20:00:19 <johnsom> #startmeeting Octavia 20:00:20 <openstack> Meeting started Wed May 3 20:00:19 2017 UTC and is due to finish in 60 minutes. The chair is johnsom. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:00:24 <openstack> The meeting name has been set to 'octavia' 20:00:27 <johnsom> Hi folks! 20:00:45 <xgerman> O/ 20:00:54 <nmagnezi> o/ 20:01:42 <johnsom> Maybe I can summon rm_work to this channel... 20:01:48 <rm_work> o. 20:01:50 <rm_work> o/ 20:01:52 <johnsom> grin 20:01:56 <johnsom> #topic Announcements 20:02:12 <johnsom> The summit is next week. I plan to cancel the weekly meeting 20:02:19 <johnsom> #link https://etherpad.openstack.org/p/octavia-pike-summit 20:02:53 <johnsom> That is an etherpad with our talk details and a list of folks (not attending in this case). 20:02:55 <rm_work> kk 20:03:29 <johnsom> The week of the 15th and 22nd I will be on vacation and likely in and out of cell service 20:03:48 <johnsom> Does someone want to volunteer to run the IRC meetings or should be cancel those too? 20:03:52 <rm_work> everyone in 'not attending' gets frowny faces 20:04:03 <rm_work> I'll also be out that week 20:04:11 <johnsom> Yes, frowny faces are appropriate 20:04:29 <johnsom> That is two weeks actually 20:04:30 <rm_work> I'll be at *another* conference 20:04:34 <xgerman> I am in the week after the summit but then out two weeks 5/19-6/2 20:04:34 <rm_work> ah k 20:04:46 <rm_work> well, week after Boston summit, I'm at an internal conference thing 20:04:57 <xgerman> nice! 20:04:58 <johnsom> Ok, so maybe we should skip a few 20:05:03 <rm_work> but will be back after that 20:05:11 <rm_work> when does the final Pike release freeze happen? 20:05:30 <johnsom> #link https://releases.openstack.org/pike/schedule.html 20:05:39 <johnsom> Pike-2 is June 05 20:05:40 <rm_work> most importantly maybe, feature freeze 20:05:41 <rm_work> I REALLY want to get single-create / cascade-delete in 20:05:42 <rm_work> for Pike 20:05:43 <rm_work> OK 20:05:46 <rm_work> well 20:05:54 <rm_work> we'll be kinda cramped for time I guess with everyone in and out 20:06:00 <rm_work> but hopefully we can do it 20:06:06 <johnsom> Feature freeze for Octavia is Jul 24th 20:06:08 <johnsom> Pike-3 20:06:12 <rm_work> aaah ok 20:06:18 <rm_work> excellent 20:06:28 <rm_work> somehow I always thought the conference lined up with a release 20:06:30 <rm_work> but i guess not 20:06:39 <johnsom> It did until the PTG happened 20:06:43 <rm_work> ahh 20:06:45 <rm_work> ok 20:07:08 <johnsom> Pike release is now Aug 28th 20:07:15 <rm_work> ok, until July is great, we should be able to 100% wrap up APIv2 in Octavia 20:07:22 <johnsom> YES! 20:07:28 <xgerman> famous last words 20:07:40 <rm_work> umm, do we also need to start looking at the driver shim stuff again? 20:07:44 <johnsom> Except maybe the vendor driver support.... But, hey, we made huge progress 20:07:46 <rm_work> I assume people will ... want that 20:07:50 <rm_work> yeah lol 20:08:20 <johnsom> Yeah. I'm pretty much expecting we will not have providers in Octavia V2 API until Queens 20:08:21 <rm_work> I wish I could help look at that, but I think the internal goodwill for me working pretty much 100% upstream recently is about to dry up 20:08:30 <johnsom> Given the huge drop in contributors 20:08:38 <rm_work> and it's a lot harder to tie that to our use cases here 20:09:08 <xgerman> indeed - 20:09:16 <johnsom> I had hoped we had some vendors that were interested in working on that, but... 20:09:21 <rm_work> :/ 20:09:37 <nmagnezi> johnsom, maybe f5? m-greene? 