cp16net 18:28:07 I've tried a couple of times to connect with vkmc and sonali but have not been able to. 18:28:17 vkmc is pto this week 18:28:18 so I fear that the two couchdb configuration group things will slip 18:28:32 well, that'd explain it! 18:28:45 And I can't reach sonali. 18:29:01 so I think we are good to call RC1 as soon as we get imandhan's changes in. 18:29:05 yeah i think we can move them to the next cycle 18:29:08 is that correct? 18:29:28 johnma, peterstac ... did we get everything else? are we missing something that should be starred? 18:29:49 any final requirements 18:29:56 updates or translations 18:30:13 looks like the Redis one got dropped somehow 18:30:16 #link https://review.openstack.org/#/c/272844/ 18:30:44 For production systems, the backup shouldn't lock out the connections ... 18:30:56 starred now 18:31:08 we need another reviewer 18:31:23 I'll review the redis patch 18:31:50 i had a concern about it but i think peter won me over. 18:32:42 cp16net, yeah, it fails over to the 'blocked' save if the bg one fails for any reason 18:32:50 (for example, due to memory issues) 18:33:08 yea 18:33:38 it's not on the list of reviews, but is the plan of enabling the new scenario tests still on track for Mitaka? 18:34:06 well, we're moving ahead with it, but my guess is it won't be ready until Newton 18:34:18 but should be early in the cycle 18:34:29 peterstac: not even a gate for the non-experimental datastore? 18:34:52 #agree scenario tests should be turned on in newton. 18:35:14 we could probably set up the mysql one as non-voting soon (since I think they're fairly stable) 18:35:35 we might want to run the Redis ones a bit more in case they need tweaking 18:35:43 we setup the other experimental jobs to run the scenario tests 18:35:47 (no sense wasting gate resources if we don't need to) 18:35:58 but they dont run normally 18:36:15 right, but it's the chicken and egg problem: tests are not stable because are not always executed, so they are not used for gates, so not executed 18:36:22 yes, and with all the mitaka push I haven't be running them manually much lately 18:36:38 but I should have a little more time now :) 18:36:42 there hasnt been anyone working to make sure they are stable in the gate tho 18:36:54 there is a job you can run but last i checked it failed 18:36:57 I didn't know about the extra job 18:37:24 if you run 'check experimental' in trove, it'll kick off the scenario tests for mysql and redis 18:37:39 yeah ^^ thats what i was going to say but peterstac beat me to it 18:38:13 maybe we can talk about it post-meeting, if the failure was not complex (or it was just for redis), it could be fixed easily 18:38:33 I suggest we take this up after the meeting please. 18:38:34 so for the features in mitaka that amrith starred we have the ones we need 18:38:38 there are other things on the agenda. 18:38:43 cp16net, if you like I can try and get those ready for a 'non-voting' switch by next week (maybe ...) 18:39:20 if we dont hear from sonali or vkmc soon on status then we will have to push that feature to newron 18:39:56 cp16net, I suggest we decide that now. 18:39:59 so we can move on 18:40:17 and cut rc1 as soon as the two mongo things, the couchdb change (imandhan) and the redis change are merged. 18:40:36 i'm in favor of pushing it wihtout any work on them in over a week 18:41:26 does anyone feel differently? 18:41:28 amrith: i know you've been talking with the release team 18:41:37 cp16net, yes. 18:41:44 is there pressure for release be cut asap? 18:42:03 i've caught parts of the convs but not heard much on when yet 18:42:13 I would like to mark rc1 so we can focus on bug fixing. yes, RC1 deadline is this week. they'll do it anyway ... 18:42:29 I would rather do it when we're ready and then focus on fixing and testing. 18:42:41 right i agree 18:42:52 that'll also mean that we can tag stable/mitaka and other things (for newton) can start to land. 18:43:02 yup 18:43:16 which is the next thing I have on the agenda ... 