14:00:37 #startmeeting tripleo 14:00:38 Meeting started Tue May 31 14:00:37 2016 UTC and is due to finish in 60 minutes. The chair is EmilienM. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:39 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:42 The meeting name has been set to 'tripleo' 14:00:47 #topic rollcall 14:00:51 hey folks 14:00:51 Hello 14:00:52 \o 14:00:56 hi 14:00:58 o/ 14:01:02 hello 14:01:11 Hi all! 14:01:20 #topic agenda 14:01:21 o/ 14:01:27 #link https://wiki.openstack.org/wiki/Meetings/TripleO 14:01:42 * two off agenda items 14:01:46 * bugs 14:01:51 * Projects releases or stable backports 14:01:56 * CI 14:01:57 * specs 14:01:59 * open discussion 14:02:19 (I just realized I created an item that is part of a topic) 14:02:27 so we have one off agenda item 14:02:34 beagles: o/ 14:02:42 #link https://etherpad.openstack.org/p/tripleo-bug-cleanup-2016 14:02:46 you wanted to chat about it? 14:02:49 o/ 14:02:55 o/ 14:02:58 hi 14:03:16 EmilienM, unfortunately I can't edit wiki and add there issues, because of security policy of wiki, maybe it's better to use a etherpad for this? 14:03:22 more of a status really 14:03:40 beagles: sure, go ahead 14:04:01 sshnaidm: that's unfortunate :( 14:04:02 I've been digging through the bugs starting at the oldest first and working my way forward 14:04:37 some of them are probably having more learned eyes go over, but aren't a high priority afaict 14:04:39 o/ 14:05:40 I just wanted to bring it up in this meeting along with the link in the agenda that if anyone is curious, go have a look and offer thoughts, guidance, correction etc 14:05:42 that is all really 14:05:45 o/ 14:05:49 beagles: ok cool 14:06:02 #topic bugs 14:06:10 #link https://bugs.launchpad.net/tripleo/ 14:06:26 hi everyone 14:06:28 * dprince is late 14:06:51 do we have outstanding bugs this week? 14:07:18 I filed yesterday: https://bugs.launchpad.net/tripleo/+bug/1586906 14:07:18 Launchpad bug 1586906 in tripleo "CI: periodic jobs fail becasue "no stack found"" [Undecided,New] 14:07:32 just wanted to mention https://bugs.launchpad.net/tripleo/+bug/1585641, which is still causing many CI jobs to fail 14:07:32 Launchpad bug 1585641 in tripleo "CI: jobs failing with Message: No valid host was found. There are not enough hosts available., Code: 500"" [High,Triaged] - Assigned to James Slagle (james-slagle) 14:07:45 slagle: right, we still have this one 14:07:45 we are still waiting on the fix in nova to land 14:07:52 yeah, still seeing lots of those myself too 14:07:53 :( sigh 14:08:54 we need https://review.openstack.org/#/c/316031/ 14:08:58 iirc 14:09:31 someone is looking at https://bugs.launchpad.net/tripleo/+bug/1586906 already? 14:09:31 Launchpad bug 1586906 in tripleo "CI: periodic jobs fail becasue "no stack found"" [Undecided,New] 14:09:32 if there is a patch we know fixes it (or think it does) we could cherry pick it 14:09:54 that would mean building a custom Nova patch for each CI run... but we've done it before 14:10:02 dprince: yes we could, as long as we catch-up from master once the patch is merged upstream 14:10:25 dprince: i wasn't sure that still worked or not 14:10:31 dprince: +1 for this proposal 14:10:42 i couldn't see how in our scripts it would know to build the nova package 14:10:47 so we might need to add some plumbing 14:10:50 slagle: well... it used to. And it is a useful thing to have 14:11:05 slagle: so if we need to re-add this capability I'd think it generally useful 14:11:11 yes 14:11:16 think of it as an upstream defense mechanism 14:11:35 (as long as it's temporary) 14:11:42 dprince: do we take action on this? 14:11:52 i can have a look at it 14:12:04 slagle: it stopped working a while ago, but if you add nova to ZUUL_CHANGES as part of the cherry-pivk function then that should be all thats needed http://git.openstack.org/cgit/openstack-infra/tripleo-ci/tree/scripts/common_functions.sh#n43 14:12:27 derekh: