14:00:09 <lajoskatona> #startmeeting networking
14:00:09 <opendevmeet> Meeting started Tue Jul  5 14:00:09 2022 UTC and is due to finish in 60 minutes.  The chair is lajoskatona. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:09 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:09 <opendevmeet> The meeting name has been set to 'networking'
14:00:12 <lajoskatona> Hi
14:00:12 <mlavalle> o/
14:00:25 <obondarev> hi
14:00:42 <rubasov> o/
14:00:56 <elvira> hi!
14:01:07 <slaweq> o/
14:01:11 <frickler> \o
14:01:24 <ralonsoh> hi
14:01:59 <lajoskatona> Ok let's start today's meeting
14:02:01 <lajoskatona> #topic Announcements
14:02:11 <lajoskatona> The usual Zed schedule: https://releases.openstack.org/zed/schedule.html
14:02:25 * slaweq is in the same time in the secure RBAC meeting
14:02:35 <lajoskatona> slaweq: ack, thanks
14:03:02 <lajoskatona> The EOLing of Queens, Rocky and Stein branches for stadiums happened
14:03:36 <lajoskatona> there are still some hanging patches (topic: #link https://review.opendev.org/q/topic:stadium_eol_qrs )
14:03:48 <lajoskatona> but that is for deleting some old legacy zuul jobs
14:04:54 <bcafarel> late o/
14:05:04 <lajoskatona> There are some interesting discussion on mailing list, like:
14:05:12 <lajoskatona> [all][tc][gerrit] Ownership of *-stable-maint groups
14:05:19 <lajoskatona> #link http://lists.openstack.org/pipermail/openstack-discuss/2022-July/029367.html
14:05:37 <slaweq> yeah, I saw it but didn't had time to read all that thread yet
14:06:27 <lajoskatona> If I understand well we wil have management right over the stable neutron groups
14:06:45 <lajoskatona> another topic this morning perhaps is this one:
14:06:59 <lajoskatona> A centralized list of usable vs unusable projects? (#link http://lists.openstack.org/pipermail/openstack-discuss/2022-July/029385.html )
14:07:36 <lajoskatona> even thinking on it is interesting as most/some of our stadium projects can be interesting to check from usability
14:08:23 <lajoskatona> they mostly perfect for the few usecase the original developers worked on, but no new adaptation happened recently, like to make them work with OVN
14:09:29 <lajoskatona> Do you have questions or comments for Announcements?
14:09:58 <lajoskatona> Ok, next topic
14:10:02 <lajoskatona> #topic Bugs
14:10:13 <lajoskatona> Report from rubasov: http://lists.openstack.org/pipermail/openstack-discuss/2022-June/029270.html
14:10:39 <lajoskatona> There were few "under discussion / more investigation needed":
14:10:46 <lajoskatona> neutron dnsmasq keeps pushing the default gateway (#link https://bugs.launchpad.net/neutron/+bug/1979528 )
14:11:40 <lajoskatona> And: l3ha router delete race condition if state changes to master (#link https://bugs.launchpad.net/neutron/+bug/1979089 )
14:12:35 <lajoskatona> rubasov: do you have any comments, or something to highlight?
14:13:05 <rubasov> I missed the newest answer on 1979528, but now I will follow up on that
14:13:23 <lajoskatona> rubasov: thanks
14:13:28 <rubasov> on the other hand I don't know enough about l3ha for the other one
14:14:43 <lajoskatona> ok, next week's report from ralonsoh: http://lists.openstack.org/pipermail/openstack-discuss/2022-July/029380.html
14:15:24 <lajoskatona> ralonsoh: do you have any comments/highlights for it? As I remember all bugs has assignee
14:15:32 <ralonsoh> I think all bugs are assigned or there is a patch under review
14:15:36 <ralonsoh> right
14:16:32 <ralonsoh> so nothing to comment, thanks
14:16:37 <lajoskatona> ralonsoh, rubasov: thanks for the report
14:16:55 <lajoskatona> This week lucasgomes is the deputy and next week jlibosva will be.
14:17:16 <lajoskatona> I am not sure if lucasgomes is aware of his deputy week, I have no answer to my mail :-(
14:17:44 <mlavalle> I can ping him internally at RH
14:17:52 <lajoskatona> mlavalle: thanks
14:18:28 <jlibosva> he was sick at the beginning of the last week so perhaps he might have missed that. I ack my duty for the next week
14:18:46 <lajoskatona> jlibosva: thanks
14:19:16 <lajoskatona> if there is no comments or questions for the bugs, we can move on
14:19:56 <lajoskatona> #topic On Demand Agenda
14:20:08 <lajoskatona> (lajoskatona/elvira): channel in questions for development topics (how to do things, how to debug things.... etc.) to IRC
14:20:23 <mlavalle> lajoskatona: lucasgomes confirms he is aware
14:20:32 <lajoskatona> mlavalle: thanks
14:20:45 <lajoskatona> this topic is quite formless, but please think about it.