20:09:45 <rm_work> is kobis still around? 20:09:47 <xgerman> *hoping* 20:09:51 <johnsom> Ok, so other announcements, the systemd as default in devstack happened. Joy 20:10:12 <johnsom> Also, the uwsgi goal we JUST finished is now a moving target..... 20:10:33 <johnsom> The py35 goal is pretty much done except for having voting gates (lbaas) 20:11:15 <rm_work> we have voting gates (octavia), no? 20:11:21 <johnsom> Any other announcements? 20:11:37 <johnsom> rm_work yes, we have the octavia gates voting, but you wanted to hold off on lbaas 20:11:40 <rm_work> i'd say it's done 20:11:43 <rm_work> why do those matter 20:11:45 <rm_work> that project is dead 20:12:02 <johnsom> Yeah, um, not technically yet.... 20:12:27 <rm_work> also for the record, i'm glad we held off, because with the roving VM build errors (OVH acting up again?) the more voting scenarios we have, the worse the recheck-hell gets 20:12:38 <rm_work> as it is, we have to recheck about 1/3 of the changes that we try to merge 20:12:56 <rm_work> with yet more scenarios that'd be like 2/3 20:12:56 <johnsom> Yeah, OVH is a problem. I'm don't know why some of their hosts work and some libvirt crashes 20:13:27 <johnsom> Maybe we get "creative" and make them voting, merge the "goal complete" then switch them back.... 20:13:39 <rm_work> lol 20:13:46 <rm_work> i don't think anyone is watching that closely anyway 20:13:52 <rm_work> i'm still not sure why those would even matter 20:13:55 <johnsom> Hey, life changes in the openstack world 20:13:56 <rm_work> we have py3x gates 20:14:00 <rm_work> in octavia 20:14:01 <rm_work> they vote 20:14:12 <rm_work> and the py3x gates in neutron-lbaas vote *in neutron-lbaas gates* so 20:14:15 <rm_work> how is this a problem 20:14:20 <johnsom> Well, I can try and see what we get for feedback. 20:14:34 <rm_work> i think the fact that we co-gate at all with n-lbaas is generous 20:14:46 <johnsom> rm_work they don't vote in neutron-lbaas 20:14:46 <rm_work> if it were up to me i'd just remove them from octavia 20:15:02 <rm_work> wait really? 20:15:07 <johnsom> Hmm, maybe we should do that, make them vote over there 20:15:40 <rm_work> yeah why wouldn't they vote on their own project, lol 20:16:01 <johnsom> Well, the same reason we didn't have octavia voting. Tons of other projects breaking us 20:16:43 <johnsom> Ha, now that is interesting. The global requirements change didn't pass pep8... lol This place will drive me crazy some day..... 20:16:50 <johnsom> #link https://review.openstack.org/#/c/462111/ 20:17:07 <johnsom> Ok, maybe I will move forward with voting in neutron lbaas 20:17:13 <johnsom> That seems "fair" 20:17:43 <rm_work> then when they start break and no one fixes them because no one cares about neutron-lbaas, we can finally admit it is dead :P 20:17:52 <johnsom> Oye, some eager beaver: N537 Log messages should not be translated! 20:18:09 <rm_work> lol they're trying to ENFORCE THAT now? 20:18:12 <rm_work> seriously 20:18:13 <johnsom> Great, another housekeeping project, go fix those crapy log change patches 20:18:17 <rm_work> they can't just let us choose 20:18:27 <johnsom> Evidently 20:18:34 <nmagnezi> johnsom, maybe this will resolve the issue (I have not looked at the code) https://review.openstack.org/#/c/447921/ 20:18:51 <nmagnezi> #link https://review.openstack.org/#/c/447921/ 20:19:10 <johnsom> Yeah, those patches need some work.... 20:19:22 <johnsom> They were just global replacements that broke a bunch of stuff 20:19:36 <johnsom> Well, heck, throw one more cleanup on the pile.... 20:19:50 <johnsom> We have uwsgi, the stable branch issues, and now this 20:19:58 <johnsom> I guess my afternoon is booked. 