yea, that's what i had tried, but i think i got the syntax wrong for ZUUL_CHANGES 14:12:59 #action slagle to cherry-pick nova/316031 into tripleo CI 14:13:11 slagle: ok, iirc depending on where you do it, it may need to be space delimited or "^" 14:13:43 anything else about bugs? 14:14:45 #topic Projects releases or stable backports 14:15:06 shardy asked us if we could produce a newton-alpha this week 14:15:25 slagle: do you still have time this week to work on it with me? 14:15:40 I never did it for tripleo (only puppet modules), but want to learn 14:15:44 slagle: also there might be something useful in here (or maybe not) https://review.openstack.org/#/c/227563/6 14:16:12 EmilienM: yea, it should be relatively straightforward, get the hashes from the last successful promote, and release those 14:16:17 * beagles lost network connectivity for a couple of minutes 14:16:19 #link http://releases.openstack.org/newton/schedule.html 14:16:41 EmilienM: hopefully we can get a promote this week, if not, we can use what was promoted last week 14:16:43 slagle: cool, let's do it this afternoon or tomorrow, depending if your agenda (mine is flexible) 14:17:16 #action EmilienM & slagle to promote CI & release newton-alpha 14:17:30 any question/feedback about release management? 14:17:44 do we have outstanding backports this week? 14:18:14 ok, let's move then 14:18:30 #topic CI 14:19:05 so we have some bugs already mentionned earlier that affect CI, but otherwise, do we have outstanding work regarding CI? 14:19:40 I'm asking for advise who can review and approve tripleo testing jobs: https://review.openstack.org/#/c/320957/ because talking "to all" doesn't work, I'd prefer to ask specific people. If somebody knows how it could be pushed, please advise me 14:20:34 sshnaidm: you need to ping infra prople about this patch, pabelanger for example 14:20:47 people* 14:21:11 sshnaidm: /join #openstack-infra and ask for review 14:21:16 EmilienM, ok, will ask pabelanger 14:21:19 cool 14:21:21 I have nothing on the current ci 14:21:23 rh2 deployment is making progress, I now have a publically available working rh2 overcloud 14:21:27 EmilienM, I do it every day 14:21:30 today started testing OVB on it 14:21:34 sshnaidm: sad story 14:21:41 will be redeploying several times over the next week or so, so still a WIP (but no blockers yet) 14:21:59 derekh: excellent progress :) 14:22:13 derekh: do you need help for something? 14:23:07 not at the moment, when I get it all up, I'll send on my scripts and take suggestions etc.. on how things could better be configured 14:23:16 cool 14:23:19 thx for this work 14:23:27 let's move to next topic then 14:23:27 EmilienM: and if I hit any blockers I can't figure out I'll ping #tripleo 14:23:32 ++ 14:23:36 #topic Specs 14:23:43 #link https://review.openstack.org/#/q/project:openstack/tripleo-specs+status:open 14:24:26 if you have a spec you want to discuss, please go ahead 14:25:44 most of specs already had nice insights in gerrit 14:25:59 if there is anything outstanding to discuss go ahead or we move to open discussion 14:26:52 #topic Open Discussion 14:27:22 (quite meeting today) 14:27:30 So I've got one 14:27:40 many of the developers have their own blogs 14:27:57 but would it be useful to setup a "team" blog and tie it in as well? 14:28:25 for tripleo posts? 14:28:33 I was thinking of setting up a github pages blog (simple one) and we could host it under the github.com/tripleo account 14:28:56 hey o/ am testing M..N upgrades using the current upgrades process... some rework needed in the 'start/stop all the things' logic we use tracked at https://blueprints.launchpad.net/tripleo/+spec/overcloud-upgrades-workflow-mitaka-to-newton currently testing. reviews welcome. would be nice to land sthing in N1 for the parent upgrades BP at https://blueprints.launchpad.net/tripleo/+spec/overclo 14:29:02 ud-upgrades 14:29:05 yeah, for team posts, that we perhaps collaborate on. Or for individuals who are part of the team and want to blog but don't have their own. 14:29:41 dprince: +1 to reminder about writing _sthing_ (thinking about the recent experience doing the upgrades process 'for real') 14:30:06 marios: it's all on https://review.openstack.org/#/q/topic:fixup_upgrades_scripts,n,z ? 