14:21:28 <lajoskatona> It is about to give more life to the channel, and make it a welcoming channel for dumb developer questions, or newcomer topics also
14:22:11 <lajoskatona> perhaps it is now, but we have few newcomers, but would be good to know if we can make it more easy to ask
14:22:57 <lajoskatona> I was thinking on a period of like office hours for some projects, but as "help you hacking hour" or something like this
14:23:30 <ralonsoh> nice idea
14:23:45 <lajoskatona> it was elvira who mentioned that if a channel is quiet, it's not easy to go there and ask a question becasue it feels like bothering others
14:23:50 <slaweq> lajoskatona maybe I'm not aware but were there any attempts to contact with such questions already?
14:23:54 <slaweq> and we didn't reply?
14:24:25 <slaweq> my concern is that how such newcomer will know about such "office hours" in the channel?
14:24:26 <lajoskatona> slaweq: It is more about the questions that are not even asked
14:24:37 <elvira> slaweq: I think is more on the side of: if no one talks, no one asks. Not really that the channel is not welcoming
14:24:39 <lajoskatona> slaweq: yeah that is a good question
14:25:30 <slaweq> First of all we have this doc https://github.com/openstack/neutron/blob/master/doc/source/contributor/contributing.rst
14:25:37 <slaweq> maybe we should make it somehow more visible?
14:25:43 <lajoskatona> yea the not welcoming was not a good starting from me, I think there's always answer if there is a question, but to make it easier to ask the first question
14:25:46 <slaweq> like link to it on the main page of the repo?
14:26:23 <lajoskatona> slaweq: thats a good idea
14:26:33 <slaweq> now it's linked in https://github.com/openstack/neutron/blob/master/CONTRIBUTING.rst
14:26:46 <slaweq> but not in the https://github.com/openstack/neutron/blob/master/README.rst
14:27:07 <slaweq> lajoskatona  elvira I'm not against such "office hours" and I'm happy to help with it
14:27:27 <slaweq> but we should first somehow advertize it
14:27:51 <slaweq> maybe we can also add such info to the channel's topic
14:28:05 <lajoskatona> slaweq: I don't want to have extra load if it is useless, just first to think about it :-)
14:28:30 <slaweq> sure
14:28:34 <lajoskatona> slaweq: channel topic is also good idea, if it is working with oftc
14:29:15 <frickler> I can set channel topic if you tell me what you want there
14:29:19 <elvira> Maybe we can come up with more general approaches. Like adding a message in the topic of the IRC where we encourage asking
14:29:20 <frickler> or any other infra-root
14:29:32 <lajoskatona> ok, thanks for the ideas, I will think about it also in the coming weeks
14:29:40 <lajoskatona> frickler: thanks
14:30:05 <slaweq> elvira++ frickler++ good ideas
14:30:33 <frickler> would also be good to link to the channel log IMHO
14:30:52 <lajoskatona> Next topic:
14:31:01 <lajoskatona> Sorry, wrong copy/paste
14:31:29 <mlavalle> regardless of the mechanism, making the Neutron channel and the team more approachable is always good
14:31:41 <lajoskatona> frickler: you mean to https://meetings.opendev.org/irclogs/%23openstack-neutron/ ?
14:31:50 <frickler> yes
14:32:16 <fungi> also the neutron team can delegate additional channel operators by amending the accessbot config in openstack/project-config
14:32:29 <lajoskatona> by the way is there a way to search irclogs? I missed that feature sometimes
14:32:58 <fungi> lajoskatona: site:meetings.opendev.org in a search engine maybe. but no we don't have a standalone index of the channel log files
14:33:16 <lajoskatona> fungi: thanks, I will check it
14:34:29 <lajoskatona> Ok, next topic:
14:34:37 <lajoskatona> (rubasov): Neutron performance measurements with rally back to Big Bang :-)
14:34:43 <lajoskatona> http://lists.openstack.org/pipermail/openstack-discuss/2022-July/029384.html
14:34:58 <rubasov> I just have these api performance measurements for the last few releases, maybe you find it interesting
14:35:09 <rubasov> direct link to the summary chart: http://lists.openstack.org/pipermail/openstack-discuss/attachments/20220704/6838fd60/attachment-0001.png
14:36:20 <lajoskatona> Note: you have to scroll to the bottom to see the x axis labels
14:37:03 <rubasov> yeah the x axis is shared between the plots
14:37:36 <slaweq> rubasov++ thx for that, it's really good
14:38:18 <lajoskatona> exactly: thanks for the efforts, and even the news is good :-)
14:38:20 <mlavalle> rubasov: +++
14:38:27 <slaweq> since like between Queens and Victoria most of the things were really slow, and it improved in Wallaby and newer versions
14:38:39 <rubasov> still thinking about re-uploading the tarball somwehere more stable
14:39:13 <rubasov> will check if it fits github policies