20:20:10 <nmagnezi> sorry about that :) 20:20:34 <nmagnezi> if it was not a bit late for me i would tackle those log translation issues 20:20:45 <johnsom> nmagnezi Ah well, I signed up for this.... grin 20:21:48 <johnsom> rm_work maybe I should not sign up for the stats patch. Maybe I should leave that up to someone with time and do this stuff and try to get back to the api-ref 20:22:39 <johnsom> Ok, moving on... 20:22:47 <johnsom> #topic Brief progress reports / bugs needing review 20:23:15 <johnsom> I have been doing developer training for some new folks interested in working on Octavia. So hopefully some new faces soon. 20:23:46 <rm_work> I've got Single-Create and Cascade-Delete ready for review... 20:23:49 <johnsom> Otherwise it's been trying to track down some devstack problems I was having and doing reviews.... 20:23:49 <rm_work> Single-Create: https://review.openstack.org/459032 20:23:56 <rm_work> Cascade-Delete: https://review.openstack.org/461182 20:24:01 <xgerman> worked on presentations, smaller stuff 20:24:11 <johnsom> Yeah, I will get back to that too. The devstack issues screwed up my testing for that 20:24:13 <rm_work> Trying to fix stable gates for newton 20:24:22 <rm_work> and a small direction vote: https://review.openstack.org/#/c/461344/ 20:24:28 <johnsom> Same, worked on the 101 presentation and the project update presentation. 20:25:00 <rm_work> also Pagination is ready for eyes / testing: https://review.openstack.org/#/c/382147/ 20:25:01 <johnsom> rm_work should we remove "vote" from the commit message? 20:25:05 <rm_work> I have it on top of single-create 20:25:09 <rm_work> ah maybe 20:25:10 <xgerman> worked on lab presentation and soms small stuff 20:25:15 <rm_work> I guess maybe it's less of a vote? dunno 20:25:24 <johnsom> I think we vote with gerrit 20:25:40 <rm_work> johnsom: right, that's why it says vote :P 20:25:49 <rm_work> johnsom: did we decide if i could remove the tenant_id stuff already tho? lol 20:25:52 <rm_work> it was deprecated in like 20:25:53 <rm_work> Liberty 20:25:54 <johnsom> Yeah, I think sort / pagination is working I tested it. I just want to test the single-call stack with it and call it good. 20:26:02 <johnsom> Where are we at with filters? 20:26:06 <rm_work> I have to start it 20:26:16 <johnsom> Ok 20:26:17 <rm_work> I really wanted to get the big stuff I have sitting merged first though 20:26:21 <rm_work> before I start yet another thing 20:26:25 <rm_work> that'd be on the end of the chain 20:26:29 <johnsom> Was it REALLY liberty???? 20:26:33 <rm_work> I am pretty sure 20:26:37 <johnsom> When I looked it gets removed in R 20:26:42 <rm_work> umm 20:27:00 <johnsom> Hmm, not sure now, R seems wrong 20:27:39 <rm_work> Ah it was deprecated with keystone v3 20:27:46 <rm_work> or rather, when keystone v2 was deprecated 20:27:48 <rm_work> which WAS icehouse 20:27:56 <rm_work> but the deprecation was reverted 20:28:01 <rm_work> and re-deprecated in Mitaka 20:28:08 <johnsom> Yeah, it was newer than that 20:28:17 <nmagnezi> re-deprecated.. lol 20:28:23 <johnsom> Yeah, I guess it can get pulled 20:28:24 <nmagnezi> gotta love openstack. 