08:02:03 #startmeeting ptl_sync 08:02:04 Meeting started Tue Oct 14 08:02:03 2014 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:02:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 08:02:08 The meeting name has been set to 'ptl_sync' 08:02:11 #topic Nova 08:02:13 Heya 08:02:37 John is fighting technology, so he might be late 08:02:55 hah 08:03:01 so, RC2 out 08:03:09 Hurrah 08:03:21 https://bugs.launchpad.net/nova/+bugs?field.tag=juno-rc-potential 08:03:38 A few bugs there, but nothing mature enough to justify RC3 I'd say 08:04:15 I should read the live migrate one 08:04:20 1377644 08:04:38 How up to date are you on the release notes ? https://wiki.openstack.org/wiki/ReleaseNotes/Juno 08:04:45 Oh, that's the exception thing 08:04:54 So, I need to start on the release notes 08:04:58 Its basically not attempted yet 08:05:18 I do have a summary of landed BPs though 08:05:37 ok, the sooner the better for the release notes, to avoid last-minute edit conflicts 08:05:38 http://www.stillhq.com/openstack/juno/000018.html 08:05:58 So, the "Key New Features" section is just that blog post, right? 08:06:04 Or do I need to stick with the headings there? 08:06:31 no, thta blogpost with a few edits should do 08:06:36 Excellent 08:06:39 I was hopign that was the case 08:06:46 not sure our users care about the migration reserve 08:06:47 I will cut and paste that in now then 08:06:53 Yeah, fair point 08:06:56 and some titles could use a bit of elaboration 08:07:02 That's the problem with including literally all 08:07:03 but the structure should be the same 08:07:12 Well, I shall dump and then iterate 08:07:17 exactly 08:07:25 it's a good first draft :) 08:08:09 starting EOD today we won't respin unless there is some critical issue like catastrophic data loss, or legal problems 08:08:12 or missing files 08:08:23 Ok 08:08:31 "regular" bugs will be documented as known issues 08:08:45 so that leaves today to do a RC3 over normal regressions 08:09:07 but I wouldn't consider bugs that are not already fixed in master 08:09:13 I have seen zero discussion of requiring a rc3 08:09:15 So I think we're ok 08:09:22 right 08:09:55 johnthetubaguy: we were saying nothing justifies a late respin at this point 08:10:27 * johnthetubaguy grumpy that IRC forced him to use webchat 08:10:49 ttx: mikal: yeah, I haven't seen anyone screaming about something 08:11:09 johnthetubaguy: we'll keep an eye on that today. Like I told mikal: 08:11:38 starting EOD today we won't respin unless catastrophic dataloss, legal issues, missing files 08:11:56 ttx: cool, makes sense 08:12:06 so that leaves today for "normal" respin -- but we wouldn't consider any backport that's not already in master 08:12:16 and lookgin at the list right now, nothing jumps out 08:12:38 https://bugs.launchpad.net/nova/+bugs?field.status%3Alist=FIXCOMMITTED 08:13:02 we'll keep an eye on that list today just in case 08:13:26 OK, anything else ? 08:13:40 For the release or in general? 08:13:47 in general 08:14:02 So, I feel like I need to start paying more attention to the summit schedule 08:14:07 But I will get on that ASAP 08:14:25 mikal: are we going to cover that in some nova-meeting? 08:14:39 johnthetubaguy: so, we have a page of ideas 08:14:46 johnthetubaguy: we now need to work out how to turn that into a thing 08:14:57 johnthetubaguy: which I guess does involve our meetings 08:15:00 you need to split the topics between scheduled slots and the meetup basically 08:15:05 Yeah 08:15:06 mikal: yeah, I am thinking we discuss the proposed thing in the nova-meeting, see if people agree 08:15:12 Kyle pinged me about the nova-net upgrade path 08:15:17 I suggested some shared time on Friday 08:15:24 Insteafd of trying to clash with another nova session 08:15:36 scheduled slots are good when the discussion is easily timeboxed and would benefit from being advertised on the schedule (needs nova-external input) 08:15:53 I am not sure the nova-net thing can be timeboxed 08:16:04 I don't have a good read on that 08:16:21 "meetup" slots are good for project process discussion, setting cycle priorities at the end of the week etc 08:16:33 mikal: some shared nova + neutron meet up time sounds good though, even if its a slot where we have nova sessions and no neutron sessions, or something like that 08:16:47 a number of things obviously fall in the middle between those two extremes 08:17:07 ttx: is neutron two full days of sessions too? 08:17:12 so I would split the ideas between 3 groups 08:17:18 Or is there a gap we could schedule the upgrade thing into? 08:17:27 there is a gap 08:17:33 * ttx shares 08:17:42 Last summit there was a spreadsheet of doom 08:17:46 Do we have one of those again? 