14:01:32 #startmeeting kuryr 14:01:33 Meeting started Mon Nov 19 14:01:32 2018 UTC and is due to finish in 60 minutes. The chair is dmellado. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:34 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:36 The meeting name has been set to 'kuryr' 14:01:52 o/ 14:01:59 Hi fellow kuryrs, who's around here! 14:02:11 o/ 14:02:48 o/ 14:03:07 #chair dulek 14:03:08 Current chairs: dmellado dulek 14:03:27 o/ 14:03:29 o/ 14:03:43 o/ 14:04:00 so, first of all thanks for making to the Summit for those of you who were around and got to hear dulek and me 14:04:20 3 times saying pretty much the same. :D 14:04:29 exactly 14:04:37 thanks dulek for your help on the sessions and the beers 14:04:39 xD 14:05:03 so, let's first start by summarizing a little bit some stuff that happened there 14:05:19 1) We'll get to change the project to independent again and we'll try to follow the k8s releases 14:05:41 this hasn't still been set in stone and I'll be handling the process with dhellmann and basically over the upstream mailing list (now all merged!) 14:05:46 so please stay tuned out there 14:05:47 dmellado: Oh, you've discussed that with release team? 14:05:59 dulek: in a slightly and rushed way yeah 14:06:08 but we'll discuss over the ML, basically 14:06:21 I do plan to meet wit dhellmann over the week and discuss a few things 14:06:23 dmellado: I guess that at the moment smcginnis is the person to talk about it? 14:06:27 and afterwards I'll start the process 14:06:29 dulek: yeah 14:06:43 basically both, but let's see how we can handle that xD 14:07:53 2) We held a SRIOV discussion with danil and aperevalov in which we wrote some actions on an etherpad at https://etherpad.openstack.org/p/kuryr_sriov 14:08:07 garyloug: you might be interested on it as well ;) 14:08:32 dmellado, yes sure :) 14:09:22 and what else from the summit... interest on having a power 3rd party CI, cisco interest on integrating contiv, for which I'd expect a 3rd party CI as well 14:09:37 and quite a few interesting sessions about having openstack infra create k8s cluster 14:09:49 I'll need to speak with mordred on the topic around 14:09:55 dulek: anything to add around? 14:10:50 Sec… 14:11:50 I've asked folks from Catalyst Cloud about why they're not using Kuryr-Kubernetes in their K8s on OpenStack offering. 14:11:59 The answer was - no network policy support. 14:12:11 well, that's gonna be changing like, now 14:12:18 and I forgot, regarding network policy 14:12:31 Cisco folks did a demo with cillium for which they'd like to add ebpf np support 14:12:43 they said they'd be contributing, so rsblendido! looking forward to it! 14:12:46 dmellado: Not Cisco, SUSE. 14:12:55 s/Cisco/SUSE 14:13:03 was thinking right, typing wrong 14:13:04 xD 14:13:37 all right, so I guess that was somehow a quick summary but feel free to go thru our videos/slides from the summit if you're intersted! 14:13:47 #topic kuryr-kubernetes 14:14:02 Also some guys from Ericsson had a comparison of CNI plugins., 14:14:13 yboaron_: thanks for taking a look at the issue on the freezing ssh exec command 14:14:16 They've gave us some good press, which is nice. :) 14:14:19 #link http://git.openstack.org/cgit/openstack/kuryr-tempest-plugin/commit/?id=a27fe3d35f1fd1ac2b5d18a21575a67487a5c53c 14:14:23 dulek: totally! 14:14:39 I forgot to tell, but overall kuryr was mentioned on a few keynotes, so kudos team! 14:15:13 you're all good professionals besides celebdor, who is a lazy bum whom we all love anyway 14:15:18 xD 14:15:21 dmellado, YW! see https://review.openstack.org/#/c/618635/ 14:15:45 yboaron_: yeah, I was pointing to the same commit ;) 14:15:50 so it should be all fine now 14:15:55 yboaron_: care to give out a summary on this? 14:16:40 dmellado, sure , so, sometimes our tempest tests that call to 'connect_get_namespaced_pod_exec' were stuck 14:17:34 after investigation is seems that from some reason it was stuck on OS select although the operation in pod completed 14:18:21 so, what we did we had a timeout that is passed to select, so in case we hit this issue we"ll leave select due to timeout 14:18:38 well, that's all regarding this issue 14:18:45 yboaron_: I see, thanks for digging into it. Could you use the machine we asked the infra folks about? 14:19:07 yes 14:19:40 awesome, if you're finished with it just let clarkb know 14:20:09 I will keep him posted, I still want to check something there 14:20:15 yboaron_: thx! 14:20:42 all right, regarding NP we are basically finishing the feature and have a few open patches 14:21:04 basically we're finishing unit tests now and we'll start closing patches and making it merged 14:21:06 \o/ 14:21:13 thanks for your help on this while I was away ltomasbo maysams 14:21:36 np, happy to help! 14:22:23 all risht, anyone has anything to add? ;) 14:22:49 dmellado: No problem :-) 14:23:38 garyloug: anything from your side? 14:23:58 dmellado, not much from me - the os-vif was released with out nested-dpdk type included. Przemek on my tema is going to update our kuryr patch this week and also give it a long overdue rebase. 14:24:09 garyloug: ack, thanks! 14:24:17 Other than that we're looking around internally for someone to take the tempest work 14:24:32 awesome, part of the discussion with aperevalov_was about that 14:24:36 so please sync with him on the topic 14:24:40 cool 14:24:51 dmellado: From me side, I started working with NP but got some problems related to quota 14:25:01 s/me/my 14:25:05 maysams: security group rules quota? 14:25:05 garyloug, I'm trying to dive into tempest right now ) 14:25:19 that's pretty normal, we're going to add some changes to the doc 14:25:31 for now you can just deff disable sg ans sgr quota while working on this 14:25:32 dmellado:yes, so I decided to add a readiness check 14:25:48 regarding neutron resources quota 14:25:51 maysams: awesome, what does it check, quotas availability? 14:26:00 yes 14:26:13 maysams: sounds neat! let me know if you're missing any urgent review 14:26:14 for the resources needed in each handler 14:26:20 I'm still going through my review backlog 14:26:26 okay 14:26:29 thank you! 14:26:30 thanks! 14:29:09 so! 14:29:14 I guess that was all 14:29:19 #topic open discussion 14:30:32 Overall, and as I was away for a while, please, core reviewers 14:30:45 try to go over open reviews if you have the time 14:30:53 as the queue got a little bit long 14:30:55 thanks! 14:30:59 #endmeeting