17:00:25 #startmeeting murano 17:00:26 Meeting started Tue Apr 28 17:00:25 2015 UTC and is due to finish in 60 minutes. The chair is sergmelikyan. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:27 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:30 The meeting name has been set to 'murano' 17:00:42 Hi Team o/ 17:00:46 hi! 17:00:55 Hi 17:01:26 o/ 17:01:36 I suggest we deside on a secret murano greeting 17:01:47 kzaitsev, like? :) 17:01:52 yo! 17:01:56 dunno 17:01:58 ativelkov, was proposing usage of muranoers 17:02:02 0/ 17:02:20 o%// 17:02:32 hi 17:03:28 do OS-users have a secret greeting? 17:03:36 (⌐■_■) 17:03:46 kzaitsev, dunno :) 17:03:54 #topic Agenda 17:04:16 As usual today's agenda is available here: https://wiki.openstack.org/wiki/Meetings/MuranoAgenda 17:04:42 I would like to go through few bugs today that are considered critical, but we are so close to release... 17:04:57 #topic Action Items 17:05:38 #1 meeting about blueprint/abstract-env-template is scheduled for ... 17:06:07 henar was not able to attend that meeting and we didn't agreed on a new time :( 17:06:48 In scope of this topic, let's discuss for a minute what is better way to have discussions around new features given that we are not able to dicuss them on summit? 17:07:09 I'd prefer hangout 17:07:41 I mean, if we need some way other then spec review 17:08:22 ativelkov, yeah I am talking about pre-spec discussions that usually held during summit 17:09:09 yup, the hangout should work then 17:09:21 ativelkov, I also like hangout 17:09:34 slagun, katyafervent any thoughts? 17:11:29 +1 17:11:42 * sergmelikyan has issues with eyesight not ears, but he is wandering if he missed response from katyafervent 17:11:43 I agreed with hangout 17:12:26 Then our workflow is: announcement -> discussion (hangout) -> meeting minutes 17:12:31 but even better to have rich chat + hangout 17:12:38 rich chat? 17:12:50 I propose to have some etherpad to gather discussion proposals 17:13:29 something like slack. Something better than IRC 17:13:40 ativelkov, agree 17:13:42 slack is internal only 17:13:55 there are a lot of similar projects 17:13:56 you cannot use for intercompany communications 17:14:02 ah, got it 17:14:06 slagun, can't understand why you consider slack is better option than IRC, but let's skip this 17:14:16 well, we used to have "virtual summits" in Glance 17:14:28 ativelkov, using? 17:14:33 IRC should be the default choice. it's the only open channel of real-time communication 17:14:48 ruhe, for chat - sure, +1 17:14:55 yeah, for chatting 17:15:16 it was called vidyo 17:15:30 I also prefer IRC 17:15:45 www.vidyo.com 17:15:48 the problem with IRC is its single-line message. You cannot paste code samples etc 17:15:59 Last question before moving on... 17:16:05 ativelkov, how did you like it? 17:16:24 well, nothing special. ± the same as webex 17:16:40 huh... 17:16:46 #topic Release Status 17:16:59 https://launchpad.net/murano/+milestone/kilo-rc2 17:17:25 We still have several open issues, three of them I would like to discuss specifically 17:17:43 I think aderyugin fixed issue with our devstack, and that is a relief 17:18:03 kzaitsev is working on his issue, but I would consider it also fixed 17:18:11 link to patch to devstack: https://review.openstack.org/#/c/178290/ 17:18:23 Finally! 17:18:36 Yeah... it's so sad that issue was so simple :( 17:19:37 https://review.openstack.org/177796 & https://review.openstack.org/177869 for catalog issues fixed by kzaitsev 17:19:57 freerunner, can you help us verify this issues? 17:20:37 sergmelikyan: sure 17:20:43 We are really close to release, like 2 days left, and it's big risk to include fixes right now, but from other point of view this bug is a major flaw in core Murano functionality 17:20:45 We're currently doing that =) 17:20:53 There are no more new bugs in rc2, right? 17:20:56 kzaitsev, awesome! 17:21:24 katyafervent, not exactly :( Today we found three more critical bugs that I would like to discuss today. 