20:28:27 <rm_work> so M 20:28:34 <rm_work> so N/O and P is good :) 20:28:38 <rm_work> I'll make that changeeeeee 20:28:43 * rm_work is excite 20:28:50 <johnsom> Please put a release note in for that 20:29:12 <rm_work> ok 20:29:45 <johnsom> nmagnezi how is tripleo coming? Is that all merged now? 20:29:51 <rm_work> I am super super pleased that we'll be tenant_id-less tho 20:30:02 <johnsom> Yeah, I HATE it too 20:30:02 <rm_work> i know nmagnezi still has some o-hm cleanup patch pending 20:30:41 <nmagnezi> johnsom, I was OOO most of this week (and also for the rest of this week), but I'll try to push something to https://review.openstack.org/#/c/460524/ tonight (which is now, my time) 20:31:06 <nmagnezi> johnsom, I hope to wrap this up next week at most 20:31:17 <johnsom> Ok, cool 20:31:24 <xgerman> +! 20:31:36 <nmagnezi> johnsom, we deployed Octavia with tripleO and ran all scenario and api tests, which passed OK 20:31:37 <xgerman> were wndering if we can announce in BOS — guess we better skip ;-) 20:31:39 <nmagnezi> so we are in a good shape 20:31:58 <xgerman> maybe we can mention it… 20:32:02 <nmagnezi> xgerman, up to you :) 20:32:07 <johnsom> Is there a link to some docs for the tripleo stuff? 20:32:14 <nmagnezi> xgerman, If anything changes before the summit I can ping you for that 20:32:20 <xgerman> ok 20:32:28 <nmagnezi> johnsom, I think so, I'll have to check with beagles 20:32:44 <johnsom> Ok, if someone can forward I can advertise a bit 20:32:53 <johnsom> #topic Boston Summit planning 20:32:58 <nmagnezi> #link https://review.openstack.org/#/c/447496/ 20:33:01 <nmagnezi> johnsom, xgerman ^ 20:33:15 <johnsom> Speaking of, anything more we need to discuss for the summit? 20:33:32 <johnsom> I think we need to put together a deck for the main LBaaS talk 20:33:43 <rm_work> who is on what 20:33:44 <xgerman> +1 20:33:50 <rm_work> is it just me and xgerman on everything? 20:33:55 <xgerman> yep 20:34:00 <rm_work> we have ... 3 talks and a lab still? or just 2 talks and a lab? 20:34:01 <xgerman> we will throw a coinf before each talk 20:34:15 <johnsom> 3 talks and a lab 20:34:19 <xgerman> I think 3 talks + lab 20:34:22 <nmagnezi> i will look forward for the recordings :-) 20:34:24 * johnsom sounds like a band 20:35:07 <rm_work> ugh I'm not looking forward to this part... i'm starting to get sick too, i think (seriously) 20:35:18 <rm_work> hopefully will not impact my ability to speak 20:35:32 <johnsom> I hope you don't play a Doug on German 20:35:40 <xgerman> +1 20:35:45 <rm_work> T_T 20:35:59 <johnsom> cough, cough, ugh I can't do the talk, I will just watch.... My welcome to summit presenting 20:36:09 <johnsom> grin 20:37:12 <johnsom> Ok, so LBaaS talk slides... 20:37:29 <johnsom> I can work on all the code "stuff" or start a deck for that 20:37:38 <xgerman> yeah, sound sgood 20:37:42 <rm_work> did someone confirm how many talks we have? 20:37:45 <rm_work> i disconnected briefly 20:37:46 <xgerman> I can throw in some OSA stuff 20:38:01 <johnsom> rm_work three talks, one lab 20:38:06 <rm_work> ugh k 20:38:15 <rm_work> i thought we decided one of the talks was kinda redundant 20:38:20 <xgerman> well, it’s only one talk per day? 20:38:24 <johnsom> The project update talk is going to be easy 20:38:30 <rm_work> k 20:38:31 <johnsom> No, one day has two 20:38:35 <xgerman> yeah, redundant. = practice 20:38:46 <rm_work> >_< 20:39:00 <johnsom> At least I got them to move them so they weren't back-to-back 20:39:29 <xgerman> nice! 20:40:08 <johnsom> #topic Open Discussion 20:40:09 <rm_work> johnsom: is removing tenant_id a ... deprecation? 