08:18:26 yes, see pm 08:18:41 Thanks 08:18:52 from Wed 1:50pm to Thu 10:30 am, no Neutron session 08:18:53 Oh, so Wed arvo might work for example 08:18:54 Cool 08:19:42 the canonical layout is at http://kilodesignsummit.sched.org/ now 08:19:56 * ttx spots a few errors in the spreadsheet 08:20:06 Oh, nice 08:20:10 Fixy fixy 08:20:53 * johnthetubaguy_w lets hope the french don't ban me 08:21:09 Why would they? 08:21:12 Did you hurt their feelings? 08:21:28 lol, no idea 08:21:49 So, I think we sound like we're done? 08:21:58 Apart from me promising to think more on the summit? 08:22:08 I also intend to announce a mid-cycle meetup real soon now [tm] 08:22:23 mikal: let us driver folks know how we can help with that 08:23:16 johnthetubaguy: yrp 08:23:19 mikal: yes done 08:23:31 Basically two companies offered, and then one backed down 08:23:35 So the other wins! 08:23:39 mikal: does your meetup include skiing ? 08:23:57 ttx: no, that's why Company One backed down, too much snow 08:24:08 ttx: so we're going somewhere warm and will revisit Company One in Lemming 08:25:11 * johnthetubaguy looks forward to mid-cycle announcement 08:25:13 alright then, ttyl 08:25:17 Laters! 08:25:20 bye 11:23:09 ttx, hey, will any projects have rc3? 11:23:25 Heat already had one 11:23:32 Trove will likely have one too 11:23:44 ttx, heh 11:45:02 ttx: knock, knock ... 11:45:27 #topic Ceilometer 11:45:42 * ttx is busy looking up fine restaurants 11:45:56 eglynn: howdy 11:46:14 RC2 is out, a few candidates up for RC3 respin 11:46:15 no shortage of those in Paris :) 11:46:35 cool, I've no reason to think we'll need a ceilo RC3 11:46:42 ... famous last words :) 11:46:45 https://bugs.launchpad.net/ceilometer/+bugs?field.tag=juno-rc-potential 11:47:00 1379808 was proposed for backport by jd__ 11:47:38 not sure how blocking it is. It's ont of those that we can't easily backport post-release 11:47:56 yeah he mentioned it to me also 11:48:13 we can certainly respin quickly in the next hour(s) 11:48:18 if that one is wanted 11:48:34 It's certainly self-contained enough and low risk 11:48:55 actually now that I'm giving https://bugs.launchpad.net/ceilometer/+bug/1378742 a second look 11:49:31 hmmm, that one might actually justify respining 11:50:11 OK, let's go for it since both are very self contained 11:50:26 hmm, ok, let's do it fast 11:50:32 cool, thanks! 11:50:34 anything else critical in https://bugs.launchpad.net/ceilometer/+bugs?field.status%3Alist=FIXCOMMITTED ? 11:51:52 nope, we can live without those 11:52:18 eglynn: ok, propose backport for https://review.openstack.org/#/c/126871 now 11:52:21 and we'll push it 11:52:31 cool, I'm on it 11:52:39 approving julien's one 11:53:54 eglynn: after this one, we'll respin only on showstoppers that can't possibly be documented as known issues 11:54:00 like a legal issue 11:54:06 or a missing file in the tarball 11:54:13 or some catastrophic failure 11:54:17 got it 11:54:32 so after that backport you should finalize https://wiki.openstack.org/wiki/ReleaseNotes/Juno 11:54:42 and then safely proceed to Kilo design summit session planning :) 11:54:52 questions ? 11:55:30 yeah I've already done a first cut on the release notes 11:55:41 https://wiki.openstack.org/wiki/ReleaseNotes/Juno#OpenStack_Telemetry_.28Ceilometer.29 11:55:51 nothing else from me 11:56:03 cool, ping me when you have the backport up 11:57:55 https://review.openstack.org/128249 12:05:49 SergeyLukjanov: around? 12:06:57 dhellmann: ready when you are 12:24:52 ttx: sorry, running late this morning but ready now 12:26:39 #topic Oslo 12:26:44 dhellmann: o/ 12:26:49 ttx: o/ 12:26:50 Any red flag from your side ? 12:27:33 I think we're fully in Kilo mode. I'm not aware of any major issues. 