18:00:14 #startmeeting trove 18:00:15 Meeting started Wed Jan 27 18:00:14 2016 UTC and is due to finish in 60 minutes. The chair is cp16net. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:16 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:00:18 The meeting name has been set to 'trove' 18:00:25 ./ 18:00:34 hi 18:00:38 Hi all! 18:00:39 howdy everyone 18:01:08 o/ 18:01:10 egafford: welcome 18:01:30 o/ 18:01:45 cp16net: Thanks! (This is Ethan from the Sahara team; I'm hoping to serve as a Sahara liaison for when/if that might be useful.) 18:02:14 awesome 18:02:22 lets get this party started 18:02:26 And, you know, to Trove, because Trove. :) 18:02:26 #topic pulse update 18:02:41 #link http://bit.ly/1VQyg00 18:02:54 #link https://wiki.openstack.org/wiki/Meetings/TroveMeeting 18:03:27 o/ 18:04:10 we've got more reviews 18:04:37 one stat i added what on the others tab for reviews 18:05:01 tracking total vs core vs other reviews 18:05:11 o/ 18:06:12 that can be helpful to see if we have community support reviewing 18:07:24 it looks like we've done some work here but a bit more goto 18:07:48 thanks for everyone reviewing! 18:07:56 its very helpful 18:08:11 not much else from me on this 18:08:29 moving along 18:08:37 #topic announcements 18:08:53 we got the gate issue resolved last week 18:09:00 after much fun with pip issues 18:09:16 and there was an issue with the elements 18:10:30 thanks for the help from abramley we got it all resolved 18:10:46 #info fixed gate issues 18:11:14 #info trove midcycle meetup is 2 weeks away (feb 9-11) 18:11:30 #link https://etherpad.openstack.org/p/mitaka-trove-midcycle-rsvp 18:11:36 WHAAAAT??? 18:11:40 #link https://wiki.openstack.org/wiki/Sprints/TroveMitakaSprint 18:11:59 pmackinn: i hope thats a joke :-P 18:12:16 * pmackinn counts 6 participants officially 18:12:35 * amrith bites his tongue 18:12:36 i aw awaiting news of the marriott rates 18:12:45 i created the eventbright and i'll get the numbers on that later 18:12:57 i think its been duplicates of the members on the list 18:13:06 who has flights atm? 18:13:19 ./ 18:13:24 sorry, you said flights 18:13:28 I thought you said fights 18:13:36 actually 4 rsvp 18:13:44 for the live fights 18:14:47 goody 18:14:58 please signup if you plan on being there 18:15:18 it will help with food situation and planning 18:15:24 cp16net, +1 18:15:55 a few other announcements i have 18:16:15 #info [release news] mitaka-2 was released 18:16:23 #link http://docs.openstack.org/releases/releases/mitaka.html#mitaka-trove 18:17:03 #info [release news] mitaka-3 is planned for March 1-3 18:17:14 #link http://docs.openstack.org/releases/schedules/mitaka.html 18:17:36 this means we need the code for the features in asap 18:17:57 to get reviewed and updated before the rush 18:19:07 #info [release news] trove-dashboard was tagged with 6.0.0.0b2 18:19:36 it had no tags before and zigo pointed this out last night 18:20:04 this version was chosen because trove has been in horizon since havana 18:20:34 i tagged the head of master with this version last night 18:20:41 fyi 18:21:05 anything i missed? 18:21:57 alright well i guess we'll move to open discussion 18:21:59 only other thing I can think of is that zigo has other issues with trove (not dashboard). 18:22:07 and he's reported some bugs on them. 18:22:46 amrith: yeah i bet its because he is using the same vm to build all the packages across OS 18:23:10 #topic open discussion 18:23:36 anything anyone wants to bring up? 18:23:48 redstack/devstack clean this week? noticed some floating ip weirdness yesterday from a fresh install 18:24:03 * pmackinn phishs 18:24:22 i havnt messed with floating ips on devstack so i'm not sure 18:24:29 just a request to push some of the test-related backports on stable/liberty - I see that there are only two core reviewers for the stable branches, so it could be easy to miss 18:24:39 haven't noticed any recently pmackinn 18:24:55 tosky, got some #'s 18:25:17 you can send them later in #openstack-trove 18:25:22 no need to hold everyone up now 18:25:25 am looking for them 18:25:32 tosky: yeah there are a few and i think slicknik -1 asking why we were pushing test changes in stable 18:26:37 something i noticed another team doing in their meeting was a weekly update topic from everyone 18:27:07 amrith: thanks; the two you pushed and also https://review.openstack.org/#/c/262289/ 18:27:09 I could do one right now if you want ;) 18:27:15 everyone just gives a 1 sentence of what or where they are with their tasks 18:27:23 cp16net, we need a good bug scrub 18:27:29 I've been doing it adhoc for some days 18:27:36 and there are 261 oldy moldy bugs 18:27:43 and some are no longer relevant 18:27:55 i was thinking this would be really helpful with the release coming up 18:27:55 and some important ones are buried in the weeds 18:28:04 so fair warning to all 18:28:09 I'm going to be running through lp 18:28:23 and posting comments around bugs that have been growing old. 18:28:36 and fixing up status; such as change in-progress if the chagne is abandoned 18:28:52 But, I'd like people's view on unassigning bugs from people who aren't active in the project any longer. 