15:10:44 <bnemec> #startmeeting oslo 15:10:44 <bnemec> Courtesy ping for bnemec, jungleboyj, moguimar, hberaud, stephenfin, kgiusti, johnsom, e0ne, redrobot, bcafarel 15:10:44 <bnemec> #link https://wiki.openstack.org/wiki/Meetings/Oslo#Agenda_for_Next_Meeting 15:10:45 <openstack> Meeting started Mon Feb 24 15:10:44 2020 UTC and is due to finish in 60 minutes. The chair is bnemec. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:10:46 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:10:48 <openstack> The meeting name has been set to 'oslo' 15:10:52 <hberaud> o/ 15:10:58 <bcafarel> o/ 15:11:04 <jungleboyj> o/ 15:11:04 <johnsom> o/ 15:11:42 <moguimar> o/ (partialy) 15:11:45 <stephenfin> o/ 15:12:37 <ansmith> o/ 15:13:07 <kgiusti> kgiusti: o/ (must leave early for call) 15:13:17 <bnemec> #topic Red flags for/from liaisons 15:13:31 <johnsom> Nothing from Octavia 15:13:40 <jungleboyj> Nothing from Cinder. 15:13:57 <bnemec> The thing we were discussing a bunch last week was the oslo.limit 1.0 release. It's not actually a 1.0 release. :-) 15:14:16 <bcafarel> nothing from neutron (I think Rodolfo already asked last week about a question he had) 15:14:27 <bnemec> So unless you want to get involved with designing the oslo.limit API (which would be great!), don't use it. 15:14:33 <moguimar> nothing from Barbican 15:15:52 <bnemec> Should maybe just merge this with the releases topic since most of the red flags from our side are around new releases. ;-) 15:16:02 <bnemec> #topic Releases 15:17:05 <bnemec> As noted above, assume oslo.limit is actually pre-1.0 despite the version number until we say otherwise. 15:17:44 <bnemec> FWIW, it seems the consensus is to not try to "fix" this by unpublishing the release. 15:19:06 <bnemec> I probably need to take a look at what needs to be released again this week. I'd been holding off because of all the py2-removal releases, but it seems like those have slowed down and there may be stuff that needs to be released. 15:19:21 <bnemec> Will try to get that done today. 15:19:53 <bnemec> #topic Action items from last meeting 15:20:00 <bnemec> "bnemec to send mail about oslo.limit 1.0" 15:20:00 <bnemec> Done 15:20:06 <bnemec> "hberaud to write release note for oslo.limit" 15:20:14 <hberaud> done 15:20:32 <bnemec> Sorry I haven't reviewed it. :-/ 15:20:40 <bnemec> "hberaud to investigate whether to fix or remove oslobot from oslo.tools" 15:20:40 <hberaud> https://review.opendev.org/708326 15:20:45 <hberaud> not done 15:23:42 <bnemec> I approved the release note. 15:23:51 <bnemec> And that was all the action items. 15:23:57 <bnemec> #topic Vancouver PTG 15:24:29 <bnemec> Still haven't heard from anyone who is going to be there, so at this point I don't expect to ask for a room. 15:24:54 <bnemec> We can maybe do a virtual PTG if the next PTL wants to organize one. 15:25:14 <bnemec> But I'll leave that up to that person. :-) 15:25:37 <jungleboyj> Cinder team is planning to be there. 15:25:38 <bnemec> We already covered oslo.limit so we can skip that one. 15:25:40 <jungleboyj> Just FYI. 15:26:10 <johnsom> I am planning to be there, but will be focused in the Octavia room. 15:26:16 <bnemec> Cool, maybe the Oslo refugees can hang out there. :-) 15:26:49 <bnemec> #topic Meeting time poll 15:27:27 <bnemec> #link https://doodle.com/poll/zmyhrhewtes6x9ty 15:28:05 <bnemec> So there are only two times that work for moguimar. 15:28:26 <moguimar> yeah, but checking on my callendar the meeting will stop to overlap 15:28:31 <moguimar> in about two weeks 15:28:34 <bnemec> One of them doesn't work for me at all because I have a 1-1 at that time that I can't really multitask. 