14:00:23 #startmeeting networking 14:00:23 Meeting started Tue Feb 8 14:00:23 2022 UTC and is due to finish in 60 minutes. The chair is lajoskatona. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:23 The meeting name has been set to 'networking' 14:00:28 o/ 14:00:29 Hi everybody! 14:00:33 HI 14:00:35 hi 14:00:35 Hi! 14:00:39 hi 14:01:11 o/ 14:01:35 o/ 14:02:05 I think we can start 14:02:15 #topic Announcements 14:02:25 The usual Yoga cycle calendar https://releases.openstack.org/yoga/schedule.html 14:02:35 Release countdown, R-7: http://lists.openstack.org/pipermail/openstack-discuss/2022-February/027057.html 14:02:53 Again the freeze dates which are closer 14:03:01 Non-client library freeze (February 17th, 2022) 14:03:11 Client library freeze / Yoga-3 milestone: (February 24th, 2022) 14:03:21 Yoga final release: March 30th, 2022 14:04:15 for the non-client library we are quite good , 14:04:34 we have n-lib, os-ken and ovsdbapp release recently 14:05:05 If you know anything urgent please make sure that we have review focus and perhaps review priority on it 14:06:06 I list again a few features which has hanging patches: 14:06:12 Local IP: https://review.opendev.org/c/openstack/python-openstackclient/+/805618 14:06:49 It has 2 +2s as I remember, so we have to find somebody with wf+1 right in python-openstackclient 14:07:11 Min PPS: https://review.opendev.org/q/topic:bp/qos-minimum-guaranteed-packet-rate+status:open 14:07:47 the neutron and neutron-tempest-plugin patches can be merged I think they are in good shape 14:09:03 the other features which I listed last wee, ndp proxy and smart-nic for OVN as i see need much more things: 14:09:47 Do you have comments or questions for the release dates 14:10:06 or perhaps do you know more features which we should focus more? 14:10:34 none from me 14:10:45 #link https://review.opendev.org/q/topic:packet_rate_limit 14:11:20 Maybe we can pay more attention on neutron new features? 14:11:42 liuyulong: thanks, I forgot that recently, I will add them to my todo list 14:11:47 Nova guys had done a really quickly develop cycle of 14:11:47 bp/qos-minimum-guaranteed-packet-rate 14:12:30 IMO, for those configurable new features, we should not block them for too long time. 14:13:04 Since they does not break exsiting features, if you not enable it. 14:13:10 liuyulong: bp/qos-minimum-guaranteed-packet-rate from neutron perspective is ready, we have few tempest tests and the qos alias patch 14:13:36 #link https://review.opendev.org/q/topic:bp/qos-minimum-guaranteed-packet-rate 14:13:55 as you can see, nova guys had merged more than 75+ patches. 14:15:05 The development speed of neutron is relatively slow. 14:15:43 liuyulong: we can check that, if you feel 14:16:28 sorry for a dumb question: what is the difference between https://bugs.launchpad.net/neutron/+bug/1912460 and https://bugs.launchpad.net/neutron/+bug/1922237 ? 14:16:28 From my personal view of an upstream project, in every develop cycle, if we cloud introduce 14:16:28 some new features, instead of fixing irrelevancies or moving some working-fine exiting code, 14:16:28 the project will get more attention in real world. 14:16:29 what I see that review speed is sometimes not that much, but that really is a function of the active members/developers 14:16:59 More users and more developers can be attracted to our community. 14:17:34 obondarev: minimum guaranteed packet rate QoS is like minimum_badwith but for pps, and as min_bw it has placement/scheduling support 14:18:36 Any way, Neutron has new features this dev cycle, "non-Local IP"? : ) It's good to see that landed. 14:18:42 lajoskatona: that's clear, but two RFE's sound quite similar to me 14:19:30 obondarev: isn't that one provides a maximum limit bit the other provides a guarantee of a minimum? 14:19:34 liuyulong: perhaps that's true, what I know is that as more users use openstack we have to be careful to keep it running, that means heavy maintenance 14:20:17 rubasov: ah, yes, it's about where to limit, thanks :) 14:20:24 obondarev: https://bugs.launchpad.net/neutron/+bug/1912460 but the implementation is different :-) 14:21:08 sorry for noise, I'm a little late to this min pps party I guess 14:21:08 As you can see bug bug/1912460 was set eailer that bug/1922237, but nova guys had done almost all works. 14:21:19 We can discuss this topic on the PTG, to collect different views 14:21:51 how to make the most out of Networking in Openstack 14:22:04 And I left some comments there: https://bugs.launchpad.net/neutron/+bug/1922237/comments/1 14:22:28 liuyulong: I wil lcheck the remaining qos-minimum-guaranteed-packet-rate patches, thanks for the reminder 14:22:45 Limit the min pps is the premise of "guarantee min pps". 