15:04:00 <bnemec> #startmeeting oslo 15:04:01 <openstack> Meeting started Mon Nov 5 15:04:00 2018 UTC and is due to finish in 60 minutes. The chair is bnemec. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:04:03 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:04:06 <openstack> The meeting name has been set to 'oslo' 15:04:08 <bnemec> courtesy ping for amotoki, amrith, ansmith, bnemec, dansmith, dhellmann, dims 15:04:08 <bnemec> courtesy ping for dougwig, e0ne, electrocucaracha, flaper87, garyk, gcb, haypo 15:04:08 <bnemec> courtesy ping for hberaud, jd__, johnsom, jungleboyj, kgiusti, kragniz, lhx_ 15:04:08 <bnemec> courtesy ping for moguimar, njohnston, raildo, redrobot, sileht, sreshetnyak, stephenfin 15:04:10 <bnemec> courtesy ping for stevemar, therve, thinrichs, toabctl, zhiyan, zxy, zzzeek 15:04:15 <bnemec> #link https://wiki.openstack.org/wiki/Meetings/Oslo#Agenda_for_Next_Meeting 15:04:18 <moguimar> o/ 15:04:21 <phuongnh> o/ 15:04:25 * stephenfin lurks 15:04:26 <hberaud> o/ 15:04:30 <ansmith> o/ 15:04:36 <ducnv_> o/ 15:04:37 <kgiusti> o/ 15:05:30 <dhellmann> o/ 15:05:58 <bnemec> I think this is the right time for the meeting. 15:06:01 <bnemec> Stupid time changes. 15:06:27 <bnemec> #topic Red flags for/from liaisons 15:07:00 <bnemec> We've had a number of reports of issues with SSL and rabbit in Pike. 15:07:48 <kgiusti> :( 15:08:04 <jungleboyj> o/ 15:09:00 <bnemec> Obviously that isn't affecting any development though or someone would have reported it before now. 15:09:10 <kgiusti> I've not been able to reproduce the issue reported about ocata->pike locally 15:10:08 <kgiusti> I'm not sure if zigo's problem is the same - need to reach out to him today 15:10:50 <bnemec> kgiusti: Okay, sounds good (or not good ;-). Thanks for looking into this. 15:11:16 <kgiusti> pike's been out for awhile now if this is a hard failure I would've expected to hear about this sooner. 15:11:34 <kgiusti> Something is rotten in the state of Oslo (Messaging) 15:12:28 <bnemec> kgiusti: Yeah, although you would be surprised how long it takes people to adopt new releases. 15:12:39 <kgiusti> true that 15:12:40 <bnemec> Having major bugs reported a year and a half after release isn't as uncommon as you might expect. 15:14:07 <dhellmann> this smells like an openssl (or other library) upgrade problem 15:15:10 <bnemec> Could be. That would also explain why no one has reported it until now. 15:15:13 <kgiusti> dhellmann: good point, I seem to recall a breaking change at some point in the openssl release history... 15:15:23 <zigo> kgiusti: I'm here ! 15:15:36 <kgiusti> Hi zigo! 15:15:47 <kgiusti> Did you have any luck getting your SSL config working? 15:16:00 <zigo> It's working, no problem. 15:16:10 <zigo> Though my trouble was: no report at all that something was going wrong. 15:16:17 <zigo> It took me a real long time to understand. 15:16:33 <zigo> I would have very much prefer not to loose all of that time with something in my log... 15:16:57 <zigo> In my case, there was absolutely nothing in heat-api and heat-engine saying it wouldn't connect to rabbit. 15:16:58 <kgiusti> Yeah I'll work on a fix for that - SSL failures are a pain to debug (even with logs!) 15:17:21 <zigo> I had to double-guess that it wouldn't connect and that it was the root cause. 15:18:11 <kgiusti> zigo: what release are you using? We've had a report of problems with SSL going from ocata->pke 15:18:15 <kgiusti> pike, that is 15:18:28 <zigo> kgiusti: Queens, then more recently, Rocky. 15:18:32 <zigo> Both had the problem. 15:19:07 <kgiusti> zigo: thanks 15:19:15 <zigo> My pleasure. 15:19:52 <kgiusti> I suspect the ocata->pike failures may be different, since the same SSL config worked on ocata failed when upgrading to pike 15:20:11 <kgiusti> I'lll continue to work on it 15:20:41 <bnemec> kgiusti: I think I saw a patch on pike that moved some parameters around too, and I think it may have dropped some deprecated_names in the process. 