12:27:38 dhellmann: I actually had a question for you about openstack/requirements 12:27:43 k 12:27:55 there seems to be some confusion about the use of it by non-integrated projects 12:28:17 random projects asking for additional requirements that are not actually wanted by integrated projects 12:28:43 It feels like we supported that in the past but are clamping down 12:28:51 yeah, they used to need to be included for the mirror 12:29:07 but they don't anymore ? 12:29:16 no, we have a full pypi mirror now 12:29:25 I forget which package we're using, bandersnatch maybe? 12:29:31 so why are they asking ? 12:29:31 we had built our own before 12:29:43 there are a couple of reasons 12:29:46 do they have some stale requirements check job ? 12:29:58 the doc team wants to sync requirements into their projects 12:30:08 they have 10+ repos 12:30:40 fungi, AJaeger, and I sort of decided yesterday that would be ok to keep doing, since debian at least does build the docs 12:30:41 so ideally a project could "consume" common requirements and add extra ones on their side 12:30:54 well, the way it is built now, it's all or nothing 12:31:05 hrm, maybe that's not true 12:31:20 we could have the sync job run but leave the check job out of the project 12:31:34 dhellmann: how about we abuse the crossproject meeting today to discuss that a bit ? Unless the release is on fire 12:31:36 of course that means they couldn't have a different version of anything than we have globally 12:31:56 that works, although I think the projects that care won't necessarily be in the meeting 12:32:09 solum was another one that came up 12:32:18 sure, it's more so that infra/QA/us are all on the same page before we push that down their throats 12:32:47 makes sense -- I thought the decision was already more or less agreed to, but let's confirm that instead of me assuming it 12:32:53 currently I'm not even sure what to answer them ;) 12:33:01 cool. 12:33:04 Anything else ? 12:33:37 ttx, I'm here, sorry for the delay, forgot to enable notifications :( 12:33:45 I'll remind everyone to send liaisons, but otherwise nothing 12:34:01 dhellmann: ok, have a good day then 12:34:05 SergeyLukjanov: hi! 12:34:07 ttx: ditto 12:34:12 #topic Sahara 12:34:28 SergeyLukjanov: haven't seen anything justifying a RC3 ? 12:34:53 Nothing in rc-potential, a few bugs fixed in master but nothing stand out: https://bugs.launchpad.net/sahara/+bugs?field.status%3Alist=FIXCOMMITTED 12:35:01 ttx, we have a regression with direct engine and nova-network, but I don't think that we need to delay release because of it 12:35:18 ttx, I think we'll test and bacl port it to stable/juno 12:35:27 SergeyLukjanov: is that fixed in master already? 12:35:34 ttx, not yet 12:35:48 SergeyLukjanov: ok, then it shall be documented as a known issue in the release notes 12:36:02 https://wiki.openstack.org/wiki/ReleaseNotes/Juno 12:36:07 ttx, the fix will be not very safe, so, yeah, we'll document it 12:36:21 ttx, we have release notes draft https://wiki.openstack.org/wiki/Sahara/ReleaseNotes/Juno 12:36:31 ttx, will move it today to https://wiki.openstack.org/wiki/ReleaseNotes/Juno 12:36:38 sounds good 12:36:51 SergeyLukjanov: anything else ? 12:37:02 ttx, I think nope, everything going well 12:37:21 SergeyLukjanov: you'll have to start working on summit agenda 12:37:35 dhellmann: that's true for you as well ^ 12:37:39 ttx, in fact this regression is with the deprecated non-heat engine with custom configs, so, non critical 12:37:53 ttx, already working on it, should have a good results for the next meeting 12:37:55 ttx: we've started 12:38:08 ok, let me know if you have questions 12:38:34 I'll send you all an email with the link to the website where you can update the (tbd) sessions 12:38:49 ttx, ack, thx 12:38:51 when the final schedule is ready 12:40:49 mestery, nikhil_k: ping me when around so that we anticipate your need for RC3 13:24:44 ttx: pong 13:27:19 mestery: hey, wanted to ask you in advance of our sync if you think we'll need a RC3 13:27:31 if yes, we want to push it to gate asap 13:27:40 It's not looking likely, there were no release critical bugs we as a team could find in our meeting yesterday 13:27:46 I think we're good with RC2 being Juno final. 13:28:04 A quick search this morning didn't turn anything up either. 13:28:06 Someone proposed https://bugs.launchpad.net/neutron/+bug/1378525 for backport 13:28:18 should we still ignore that ? 