18:28:56 that's a touchy subject 18:28:57 amrith: yeah that should be done a little before m-3 release so that we know whats left we can work on in the m-4 time frame 18:29:23 no bug yet, but am dusting off fedora in redstack (f23) in advance of pending mariadb feature set 18:29:46 amrith: i think if there is no review for the bug up then we dont know what the progress is 18:30:11 pmackinn, there's a bug to get redstack/devstack for trove completely working for fedora 18:30:14 I entered it a while ago 18:30:18 currently assigned to vkmc. 18:30:40 1419089 18:30:40 "completely" 18:30:50 https://bugs.launchpad.net/trove/+bug/1419089 18:30:52 Launchpad bug 1419089 in Trove "Get redstack working completely with fedora" [High,Triaged] - Assigned to Victoria Martinez de la Cruz (vkmc) 18:31:23 you pushed some of the trove-integration stuff 18:31:27 a while ago 18:31:39 working on that certainly, still hitting a lot of bugs related to hard wired Ubuntu paths 18:31:44 we are almost there I'm dare to say though 18:32:02 is there a change or set of changes to converge to a complete solution? 18:32:11 i think the trove project should be removed from that bug imho 18:32:16 long march 18:32:19 so my update is the pxc review of grow/shrink(working on getting current networks on instances in the cluster) and root-enable(ready for review). 18:32:50 probably a set of changes is going to be needed, I'll submit the ones I have in my queue this week 18:33:20 that would be great vkmc 18:33:26 :) 18:33:59 if you can squash the changes together for the bug that would be good 18:34:21 k k, will do 18:34:30 unless its across projects then create bugs for them across the projects and link them in lp 18:34:51 #action amrith to chat with pmackinn about removing trove as an affected project from 1419089 18:35:12 +cp16net to that convo 18:36:09 ok, right after this mtg? #openstack-trove? 18:36:20 oh one thing i forgot to mention there was an ask for a liason for cross-project issues by default its me but curious if anyone else was interested in helping out? 18:36:54 if you are interested let me know. 18:37:06 my status: I'm playing with the integration tests, even if more on the liberty branch (but with an eye on mitaka) on RHEL platforms 18:37:18 https://wiki.openstack.org/wiki/CrossProjectLiaisons describes the full set of liaisons it's nice to have. 18:37:22 cp16net, was that a per-project issue? 18:37:30 I'm supposed to be one oslo 18:37:47 there is the cross project effort as well 18:37:50 * egafford politely suggests that one person would be hard-pressed to fill all these roles for any one project. 18:38:01 #agrees 18:38:18 cp16net, I thought Peter was the cross-project contact. 18:38:36 actually, I've been the contact for a while now ;) 18:38:42 vgnbkr: no this is relatively new 18:38:53 for oslo that is. 18:39:16 cp16net, maybe that was just for the incubator stuff or something 18:39:19 Ok, timeout! 18:39:30 if you are interested in helping out this would be helpful! :) 18:39:31 What is this effort you're talking about cp16net ? 18:39:51 egafford sent a link to a page with many liaisons already identified. 18:39:52 The OpenStack project relies on the cross-project spec liaisons from each participating project to help with coordination and cross-project spec related tasks. See full set of responsibilities The liaison defaults to the PTL, but the PTL can also delegate the responsibilities to someone else on the team by updating this tableː 18:39:56 are you talking about the same thing? 18:40:05 look at the bottom on the list of liasons 18:40:25 if interested let me know 18:40:39 thats all i have 18:40:44 I just wanted to say hi to all. I'm Dale Smith from Catalyst IT and I'm starting to look at existing trove single tenant features and how I can help improve these. (with a view to address https://bugs.launchpad.net/trove/+bug/1445295). 18:40:45 Launchpad bug 1445295 in Trove "Guestagent config leaks rabbit password" [Undecided,New] - Assigned to Amrith (amrith) 18:40:45 Ok cp16net 18:40:51 that's the SPEC liaison 18:40:57 liaison 18:41:05 specifically cross project specs. 18:41:21 Happy to help with that. 18:41:27 amrith: sorry yeah thats what i meant 18:41:29 :-P 18:41:43 vgnbkr, I'm the API liason (FWIW) 18:41:52 dalees: welcome and we appreciate your help 18:42:01 Looks like Trove is missing Product WG and Logging WG liaisons as well at the moment. 18:42:04 peterstac, Ah, I knew there was something :-) 18:42:06 And since I'm the guy who started the mail thread that prompted a Sahara/Trove liaison, I'll volunteer for the Trove sie. 18:42:08 i see amrith was assigned that bug but not sure where he is on that 18:42:17 so it might be good to chat with him about it 18:42:29 amrith: o/ 18:42:49 egafford: it all defaults to me if no one picks it up but yea there are a few openings 18:42:53 cp16net, I had a couple of conversations with Bruno on that bug 18:43:12 cp16net: Indeed; just noting. :) 18:43:15 and it seemed like there was (at the time) a disconnect between the capabilities of OSLO and what Catalyst wanted to propose. 