14:30:27 EmilienM: for now... i also fixed the nova pinning etc today... see what else we find 14:30:38 EmilienM: that is the purpose of the blueprint... track work needed to adapt the current workflow 14:30:45 or we could just aggregate the tripleo related blogs (like we do for most planet) 14:30:46 dprince "but don't have their own" +1 14:31:20 EmilienM: see the whiteboard at https://blueprints.launchpad.net/tripleo/+spec/overcloud-upgrades-workflow-mitaka-to-newton 14:31:25 i mean ghpages would need additional reviewing 14:31:33 coolsvap: we would still aggregate them w/ planet. And also this would go into planet.openstack.org too 14:31:47 coolsvap: it would just be a team blog in addition to any individual blogs 14:31:53 marios: do we have plans to have a CI that test it? 14:32:15 EmilienM: good questino. matbu is looking at that i believe 14:32:27 EmilienM: and yes that is totally missing atm 14:32:32 because, yeah, we can review it, but without manual testing it's blind review 14:33:04 dprince, I am +1 on it just thinking about options 14:33:43 marios: sorry, what were you reminding me for again? The 'overcloud execute' stuff? 14:33:45 because any additional information on tripleo is useful 14:33:46 EmilienM: ack, let's take it up on the review. fwiw am testing it locally, which is how we landed these scripts originally (for liberty upgrades), because, we didn't have ci, same reason 14:34:10 right, I don't say we should block your patches because we have no CI for it 14:34:14 I just ask if we have plans for it 14:34:16 dprince: ha, nothing, you reminded me about saying i'd write sthing about tripleo when you said you asked ppl to 14:35:02 matbu: please let us know your plans, mailing-list for example 14:35:17 EmilienM: i think he just started looking 14:35:22 EmilienM: we have a tracking bug 14:35:22 sec 14:35:39 https://bugs.launchpad.net/tripleo/+bug/1583125 14:35:39 Launchpad bug 1583125 in tripleo "There is no 'major version' upgrades job for ci " [High,Triaged] - Assigned to mbu (mat-bultel) 14:35:46 (we had a CI topic and it was not raised :-P) 14:35:54 sry :( 14:36:29 cool work 14:36:46 #link Major upgrade work https://bugs.launchpad.net/tripleo/+bug/1583125 14:36:46 Launchpad bug 1583125 in tripleo "There is no 'major version' upgrades job for ci " [High,Triaged] - Assigned to mbu (mat-bultel) 14:36:55 EmilienM: yep i'll try to figure out this part asap 14:36:59 cool 14:37:18 EmilienM: i just discover the ci tripleo.sh script yesterday 14:37:33 dprince: maybe we could use mailing-list to poll people who would be interested by blogging in a common place 14:37:47 EmilienM: but afaik there is not much work there ... just hoping the upgrade will work 14:37:59 EmilienM: yep, I'll ask there too 14:38:14 do we have other thoughts for today? 14:39:32 dprince, gfidente: do you guys have something to mention about composable roles? 14:39:33 not from me. Thanks EmilienM 14:39:46 no blocker? progress is good? 14:40:00 EmilienM, we got swift proxy 14:40:13 I think we got to wire chains in cephstorage and compute nodes 14:40:20 so I'm getting back at the ceph roles 14:40:23 good 14:40:34 slow progress 14:40:36 EmilienM: oh, so we need to prototype a solution that works better to enable/disable things with Heat environments 14:40:37 but progress 14:40:47 slow progress, largely due to CI I think 14:41:01 dprince: let's fix nova, it should be better 14:41:07 yep 14:41:27 I'll keep meeting open 30 more seconds, if anything go ahead now 14:42:24 ok folks, have fun! /peace 14:42:26 EmilienM: ++ thanks 14:42:26 #endmeeting