20:40:13 <rm_work> or a feature? :P 20:40:30 <rm_work> i think it was ALREADY a deprecation, right? 20:40:38 <rm_work> but does it fit under that category again now? 20:41:24 <johnsom> Yeah, that is awkward isn't it. I would say either of those or "other" 20:41:32 <rm_work> k 20:41:57 <johnsom> I guess deprecation can just say, timer expired, this is history 20:41:58 <johnsom> ha 20:42:21 <johnsom> I'm going to skip the API thing again (it wasn't on the agenda for today 20:42:44 <xgerman> yep 20:43:12 <johnsom> Evidently there is more info on the systemd stuff here: 20:43:13 <johnsom> #link https://review.openstack.org/#/c/461716/2/doc/source/development.rst 20:43:30 <rm_work> there's GOT to be a way to get logs working properly 20:43:41 <rm_work> woot: https://review.openstack.org/461344 20:43:53 <johnsom> Any other topics for today? 20:44:40 <nmagnezi> nothing on my end 20:44:55 <m-greene_> apologies for being off-grid for several weeks. I saw my nick referenced in my scrollback history, but not sure the context 20:45:00 <nmagnezi> I'll keep you posted if the ovs cleanup patch makes progress before the summit 20:45:06 <rm_work> k 20:45:07 <xgerman> k 20:45:09 <m-greene_> something about hoping a vendor would contribute 20:45:16 <rm_work> yes :P 20:45:22 <rm_work> the provider work in octavia 20:45:24 <xgerman> yeah, we were talking about the shim/plugin layer 20:45:27 <rm_work> so vendor drivers are possible 20:45:34 <m-greene_> I trying to line up GUI dev from my team starting next quarter 20:45:37 <rm_work> vendors know best what they're looking for 20:45:47 <m-greene_> (octavia GUI, missing L7 wizard, etc.) 20:45:50 <johnsom> Ah, GUI help would be great 20:45:56 <rm_work> yes 20:45:58 <xgerman> +10 20:46:20 <m-greene_> we’re in a tough spot though right now… I only have 4 devs active 20:46:42 <rm_work> yeah, i think basically everyone is feeling the same pain currently :( 20:46:49 <xgerman> :-( 20:47:53 <rm_work> alright, we pretty much done? 20:47:57 <rm_work> about to be boarding I think 20:48:00 <johnsom> m-greene_ I assume they are not working on OpenStack 20:48:16 <johnsom> Yeah, I think we are pretty close to wrapped up for the week. 20:48:27 <xgerman> how long are we skipping now? 20:48:34 <johnsom> Reminder, I think we decided no meetings for the next three weeks 20:48:50 <xgerman> yeah, sounds good — 20:48:54 <rm_work> ok 20:48:56 <johnsom> I will send out an e-mail 20:48:59 <nmagnezi> ack 20:48:59 <rm_work> sounds good to me 20:48:59 <m-greene_> I have 6 technical reports, with 2 temporarily re-allocated. Of my 4, 1-2 are usually handling hot customers cases… :( 20:49:00 <xgerman> k 20:49:20 <johnsom> m-greene_ yeah, understood 20:49:21 <m-greene_> but it’s 4 on OpenStack 20:49:23 <xgerman> customers… it’s always them 20:50:15 <johnsom> m-greene_ I just want to make sure they aren't working on something that they don't have assigned as we have some people coming on that may be picking up some stuff. I just don't want duplicate work 20:51:46 <johnsom> Ok, I think that is it. Safe travels to the Summit folks! 20:51:58 <johnsom> #endmeeting