15:28:52 <bnemec> Oh, is this only a daylight savings time conflict? 15:29:00 <openstackgerrit> Nick Maludy proposed openstack/tooz master: Adds heartbeating to the consul driver. https://review.opendev.org/707681 15:29:28 <moguimar> yeah 15:29:45 <moguimar> it is an issue during nov-feb 15:30:01 <bnemec> It occurs to me that I hadn't even considered the DST change when I filled out the survey... 15:30:03 <moguimar> but my meeting was also migrated to monday during nov 15:30:09 <moguimar> so I didn't catch that until now 15:30:51 <bnemec> Yeah 15:30:56 * bnemec shakes fist at DST 15:30:59 <moguimar> so I guess we can keep this time 15:31:12 <moguimar> and my other meeting might move around before next November 15:31:35 <bnemec> Yeah, and actually the next hour on Monday is the one with the highest number of people available. 15:32:15 <bnemec> So if DST will take care of moving it to then in a few weeks anyway, that seems like the way to go. 15:32:50 <bnemec> That should also reduce the meeting overrun problems I've had lately as my previous meeting shouldn't move. 15:33:36 <openstackgerrit> Merged openstack/oslo.limit master: oslo.limit mistakenly released as 1.0.0 (release note) https://review.opendev.org/708326 15:33:56 <bnemec> Okay, if no one has any objections then let's let DST solve the conflict for us and leave the meeting where it is UTC-wise. 15:34:05 <moguimar> +1 15:34:13 <bnemec> I don't expect any objections since almost everyone was green on that time. :-) 15:34:24 <bnemec> I'll send a followup email to the list with this result. 15:34:33 <bnemec> #action bnemec send email about Oslo meeting time 15:34:41 <bnemec> Thanks to everyone who filled out the survey! 15:35:32 <bnemec> That's the one-off topics for this week. 15:35:34 <bnemec> #topic Weekly Wayward Review 15:36:30 <bnemec> #link https://review.opendev.org/#/c/699694 15:36:35 <bnemec> I just approved it. :-) 15:36:59 <bnemec> I could swear I had looked at that before, but maybe it was before the py2 drop had happened or something. 15:38:12 <bnemec> If anyone else is looking for something to review, it would be nice to make some progress on https://review.opendev.org/#/c/691820/ 15:38:41 <bnemec> This is part of Line's changes to reduce messaging bottlenecks in their deployment. 15:38:54 <bnemec> Would be nice if we could get it available upstream. 15:39:48 <bnemec> Anyway, that's it for the regular agenda. 15:39:49 <bnemec> #topic Open discussion 15:39:53 <bnemec> Anything else this week? 15:40:00 <moguimar> yep 15:40:02 <moguimar> o/ 15:40:10 <moguimar> who are our oslo.cache experts? 15:40:22 <moguimar> I'm trying to push TLS support for memcached 15:40:35 <bnemec> With the departure of kmalloc I think that would be hberaud. 15:40:40 <hberaud> bnemec: (or other core) can you validate this one => https://review.opendev.org/#/c/704508/ (backport) 15:40:48 <bnemec> Maybe someone else from the Keystone team too. 15:40:51 <moguimar> got that merged into python-binary-memcached so far 15:41:34 <hberaud> moguimar: I seen your dogpile.cache discussion 15:42:18 <moguimar> I was reading oslo.cache docs last week 15:42:25 <moguimar> and found out that mongodb has TLS support 15:42:28 <moguimar> gotta try it 15:42:40 <moguimar> support already in oslo.config 15:42:45 <moguimar> oslo.cache* 15:43:15 <moguimar> but I also couldn't find many oslo.cache examples 15:43:35 <moguimar> there is only one in the docs and not that straightforward 15:43:55 <moguimar> not just copy-paste and watch it burn 15:44:59 <hberaud> moguimar: what do you thing to start by initialize some etherpad and start centralize things on it and collect our needs etc...? 15:45:00 <moguimar> I'll be synching up with Lance Bragstad later today on this topic 15:45:20 <moguimar> hberaud, we have one already right? 