17:21:34 Main question - should we have them fixed in kilo or not 17:21:58 #1 https://bugs.launchpad.net/murano/+bug/1449500 17:21:58 Launchpad bug 1449500 in murano "AgentListener doesn't stop after GC cleanup" [Critical,Confirmed] - Assigned to Stan Lagun (slagun) 17:22:45 sergmelikyan: I already know why it happens. Will fix today 17:23:04 This bug essentially breaks application deployment and it is really critical... I personally think that we need to focus and fix this in Kilo 17:23:06 we can always back port them later, as we did for some bugs found in stable/juno 17:23:06 what are the chances to introduce a regression with your fix? 17:23:31 katyafervent, not exactly - many people use released version not a stable branches 17:23:45 ruhe: very unlikely 17:24:04 so you will be use latest version of juno not 2014.2 17:24:57 now it's 2014.2.2 and it has fixes for some critical bugs 17:25:05 opposite - people usually use and include in the distribution 2014.2, 2014.2.1 and so on and rarely pull code from stable/juno 17:25:34 katyafervent: we don't want to wait for 2015.1.1 for such a critical fix 17:25:41 if we would not fix this bug in kilo right now it means that fix would be available at least 1-2 month later 17:25:46 ativelkov, +1 17:26:10 so, I'd prefer to risk and include it in RC3 17:26:24 ativelkov, we still may release in RC2 :) 17:26:37 s/RC3/RC2 :) 17:26:45 slagun, so any hints how you plan to fix this issue? 17:27:31 But that's a lesson lerned: we definetly have to many critical issues at RC stage. Probably because we had too many freeze exceptions 17:28:01 or very late QA cycle & not so many tests :( 17:28:11 sergmelikyan: we are using environment.start()/stop() to execute pre/post deploy hooks. The fix is to have GC as an explicit step with pre/post-GC hooks similar to deployment 17:28:25 sergmelikyan: that's the reason of that: we were still doing feature development when the schedule assumes QA 17:28:34 currently GC happens implicitly upon model load 17:28:45 let's remember this and try to avoid such a pattern at L 17:29:16 ruhe, is it a common practice to fix such kind of bugs in rc releases? do all the projects have a lot of bugs in rc releases? 17:29:18 ativelkov, yeah... 17:29:37 katyafervent, it depends... horizon goes for RC3 AFAIK 17:31:08 in what scenarios will this issue be reproduced? would it affect *all* users? 17:31:59 it affects any application that have destruction logic 17:32:15 slagun, ruhe Kubernetes? 17:32:53 I've encountered this issue when I started to implement Docker applications removal 17:33:25 we don't have distruction in any other applications in murano-apps but users may have it in their own apps 17:33:50 then, imho, fix should be merged now in kilo 17:34:24 #startvote Fix #1449500 in RC2? Yes, No 17:34:24 Begin voting on: Fix #1449500 in RC2? Valid vote options are Yes, No. 17:34:25 Vote using '#vote OPTION'. Only your last vote counts. 17:34:32 #vote Yes 17:34:40 #vote yes 17:34:42 #vote Yes 17:34:49 #vote Yes 17:35:16 #vote yes 17:35:19 #endvote 17:35:19 Voted on "Fix #1449500 in RC2?" Results are 17:35:21 Yes (5): katyafervent, ruhe, ativelkov, slagun, sergmelikyan 17:35:21 #vote yes 17:35:36 I think we had majority here :) 17:35:40 freerunner, sorry :) 17:35:51 sergmelikyan: Oops =) 17:36:08 #2 https://bugs.launchpad.net/murano/+bug/1449497 17:36:08 Launchpad bug 1449497 in murano "[MuranoPL] pselect exits on first exception" [Critical,Confirmed] - Assigned to Stan Lagun (slagun) 17:36:37 this is even more important than the previous one 17:36:56 This bug is nasty, we should definitely fix it 17:37:18 We can't get real exception because of this bug 17:37:29 Yeah, this affects almost every deployment 17:37:35 Well, there is another option: replace pselect with select in core library. This will not introduce any regressions, we'll just decrease performance 17:37:47 we had this bug since day 1 of MuranoPL :) 17:37:53 I mean, in the Environment.deploy() 17:37:55 Moreover, this bug shows misleading exception 17:38:05 slagun, you are proud of that? :) 17:38:30 sergmelikyan: :-)) 17:38:37 slagun, if you tell me that you know about this issue long time ago I will not buy a beer for you on Summit 17:38:42 slagun: may i ask you to update bug description and include steps to repro, symptoms, etc. so that users of murano might understand it and figure out they're hitting exactly this issue? 