13:28:22 Yes, we spoke about that one, and we determined it can be release noted. 13:28:25 I'll comment in the bug 13:28:51 mestery: ok then 13:29:01 mestery: we can formally do our sync now if that suits you 13:29:17 ttx: Actually, one sec, looking at the bug. 13:29:20 It's a 1 line change. 13:29:28 Hmmm .... 13:29:38 Still not sure it's worth spinning a new release for. 13:29:43 And yes, lets sync now. 13:29:48 #topic Neutron 13:30:11 It's sufficiently self-contained that we can respin on that, if you think it's a critical issue 13:30:47 Let me talk to Akihiro, he made a comment about Horizon. 13:30:53 I will know in < 1 hour I think once I get a hold of him. 13:31:02 I agree, it may be worth respinning just for this bug since it's contained. 13:31:11 Shall I propose the cherry-pick in anticipation? 13:31:19 it's already cherrypicked 13:31:26 Ah, cool. :) 13:31:28 https://review.openstack.org/#/c/126711/ 13:31:54 Perfect. 13:32:02 so it's not too disruptive as long as we push it now 13:32:06 Right. 13:32:08 I'll know soon. 13:32:13 <1 hour is probably ok 13:32:26 checking https://bugs.launchpad.net/neutron/+bugs?field.status%3Alist=FIXCOMMITTED for other easy ones 13:33:06 Me too 13:33:39 Sukhdev asked me to consider this one: https://bugs.launchpad.net/neutron/+bug/1373652 13:33:45 It's a larger change, but contained to the Arista driver 13:34:03 I'm inclined to include that one if you are ok with it. 13:34:07 looking 13:34:55 more risky 13:35:05 also easily backportable imho 13:35:14 and documenatble 13:35:17 Yeah, lets leave it. 13:35:19 I agree 13:35:23 and .1 is 6 weeks out or so? 13:35:26 So it's fine to land then. 13:35:29 With more testing, etc. 13:35:59 yeah 13:36:19 https://bugs.launchpad.net/neutron/+bug/1232525 maybe 13:36:35 https://bugs.launchpad.net/neutron/+bug/1365226 also 13:36:39 although it's probably just cosmetic 13:37:09 Yes to 1232525, it may make backports easier 13:37:13 What about 1365226? 13:38:11 hmm, was never merged to master ? 13:38:18 * mestery checks again 13:38:34 Right 13:38:35 Sorry 13:38:36 :) 13:38:43 bit weird 13:38:57 oh, driver removed 13:38:58 bah 13:39:05 so just one, if you confirm it 13:39:10 Yeah 13:39:19 unless we also do 1232525 13:39:28 Yes, I think 1232525 is good as well 13:39:37 I am ok with that one, and I can propose the cherry-pick this morning yet 13:39:53 propose it now, add the bug reference to the commit message 13:40:02 if it's ready when the other is, it won't delay 13:40:10 Ack 13:40:51 #info Potential RC3 over 1378525 and 1232525, to be confirmed in the next hour 13:41:48 mestery: what's the state of your release notes ? 13:42:11 ttx: I was going to release note the L3 HA bug, but now we're merging it. Otherwise, I'll take another run at them today. 13:42:19 They are in good shape, but I'll confirm today. 13:43:00 ok, then you need to start working on the design summit agenda 13:43:06 once the release is mostly done 13:43:21 Way ahead of you :) 13:43:25 We have that almost wrapped too 13:43:27 On an etherpad 13:43:36 By end of tomorrow it will be finalized I think 13:43:37 :) 13:44:19 ttx: https://review.openstack.org/128288 13:45:15 mestery: OK, let's wait until we confirm the RC3 13:45:32 ttx: Ack 13:45:37 But it's in the queue now at least 13:45:48 ok then, let me know when you've come to a decision about it 13:46:25 ttx: Will do, email out to amotoki. 13:46:31 ttx: The default is we don't do it, which is fine too. 13:46:54 agreed 13:47:34 jgriffith: ready when you are 13:47:39 ttx: Thanks! 13:58:33 ttx: amotoki just confirmed, RC3 is good to go, lets merge those two fixes. 13:58:59 ok, opening 13:59:24 ttx: ACK 14:00:57 https://bugs.launchpad.net/neutron/+milestone/juno-rc3 14:01:23 Looks good ttx, thanks! 14:01:49 ttx: https://review.openstack.org/#/c/126711/ you can merge this one now or remove your -2. 14:02:02 ttx: jinx ;) 14:02:12 gating 14:02:24 jgriffith: around? 14:02:36 ttx: Thanks! 14:02:48 will tag as soon as those two merge 14:03:11 although there is quite a bit of delay in the post queue right now 14:05:19 ttx: Lets hope it makes the journey safely. 