15:45:27 <moguimar> let me find the link 15:45:39 <hberaud> moguimar: ah right I remember now 15:46:15 <moguimar> #link https://etherpad.openstack.org/p/oslo-cache-tls-support-worksheet 15:47:34 <bnemec> A mailing list thread might be good too. I wouldn't be surprised if there are other people interested in this. 15:47:38 <openstackgerrit> Merged openstack/oslotest master: tools: Default to Python 3 https://review.opendev.org/699694 15:48:04 <moguimar> good idea bnemec 15:48:25 <moguimar> I'll prepare something 15:48:46 <bnemec> Which one of the options from the etherpad are you planning to pursue? 15:49:05 <moguimar> python-binary-memcached 15:49:07 <bnemec> Last I heard on the pymemcache change was that we needed to verify it didn't break in the migration from python-memcache. 15:49:13 <moguimar> already got TLS support on it 15:49:20 <moguimar> now moving up to dogpile.cache 15:49:23 <moguimar> and then oslo.cache 15:49:50 <bnemec> Okay, so that's a third memcache client lib? 15:49:52 <moguimar> python-binary-memcached also supports SASL, so we picked it for being "feature-rich" 15:50:07 <bnemec> Is it in global-requirements? 15:51:50 <moguimar> can't find it here 15:51:51 <moguimar> https://opendev.org/openstack/requirements/src/branch/master/global-requirements.txt 15:51:55 <moguimar> am I looking in the right place? 15:52:11 <bnemec> And I guess this assumes we'll move all of oslo.cache to use pbm? I wouldn't think we'd want to use different libs for SSL and non-SSL. 15:52:27 <bnemec> Yeah, that's it. 15:52:33 <bnemec> It was kind of a leading question because I suspected it wouldn't be there. ;-) 15:53:21 <bnemec> If we're not still planning to pursue pymemcache we might want to remove that. 15:53:37 <bnemec> The requirements team doesn't generally like having a bunch of libs that do the same thing. 15:54:37 <moguimar> there are 4 different python clients to memcached 15:54:42 <moguimar> that we have documented 15:55:23 <smcginnis> If we can narrow it down to a couple officially tested in the community, that is preferred. 15:55:40 <moguimar> bnemec, what is pbm/ 15:55:42 <moguimar> ? 15:55:59 <bnemec> python-binary-memcached 15:56:03 <moguimar> ah 15:57:13 <moguimar> this is mainly driven by the fact that TLS-Everywhere forgot memcached 15:58:38 <bnemec> Oh, maybe pymemcache isn't only used by oslo.cache. 15:58:49 <bnemec> Apparently it broke the tooz gate at one point: https://opendev.org/openstack/requirements/commit/0297b9936788ac100db6787d048c6d378da377b0 15:59:29 <bnemec> At least if we migrate off python-memcached maybe we can remove that. 15:59:51 <moguimar> both python-memcached and pylibmc rely on libmemcached 16:00:08 <moguimar> a client library to memcached written in C 16:00:18 <moguimar> and it is unmaintained for now 16:00:29 <moguimar> according to folks at #memcached 16:00:37 <bnemec> Yeah, moving away from that is definitely a win. 16:01:15 <moguimar> ok, we are past the top of the hour 16:01:28 <bnemec> If we can do a 1 for 1 swap with memcached libraries then that should be okay with the requirements team. 16:01:34 <moguimar> I'll send the email tomorrow, I'm running out of time today 16:01:51 <bnemec> Sounds good. 16:02:16 <bnemec> Speaking of running out of time, we're past the end of the meeting now. 16:02:27 <bnemec> Anything else we need to talk about quickly? 16:02:40 <bnemec> #action moguimar send email about moving oslo.cache to a different memcached library 16:04:35 <bnemec> Okay, thanks for joining everyone! 16:04:37 <bnemec> #endmeeting