17:39:33 :) 17:39:36 ruhe: one second, please ;) 17:39:49 I told about it several times but until know it didn't caused any real harm 17:40:27 ruhe: there is a slightly better description here: https://bugs.launchpad.net/mos/+bug/1449635 17:40:27 Launchpad bug 1449635 in Mirantis OpenStack "[MuranoPL] pselect exits on first exception" [Critical,Confirmed] - Assigned to Stan Lagun (slagun) 17:41:09 symptom is when you [OS Errno 2]: No such file in dashboard. Sometimes this is the only error you see 17:41:22 I agree that we need to update description for #1449497... nasty :( 17:41:47 it seems that there's general agreement that this one should be fixed now 17:42:26 #agree #1449635 need to be fixed in RC2 17:42:50 #3 https://bugs.launchpad.net/murano/+bug/1449510 17:42:50 Launchpad bug 1449510 in murano "[Docker] Sometimes Docker doesn't start" [High,Confirmed] - Assigned to Stan Lagun (slagun) 17:43:16 I have doubts about that one... it's just one of the application that we developed, and does not affect core functionality 17:43:34 * sergmelikyan remembered that we need to move tracking apps bugs to separate launchpad 17:44:02 sergmelikyan: +1 and in long term we'll need to reconsider the approach to track applicaitons as part of release of murano. they're different entity 17:44:04 This is additional step in DockerStandaloneHost to start docker service. Just a workaround. I think this is some docker startup bug 17:44:29 But from the other point of view it affects K8s & Docker apps and they are most powerful apps that we have AND they enable whole different level of applications 17:44:30 sergmelikyan: ruhe: https://bugs.launchpad.net/murano/+bug/1449497 done. 17:44:30 Launchpad bug 1449497 in murano "[MuranoPL] pselect exits on first exception" [Critical,Confirmed] - Assigned to Stan Lagun (slagun) 17:44:48 slagun, so you have workaround for that bug? 17:45:16 * sergmelikyan changing his mind about buying a beer for slagun 17:46:01 freerunner: one minor suggestion - include logs into the bug description. it'll help users to google for this error 17:46:03 sergmelikyan: yes, I have. It is very simple. I will commit it with application removal 17:46:10 (but not in the same coommit) 17:46:53 slagun, ok, then beer is on me and maybe a wine :) kzaitsev says that Vancouver should have great wine 17:47:07 icewines 17:47:34 #startvote Fix #1449510 in RC2? Yes, No 17:47:34 Begin voting on: Fix #1449510 in RC2? Valid vote options are Yes, No. 17:47:35 Vote using '#vote OPTION'. Only your last vote counts. 17:47:43 #vote Yes 17:47:47 #vote yes 17:47:53 #vote Yes 17:48:00 #vote yes 17:48:02 #vote Yes 17:48:04 ruhe ? :) 17:48:14 #vote Yes 17:48:21 # No 17:48:22 sergmelikyan: don't we have any restrictions on our RCs from Openstack's release manager? 17:48:22 #endvote 17:48:22 Voted on "Fix #1449510 in RC2?" Results are 17:48:23 Yes (6): ruhe, ativelkov, slagun, freerunner, ddovbii_, aderyugin 17:48:49 katyafervent2: why no? 17:48:54 ativelkov, no, Murano is not managed by OpenStack Release Manager 17:49:24 as well as other projects joined to Big Tent recently 17:49:32 do we have time for so many bugs? 17:49:49 katyafervent2: This bug should be fixed. Sometimes it blocks deployment. 17:50:07 katyafervent: there is also https://bugs.launchpad.net/murano/+bug/1446582 17:50:07 Launchpad bug 1446582 in murano "Invalid workflow of application replacement" [Medium,Confirmed] - Assigned to Stan Lagun (slagun) 17:50:24 Looks like we have solutions for all of them, and I believe freerunner some tests that are not yet commited for last bug 17:50:26 if you are 100% sure we wikk fux it and TEST in 3 days - i dont minf 17:50:26 and I'm going to close close 4 bugs today 17:50:58 slagun: and not introduce any regressions as you promised before ;) 17:50:58 slagun, ok, Jack is also on me :) 17:51:29 ruhe: I do a lot of manual testing on k8s and Docker. So at lest those 2 wont break 17:51:44 sergmelikyan: :) 17:52:35 Let's make a deal then: if we will have fix this issues by tomorrow, and freerunner will give his +1 to reviews (obviously +1 from other reviewers as well), we agree to merge them to RC2 and release it 17:52:46 *for all of the mentioned issues 17:53:16 I'm going to commit them to master first and then we decide if we are going ti cherry-pick them to stable/kilo if you don't mind 17:53:50 deal 17:54:07 ok 17:54:32 slagun, better to propose them to stable/kilo first 17:54:50 will save time, testing phase will be same regardless of the branch 17:56:36 ok 17:57:30 #topic Summit Etherpads 17:57:52 #action sergmelikyan create etherpad for summit topics 17:58:03 We are little bit late on this task, mea culpa 17:58:07 I will do that ASAP 17:58:15 #topic Open Discussion 17:58:38 * sergmelikyan changed topic in rush to avoid throwing eggs phase :) 17:58:59 Few minutes to bring topics that we will discuss in #murano 18:00:19 time's up 18:00:50 #endmeeting