14:23:02 Sorry, Limit the pps 14:24:20 I jump back to announcements, if there is no more thoughts for the previous topic 14:24:27 lajoskatona, sure, thanks for the consideration. Please move on. 14:24:39 Next PTG: April 4 - 8, 2022 (virtual) 14:24:49 Please register: https://openinfra-ptg.eventbrite.com/ 14:24:55 it's free :-) 14:25:20 I registered today the team, and started to book slots: #link https://ethercalc.openstack.org/7yxdas7suqnd 14:26:48 the room is Grizzly and now I booked the usual 1300UTC-1600UTC, that seems something perhaps that enjoyable from all timezones 14:26:52 * mlavalle just registered 14:27:15 if you have opinion please tell, we are in time and I am sure I can move slots 14:27:22 mlavalle: +1 14:27:34 thanks for the reminder :-) 14:28:22 another thing is that sphinx jobs are failing on pike and queens 14:28:34 #link http://lists.openstack.org/pipermail/openstack-discuss/2022-February/027106.html 14:29:05 it is again something with py27 and how we build images for zuul. 14:29:39 sigh 14:30:12 at the moment it's not clear how it will evolve, possible that we have to fix those jobs if there is a way, and if we have patches to merge 14:31:32 as I see on pike we have very few patches recently bu queens is different 14:31:52 we are still backporting patches to queens 14:32:31 ralonsoh: yeah, true 14:32:33 I will soon have a patch to port to Queens 14:33:14 lajoskatona, is it possible to have [3]? the revert of the patch that removed the py2 support 14:33:33 in the mail I linked there are several options, like fix it (brute force from project to project), revert the thing that causes it or make the job non-voting/delete it 14:34:12 ralonsoh: I would not put my salary on it, but possible 14:34:18 ok, for now I think we can skip that job 14:34:20 ralonsoh: I can ask around on infra 14:34:25 is "just" the docs 14:34:55 yes this is my "less worse" idea from the list 14:35:22 I'll send a patch 14:35:24 ok, I ask around and we can check and act based on that, but skipping doc build sound like a way forward 14:36:09 and we can add back the job if things are better 14:37:00 #topic Bugs 14:37:08 Report from lajoskatona: http://lists.openstack.org/pipermail/openstack-discuss/2022-February/027074.html 14:37:32 I have a few without assignee 14:37:41 Neutron-dynamic-routing does not work with OVN: #link https://bugs.launchpad.net/neutron/+bug/1959666 14:38:11 We discussed this or similar a few times in this cycle 14:38:14 yeah, I'll ping OVN folks, at least to provide feedback on this one 14:38:35 ralonsoh: thanks 14:38:51 [stein][neutron] Static rules losed: #link https://bugs.launchpad.net/neutron/+bug/1959884 14:39:34 I try to allocate some time to try to reproduce it on master, but if anybody has ideas or time to check if it is on master that would be helpful 14:39:56 I tried it but with newer versions 14:39:56 Disallow users to allocate gateway ip of external subnets as floating ip: #link https://bugs.launchpad.net/neutron/+bug/1959699 14:40:29 ralonsh: really? sounds good 14:40:45 I don't think https://bugs.launchpad.net/neutron/+bug/1959699 is a bug 14:41:05 of course we can limit that but I don't know why 14:41:13 ralonsoh: for bug/1959699 I have the same feeling 14:41:17 the GW IP is not a Neutron resource 14:41:24 (not a port or an assigned IP) 14:41:47 in any case, I'll reply again there 14:42:47 ralonsoh: thanks 14:43:27 +1 Additionally it may break things for those using this ability on purpose 14:43:58 obondarev: or we have to add extra cfg to keep old behaviour 14:44:22 This week bcafarel is the deputy, and next week slaweq will be. 14:44:30 lajoskatona: right 14:45:08 bcafarel: is that ok for you? your colleagues said last week it is ok for you :-) 14:45:22 lajoskatona: they were right, I am on it :) 14:45:32 bcafarel: ok, thanks 14:46:16 Just one more thing from me, the PTL and TC nomination starts: #link http://lists.openstack.org/pipermail/openstack-discuss/2022-February/027070.html 14:46:43 are you proposing yourself again, lajoskatona ? 14:46:43 so if you would like to nominate yourself to the TC or to the PTL position, please go for ti 14:47:05 ralonsoh: yes, I will, I checked it with my employer 14:47:14 nice to hear that 14:47:15 good! 14:47:16 I go for another round 14:47:32 nice that you say that :-) 14:48:24 I would like to forward the warning that the CI meeting will start 30 minutes later 14:48:32 Excellent news! Thankd for your service to the community 14:48:48 mlavalle: thanks :-) 14:49:11 And nothing more from me, is there anything more to discuss? 14:49:46 nothing from me 14:50:46 lajoskatona: cool! :) 14:51:20 than lets close the meeting 14:51:34 #endmeeting