14:14:01 dolphm: ready when you are 14:14:12 ttx, o/ here as well. 14:14:19 ttx: o/ 14:14:22 morganfainberg: o/ 14:14:23 #topic Keystone 14:14:30 Looks like you're good? 14:14:36 i believe so! 14:14:55 no reason for an rc3 at this point 14:15:23 so.. by EOD today we'll only respin on show stoppers, like legal issues, missing files in tarballs etc 14:15:50 everything else will go to known issues in the release notes 14:16:10 Looks like your release notes are mostly ready ? 14:18:58 dolphm ? morganfainberg ? 14:19:09 ttx, looking at release notes, but i think so 14:19:16 OK. How is summit planning going ? 14:19:38 ttx, have most of the sessions lined up wiht one open slot to be discussed at the keystone meeting today 14:20:20 ttx, descriptions and specific time slots are still variable, at the top of https://etherpad.openstack.org/p/kilo-keystone-summit-topics i've added a tentative list/schedule 14:21:01 when is the "move this stuff to the actual schedule" date? 14:21:01 I'll send info on how to push updates to the schedule 14:21:03 k 14:21:29 date would be between reception of email and one week before summit starts 14:21:41 ttx, ack 14:21:55 any other question ? 14:22:04 i think thats it for me. 14:22:04 anything you'd like to discuss at the meeting today ? 14:22:17 ok then, talk to you later! 14:22:20 oh, once we move to kilo, mind changing this time slot? 14:22:26 not at all. 14:22:34 it's going to be 0615 for me once we timeshift 14:22:40 a bit early :) 14:22:47 we'll discuss the reshuffle them all with the DST end anyway 14:22:53 sounds good! 14:24:16 I'll bring up that keystone team should do a once over on the release notes today at the meeting, but thye look complete to me. 14:25:15 david-lyle: ready when you are 14:25:21 ttx: ready 14:25:26 #topic Horizon 14:25:39 RC2 is out 14:26:05 rc-potential has https://bugs.launchpad.net/horizon/+bug/1379761 14:26:31 yes, this has to be fixed 14:26:59 or documented, if we can't get it fixed in time 14:26:59 can't run Horizon without DEBUG=True in the settings for Django, which is bad 14:27:15 david-lyle: how is the fix looking so far ? 14:27:37 complicating factor is that the broken code is in a dependency 14:27:53 I prefer to wait for the fix to land in master before opening formally an RC3 window (prefer not to open windows I don't know how to close) 14:28:07 may require a monkey patch to get Juno released and fix appropriately upstream 14:28:39 ttx: (bah, sorry for dissappearing, but everything sounds good. thanks morganfainberg!) 14:28:40 I don't have a fix yet, still working through cleaner ways of working around upstream bug 14:29:22 I don't mind waiting on the RC3 window until we have a fix 14:29:31 hmm, ok 14:29:55 we can't really release without a fix 14:30:07 unfortunately 14:30:16 is that something we could have detected earlier ? 14:30:40 most likely, but it's an odd issue 14:30:55 we needed to run this as a production release 14:31:31 the upstream code just punted when they didn't really understand django settings 14:32:01 david-lyle: ok, so ideally we need that fix propoesed, reviewed and merged today 14:32:11 does that look even doable ? 14:32:14 that's my hope 14:32:33 it may not be elegant, but we should have something that works 14:32:37 that's for some large definition of "today" -- meaning I can get up tomorrow morning and get the backport in 14:33:04 so getting it proposed and reviewed by your EOD should be enough 14:33:52 ok 14:34:10 anythign else ? How are your release notes coming up ? 14:34:35 hmm, more work needed there but if everything goes ok today you should have time for that tomorrow 14:34:42 or find someone to delegate that to 14:34:43 I have them mostly completed, just a little cleanup and I can add them 14:34:48 ok, cool 14:35:02 then post-release you can focus on yoru design summit session schedule :) 14:35:07 your* 14:35:29 maybe we'll have that figured out by Nov ;) 14:35:46 ok, questions? 14:36:01 nope 14:36:07 going to go tackle a bug 14:36:16 go go go 14:36:19 talk to you later 14:36:26 ack 14:36:31 jgriffith: around now ? 14:51:34 ttx: hi 14:51:47 nikhil_k: hi! 14:52:05 nikhil_k: could we sync early to discuss a potential Glance RC3 ? 14:52:12 (i.e. now ?) 14:52:15 yeah sure 14:52:21 was just saying.. 14:52:24 #topic Glance 14:52:26 I'm a bit uncertain either way about 14:52:35 oh ok :) 14:52:42 RC3 ! 14:52:47