18:43:17 true! 18:43:39 So, one thing at a time. 18:43:49 with your permission cp16net I will add myself as Trove/Sahara liaison 18:43:50 ok? 18:44:01 sounds good to me 18:44:06 Thx 18:44:12 amrith, so many hats? 18:44:26 but none of them are red... 18:44:27 :-P 18:44:32 enough to talk through 18:44:54 anyone have anything else? 18:44:59 o/ 18:45:04 just a quick question for tosky 18:45:13 * amrith waits for vgnbkr to finish 18:45:18 you go first vgnbkr 18:45:24 OK. 18:45:38 * amrith graciously allows peterstac to go next 18:45:49 We've had ongoing issues and misunderstanding with API versioning between the Trove components. 18:46:23 I would like to propose a spec outlining a method for dealing with these issue, but would want some level of buyin before making the effort. 18:46:45 Do people think it worthwhile formalizing the method for changing APIs? 18:47:12 anyone? ... 18:47:32 vgnbkr: there is a method to it but its not documented really so it might be helpful to layout the path 18:47:43 for trove that is 18:48:02 vgnbkr, I think it would be a worthwhile spec and think it makes sense to get some agreement on the approach 18:48:05 before writing the spec. 18:48:16 vgnbkr, sounds like you're proposing a document not a 'spec' - rihgt? 18:48:24 I think it would be very good to formalize the method for changing the API. 18:48:44 My proposal would basically be that every API destination have a preceeding block comment that it never be changed (the destination). 18:48:50 peterstac, vgnbkr, cp16net ... would a short email to the ML make sense? 18:49:26 getting an abstract of the details before diving to far in would be good 18:49:33 I thought the ML was where ideas went to get shot down ... ;) 18:50:41 vgnbkr, that proposal seems to make a lot of sense 18:50:44 the block comment 18:50:44 vgnbkr: write up an abstract on the process you envision and we can discuss 18:50:54 Then a method for how a new API destination would be named to show its relationship to the previous version (say reset(foo) => resetMOD2(foo, bar)). 18:51:03 we could bring it up next meeting 18:51:28 I'm not doing to spend the time writing a document if the basic idea isn't what people are looking to do. 18:51:48 i think there is some value to it 18:51:49 if we go that route vgnbkr we could also add a test that would check the method signature and barf it was changed ;) 18:52:08 amrith, yes, I was going to propose that as well. 18:52:24 I think it is a good thing to do vgnbkr 18:52:31 i agree 18:52:37 and if I can help write it or the spec, you know where to find me. 18:52:57 alright peterstac had something/ 18:53:08 And you agree with the basic idea of naming methods (reset(foo) => resetMOD2(foo, bar))? 18:53:22 devil meet details 18:53:22 (of course, the MOD2 part is open to discussion) 18:53:56 makes sense 18:54:35 OK, I'll write something up. A Wiki doc rather than a spec, then? 18:54:37 I just had a question for tosky (if vgnbkr is done ;) ) 18:54:42 peterstac: sure :) 18:54:50 go ahead 18:55:24 tosky, you're trying to get some tests backported to liberty, but from what I can see, these are all 'scenario' tests 18:55:44 peterstac: yes, they are 18:56:11 These tests don't run in the gate yet (never have) and definitely weren't stable in liberty 18:56:21 (some aren't stable yet in mitaka) 18:56:34 peterstac: tru! 18:56:35 peterstac: I understand that, but they were partially backported already; with the few missing bits, they should be able to work properly in a still-supported branch 18:56:39 just curious what the motivation is 18:57:30 but that's my point - they don't run in the stable branches - never have 18:57:48 (still don't, in fact) 18:58:06 yeah they were a work in progress in the last stable branches 18:58:25 yes, but from the discussion on the channel it seemed that at least with two backports (the one going in now) at least they were not totally broken 18:58:36 and from my testing I can say that they are mostly working 18:58:39 the int-tests that we have that tested mysql thoroughly were stable 18:59:15 right, but between spending time with the old ones and adapt them to mariadb vs using the new ones, I invested my time in the new one 18:59:25 hopefully it's something that will be visible on RDO soon(ish) 19:00:45 tosky, ok, just curious (since we have limited bandwidth and lots of stuff to do for mitaka) 19:00:52 I was going to ask for few more small backports (of patches I did) but it the direction is to reduce their usage, I won't ask for that 19:01:41 are we there yet? 19:01:49 tosky, you've submitted changesets for the scenario tests? (must have missed them - will go look) 19:01:59 oops - out of time ... 19:02:12 alright times up 19:02:26 cp16net, pmackinn regroup in #openstack-trove? 19:02:27 #endmeeting