15:21:02 <bnemec> Maybe people were still relying on the deprecated stuff and it broke when that went away? 15:21:13 <bnemec> (assuming I read the patch correctly, which you shouldn't ;-) 15:21:54 <bnemec> Anyway, we should probably revisit this after the meeting. 15:22:00 <kgiusti> +1 15:22:09 <bnemec> #topic Releases 15:22:21 <bnemec> Completely forgot to do these last week. :-( 15:22:39 * bnemec blames a certain TLA for interrupting his weekly routine 15:23:01 <bnemec> I'll get that done today. 15:23:29 <bnemec> I anticipate no releases next week since it will be Summit. 15:24:39 <bnemec> #topic Action items from last meeting 15:24:51 <dhellmann> bnemec : I feel like we had a couple of things go in late last week that would be good to release 15:25:20 <bnemec> dhellmann: Yeah, I'll propose releases today. 15:25:26 <dhellmann> the oslo.config env stuff from cdent, and something that landed over the weekend as a fix that I can't quite put my finger on 15:25:27 <dhellmann> ok, cool 15:25:57 <bnemec> It just slipped my mind last week and by the time I remembered it was late enough in the week that I figured I'd just wait until this week. 15:26:12 <bnemec> "bnemec send email about project update video" 15:26:27 <bnemec> As it turns out, I didn't have to because we got a project update slot. 15:26:28 <bnemec> \o/ 15:26:40 <moguimar> did the oslo.config env stuff merged? 15:26:53 <bnemec> moguimar: I believe so. 15:27:26 <moguimar> yup 15:27:28 <moguimar> merged 15:28:00 <bnemec> I have some slides made up for the project update and I'll be sending those out today as well for people to review/make additions. 15:28:56 <bnemec> "bnemec to investigate oslomotw" 15:29:16 <bnemec> Still on my todo list, but not done. Probably going to have to wait until after Summit. 15:29:38 <bnemec> "lbragstad to ping jaypipes about oslo.limit API" 15:29:54 <bnemec> I saw some discussion around this, but I don't recall exactly where it stands. 15:32:55 <bnemec> Will have to follow up on that later. 15:32:58 <bnemec> #topic No meeting next week 15:33:17 <bnemec> It's Summit, so I expect a lot of the people attending this will be travelling. 15:33:42 <bnemec> #action bnemec to send meeting cancellation email to list 15:34:01 <bnemec> #topic Oslo project update 15:34:08 <bnemec> #link https://www.openstack.org/summit/berlin-2018/summit-schedule/events/22952/oslo-project-onboarding 15:34:27 <bnemec> It's Thursday at 2:05. 15:35:03 <bnemec> Hope to see everyone there. :-) 15:35:27 <bnemec> #topic Weekly Wayward Review 15:35:34 <bnemec> Let's see what we need to look at this week... 15:36:53 <bnemec> There's a hacking patch for those of you who are hacking cores: https://review.openstack.org/#/c/602464/ 15:37:57 <bnemec> There's also a kafka messaging patch, but that might be too big to get through right now: https://review.openstack.org/#/c/603190/ 15:39:28 <bnemec> If someone wanted to write unit tests for https://review.openstack.org/#/c/612376/ we could get that one in. 15:40:18 <bnemec> Shouldn't be too difficult. There's only one branch in the function. 15:41:57 <bnemec> So the fact that I'm struggling to find a patch we can resolve during the meeting is kind of a good thing. 15:42:12 <bnemec> Almost everything in my top 15 list is related to large feature work that just takes time to review. 15:42:29 <bnemec> Which I take to mean that we're doing a reasonably good job of keeping up with reviews overall. 15:42:38 <moguimar> There is an oslo.policy patch by ozz: https://review.openstack.org/#/c/613313 15:43:03 <bnemec> moguimar: That's approved. :-) 15:43:40 <moguimar> a few minutes ago =D 15:43:59 <bnemec> Only 21 patches out of 116 are waiting on reviewer, so high fives all around. 