14:40:09 <rubasov> and there's more data in the rally reports, may make sense to dive into that too
14:41:51 <lajoskatona> Perhaps it's worth to make a kind of discussion if this method is good enough for such measurements, and do something like this for every coming cycle, and add a "team goal" to next cycle based on that
14:42:31 <rubasov> if you have any questions, comments on the methodology, you know where to find me :-)
14:42:47 <lajoskatona> thanks rubasov
14:43:09 <slaweq> yes, we should IMO identify now slowest operations and maybe take a closer look at such
14:43:23 <slaweq> like e.g. create_and_update_routers - it is better now but still 150 seconds
14:43:42 <slaweq> I'm not sure for how many routers it was at once but seems still slow
14:44:06 <rubasov> keep in mind that is for 40 iterations
14:44:23 <slaweq> ahh, ok
14:45:07 <rubasov> the rally task definition defines these iteration counts
14:45:40 <rubasov> the chart has the iteration count next to the scenario name
14:46:38 <slaweq> rubasov: I saw it but I wasn't sure if time is summary for all iterations or maybe average per iteration
14:46:41 <slaweq> or something like that
14:46:47 <slaweq> thx for explanation
14:47:09 <rubasov> rally has two main outputs: load_duration and full_duration for each scenario
14:47:23 <rubasov> the chart show load_duration for all iterations
14:49:13 <lajoskatona> If there's no more comments for these measurements, we can move on
14:49:31 <lajoskatona> (slaweq) August 2022 OpenInfra Board Sync
14:49:42 <lajoskatona> https://lists.openstack.org/pipermail/openstack-discuss/2022-June/029352.html
14:49:42 <lajoskatona> Etherpad https://etherpad.opendev.org/p/2022-08-board-openstack-sync
14:49:42 <lajoskatona> Time slots poll https://framadate.org/atdFRM8YeUtauSgC (deadline next Friday, 15.07.2022)
14:49:55 <lajoskatona> slaweq: thanks for bringing it up
14:50:43 <slaweq> yeah, it's just a heads up
14:50:53 <lajoskatona> thanks
14:50:53 <ralonsoh> who needs to attend?
14:51:04 <slaweq> if You have any topics which You think would be good to discuss with the Board, please add to the etherpad
14:51:19 <slaweq> and add Your prefered time slots in the poll
14:51:26 <lajoskatona> I will I think (have to check the time slots)
14:51:41 <slaweq> nothing more regarding this topic from me
14:51:58 <lajoskatona> ok, next one:
14:52:05 <ralonsoh> sorry
14:52:05 <lajoskatona> (slaweq) Bare rechecks
14:52:10 <lajoskatona> Details in https://lists.openstack.org/pipermail/openstack-discuss/2022-June/029342.html
14:52:11 <ralonsoh> who needs to attend
14:52:12 <ralonsoh> ?
14:52:36 <slaweq> ralonsoh  You are asking about this board sync?
14:52:42 <ralonsoh> yes
14:52:51 <slaweq> only those who wants
14:52:56 <lajoskatona> ralonsoh: I think the PTL should be there, so I will go, if you have topics that I can bring, let me know
14:52:56 <ralonsoh> ok, thanks
14:52:59 <slaweq> there is no any requirements
14:53:21 <slaweq> but everyone is welcome :)
14:53:42 <slaweq> regarding "bare rechecks"
14:54:11 <slaweq> I just wanted to bring it here as in TC recently we started looking how many rechecks done by different projects are rechecks without any reason given in comment
14:54:41 <slaweq> such rechecks may be without reason because person who did it, didn't check what was the issue really
14:55:24 <slaweq> so please try to at least look into the failed jobs' results and try to maybe find what the issue was there, or report new bug
14:55:32 <slaweq> or find existing bug if there is any
14:55:51 <slaweq> You can always ask on the channel for help with any specific failure
14:56:06 <slaweq> and then write comments like "recheck and reason of the recheck"
14:56:35 <slaweq> as a reason it may be link to the bug, bug number, short explanation why job failed or anything else really
14:56:47 <slaweq> do You have any questions regarding that?
14:56:58 <lajoskatona> slaweq: thanks, I was thinking how to advertise it
14:57:17 <slaweq> lajoskatona I will be also sending emails about it weekly to the ML
14:57:24 <slaweq> so I hope it will be more advertized
14:57:59 <slaweq> and also if You will see someone who is doing "bare rechecks" - please try maybe educate in the gerrit comments, on irc, etc.
14:58:00 <lajoskatona> Isn't there a way to set custom banner in gerrit with such blinking message?
14:58:02 <slaweq> I think that will be the best what we can do
14:58:12 <slaweq> lajoskatona I don't think so
14:58:22 <slaweq> it was discussed in one of the tc meetings already
14:58:36 <lajoskatona> slaweq: ok, thanks
14:59:42 <slaweq> lajoskatona  yw
14:59:46 <lajoskatona> We are close to the end of our hour, so let's close the meeting for today
14:59:48 <slaweq> that's all from me
14:59:58 <lajoskatona> slaweq: ack, thanks
15:00:02 <lajoskatona> #endmeeting