18:43:32 lets push the couchdb config groups now. 18:43:40 I've let peterstac know already. the client side changes for module management won't be a mitaka thing. 18:43:41 and you can un star it 18:43:58 #agreed couchdb configuration groups moves to Newton 18:44:00 Also — no one likes to release on Fridays, so realistically we have until tomorrow to merge what we want for RC1 18:44:12 true 18:44:17 I have a personal reason for wanting to cut it today. 18:44:26 tomorrow and day after, cp16net and I are pub hopping. 18:44:38 Ah, you're going to be in Austin! 18:44:40 heh :-P 18:44:45 Yup 18:44:54 johnma, watch out ... 18:45:00 ok i think that finalizes the patches we need fixed up asap 18:45:11 yes, there are now 4 18:45:14 sorry had to step out for a bit, came back just on time :) 18:45:28 three are verified. and need a +1 18:45:31 Makes sense — at this point, I'd highly recommend pushing out stuff that won't be ready to Newton. 18:45:32 one is waiting for gate 18:45:44 ok so next thing on the agenda? 18:45:53 So, for newton, we should remember that FFE's don't extend to the client. 18:46:11 if we need changes on the client, the freeze date for the client is a HARD date. 18:46:22 #topic Summit planning 18:46:23 and even if you get an FFE for a feature, that's your server side 18:46:30 ok, on to summit planning ... 18:46:36 Is this new? I don't remember hard freeze dates for the clients before ... 18:46:55 (In fact I thought we got to release the client whenever we saw fit ...) 18:47:02 peterstac: i think its enforced more now than it was before 18:47:03 peterstac, I don't know. And I don't know that it really matters at this point. 18:47:12 and in the interest of time, I'd suggest we look ahead. 18:47:14 to summit planning. 18:47:23 sure, noted 18:47:27 #link https://etherpad.openstack.org/p/trove-newton-proposed-sessions 18:47:44 we have 3 fishbowls, 6 working sessions and a half day contributor meetup. 18:47:54 amrith: outlined the propsed sessions here that we had talked about at the meetup last time 18:47:57 I've requested that our contributor meetup be Friday morning. 18:48:08 because I know that several people have afternoon travel plans. 18:48:14 if there are other suggestions please add to this as you see fit 18:48:42 amrith: ++ 18:48:44 #action [amrith] contact Victor Stinner re: python3 18:49:04 #action [pmackinn] I've tagged you on some topics based on mid-cycle, please confirm 18:49:38 #action [amrith] get leaders for all sessions that we want to actually conduct ;) 18:50:00 amrith, noted 18:50:19 cp16net, ... I'd like a quick rollcall vote on who'll be at summit. 18:50:32 ./ 18:50:44 o/ 18:50:47 ./ 18:50:54 ./ 18:50:59 o/ 18:51:03 o/ 18:51:06 ./ 18:51:36 vmkc, ./ (pretty sure) 18:51:43 * peterstac raises hand for dougshelley66 18:51:59 ./ 18:52:35 looking forward to seeing everyone there 18:52:54 #topic open discussion 18:53:16 thx cp16net 18:53:46 anyone have anything else? 18:54:04 if not i'll end it 18:54:05 since we have the time, tosky had some questions earlier. 18:54:11 or we can go to #openstack-trove 18:54:13 and continue there? 18:54:45 either way we have about 5 min 18:54:51 I think peterstac and cp16net mentioned they will recheck the status of the current job 18:55:11 which would be enough for now :) 18:55:17 tosky, I'll work on that this week 18:55:22 peterstac: thanks 18:55:27 sounds good 18:55:28 np 18:55:50 one question for pmackinn 18:56:00 oh boy 18:56:01 pmackinn, any idea whether tellesnobrega will be at summit? 18:56:06 I didn't put the ceph thing on the agenda 18:56:15 but we should, if he's looking to do it for newton 18:56:41 amrith, will enquire and try to get you an answer tomorrow 18:56:51 there are other projects for which I've file BP's and I'll list some of these in the contributor meetup section or maybe have a session for "projects in Newton" 18:56:55 thx pmackinn 18:57:22 alright thanks everyone 18:57:26 #action [amrith] add projects in newton to agenda 18:57:45 #endmeeting