15:44:02 <bnemec> o/\o 15:44:27 * lbragstad wanders in late 15:44:59 <bnemec> So I guess my conclusion for this week is keep up the good review work. 15:45:13 <bnemec> And we'll keep chipping away at the feature reviews. 15:45:20 <bnemec> #topic Open discussion 15:45:27 <lbragstad> about the oslo.limit bits 15:45:49 <lbragstad> we have discussion on the API in oslo.limit happening in a few different places 15:45:57 <lbragstad> #link https://review.openstack.org/#/c/596520/ 15:46:01 <lbragstad> #link https://review.openstack.org/#/c/615180/5/nova/compute/utils.py 15:46:06 <lbragstad> and 15:46:08 <lbragstad> #link https://review.openstack.org/#/c/600266/6 15:47:49 <bnemec> Ah yes, the first one is where I had seen discussion this week. 15:48:21 <lbragstad> kinda of a lot to keep a pulse on.. but johnthetubaguy is a little concerned about the initial complexity of the API 15:48:28 <lbragstad> specifically with the context manager 15:48:45 <lbragstad> which means it had the opposite effect we were shooting for :) 15:49:09 <bnemec> Yeah, I'm a little surprised to hear that. 15:49:19 <bnemec> The context manager was supposed to make it easier to interact with quota code. 15:49:28 <lbragstad> but the feedback is good 15:49:31 <lbragstad> right 15:49:51 <bnemec> Yeah, as long as there is good discussion happening we're making progress. 15:49:54 <lbragstad> ultimately supposed to hide the verify in the __exit__ of the context manager 15:50:32 <lbragstad> but it sounds like john want's the verification to be optional - so i'm going to be tinkering with his review today to see if i can experience the pain first hand 15:50:52 <lbragstad> hopefully, we'll have something more concrete by the EOW is that's alright 15:51:37 <bnemec> I'm curious what quota is supposed to do if it's optional, but I'll try to go read those reviews and then maybe it will make sense. 15:52:35 <lbragstad> the verification on __exit__ is mainly to protect against races conditions between clients 15:52:35 <bnemec> lbragstad: Thanks for the update. 15:52:40 <lbragstad> no problem 15:53:04 <phuongnh> hi bnemec, dhellmann, as we discussed in Denver, we in Fujitsu would like the oslo config migrator feature have more attention. The time does not support us. Please help us. Thanks. 15:54:01 <bnemec> phuongnh: Yeah, we're getting close. I'm struggling a bit to understand the last patch in the series right now. 15:54:34 <bnemec> I suppose we could merge the rest of the code as-is, but I assume there would be a bug present if we did. 15:54:49 <phuongnh> thanks, ducnv has updated source code and answer your comments. Please tell us if you have more question. 15:55:03 <phuongnh> I know that :) 15:55:26 <bnemec> This is the review I'm talking about, for those playing along at home: https://review.openstack.org/#/c/603060/ 15:56:10 <phuongnh> yes, ducnv has updated on this patch 15:56:14 <bnemec> So I'll try to take another look at that this week. 15:56:27 <phuongnh> thank you a lot 15:56:35 <ducnv_> bnemec, thanks :)) 15:57:13 <bnemec> Okay, we have about three minutes. Anything else? 15:57:29 <phuongnh> and do you need our support in preparing slide for Oslo project update ? 15:57:40 <phuongnh> about oslo config migrator feature? 15:59:28 <bnemec> phuongnh: I have that in my slides already, but if you had information you specifically want included feel free to send it to me. 16:00:02 <phuongnh> general information is ok for us 16:00:09 <phuongnh> that is all for me, thanks 16:00:20 <bnemec> Alright, we're out of time anyway. 16:00:22 <dhellmann> phuongnh : will you or ducnv be able to make it to berlin next week for the summit? 16:00:38 <bnemec> Thanks for joining everyone! 16:00:40 <bnemec> #endmeeting