| opendevreview | Lajos Katona proposed openstack/networking-bagpipe master: SQLAlchemy: add context wrappers to SFC driver https://review.opendev.org/c/openstack/networking-bagpipe/+/891325 | 07:39 |
|---|---|---|
| opendevreview | Adam proposed openstack/neutron stable/zed: For hosts in DVR mode, only fetch bound FIPs https://review.opendev.org/c/openstack/neutron/+/891397 | 10:22 |
| opendevreview | Adam proposed openstack/neutron stable/2023.1: For hosts in DVR mode, only fetch bound FIPs https://review.opendev.org/c/openstack/neutron/+/891398 | 10:28 |
| opendevreview | Adam Oswick proposed openstack/neutron stable/2023.1: For hosts in DVR mode, only fetch bound FIPs https://review.opendev.org/c/openstack/neutron/+/891398 | 11:13 |
| opendevreview | Adam Oswick proposed openstack/neutron stable/zed: For hosts in DVR mode, only fetch bound FIPs https://review.opendev.org/c/openstack/neutron/+/891397 | 11:14 |
| opendevreview | Adam Oswick proposed openstack/neutron stable/zed: For hosts in DVR mode, only fetch bound FIPs https://review.opendev.org/c/openstack/neutron/+/891397 | 11:38 |
| opendevreview | Adam Oswick proposed openstack/neutron stable/2023.1: For hosts in DVR mode, only fetch bound FIPs https://review.opendev.org/c/openstack/neutron/+/891398 | 11:39 |
| opendevreview | Adam Oswick proposed openstack/neutron stable/2023.1: For hosts in DVR mode, only fetch bound FIPs https://review.opendev.org/c/openstack/neutron/+/891398 | 12:28 |
| opendevreview | Adam Oswick proposed openstack/neutron stable/zed: For hosts in DVR mode, only fetch bound FIPs https://review.opendev.org/c/openstack/neutron/+/891397 | 12:29 |
| *** obondarev_ is now known as obondarev | 13:48 | |
| ralonsoh | Ping list: bcafarel, elvira, frickler, mlavalle, mtomaska, obondarev, slawek, tobias-urdin, ykarel, lajoskatona, jlibosva, averdagu, amotoki | 14:00 |
| lajoskatona | o/ | 14:00 |
| mtomaska | o/ | 14:00 |
| mlavalle | o/ | 14:00 |
| ralonsoh | #startmeeting networking | 14:00 |
| opendevmeet | Meeting started Tue Aug 15 14:00:58 2023 UTC and is due to finish in 60 minutes. The chair is ralonsoh. Information about MeetBot at http://wiki.debian.org/MeetBot. | 14:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 14:00 |
| opendevmeet | The meeting name has been set to 'networking' | 14:00 |
| ralonsoh | hello all | 14:01 |
| isabek | o/ | 14:01 |
| obondarev | o/ | 14:01 |
| haleyb | o/ | 14:01 |
| ralonsoh | ok, lets start | 14:01 |
| ralonsoh | #topic announcements | 14:01 |
| ralonsoh | #link https://releases.openstack.org/bobcat/schedule.html | 14:01 |
| ralonsoh | we have a 3 interesting weeks from now | 14:02 |
| ralonsoh | next week is the non client lib freeze | 14:02 |
| ralonsoh | and the next one is bobcat-3 milestone | 14:02 |
| ralonsoh | we'll start receiving the releases patches from the releases team soon | 14:02 |
| ralonsoh | in any case, we'll need to review them | 14:03 |
| ralonsoh | next week I'll add all links we have related | 14:03 |
| ralonsoh | and I'll also propose new stable releases this week for A, Z and Y, if you agree on this | 14:03 |
| rubasov | late o/ | 14:03 |
| lajoskatona | +1 | 14:03 |
| ralonsoh | next weel I'll start collecting the highlights of this release | 14:04 |
| ralonsoh | and I'll push a patch for you to review, comment and add anything you consider | 14:04 |
| ralonsoh | I'll ping you once the patch is uploaded | 14:04 |
| ralonsoh | and, of course, the election nomination begins next week, so if you are considering proposing yourself as PTL, please be aware of the schedule | 14:05 |
| ralonsoh | Am I missing something? | 14:05 |
| ralonsoh | ahhh yes | 14:06 |
| ralonsoh | we have a new episode! | 14:06 |
| ralonsoh | please check https://openinfra.dev/live/#all-episodes | 14:06 |
| lajoskatona | +1, this one was really interesting (Exploring the Intersection of AI and Open Source ) | 14:06 |
| ralonsoh | yeah (I hope in 5 years I can keep my salary and my personal AI works for me...) | 14:07 |
| ralonsoh | ok, let's jump to the next topic | 14:07 |
| ralonsoh | #topic bugs | 14:07 |
| ralonsoh | the report | 14:07 |
| ralonsoh | #link https://lists.openstack.org/pipermail/openstack-discuss/2023-August/034739.html | 14:08 |
| ralonsoh | (you boys don't rest on August??) | 14:08 |
| ralonsoh | this is a quite extend list this week | 14:08 |
| ralonsoh | so let's check the unassigned bugs we have there | 14:08 |
| ralonsoh | #link https://bugs.launchpad.net/neutron/+bug/2031087 | 14:08 |
| ralonsoh | I didn't try/triagged this one | 14:09 |
| ralonsoh | but I'm not sure if link local interfaces NA should be forwarded | 14:09 |
| ralonsoh | I should check that with OVN core folk | 14:09 |
| ralonsoh | does anyone want to check it? | 14:10 |
| ralonsoh | haleyb, you have been working on it | 14:10 |
| opendevreview | Merged openstack/neutron master: [OVN] ovn-db-sync check for router port differences https://review.opendev.org/c/openstack/neutron/+/890799 | 14:10 |
| haleyb | yes i can take it as i already tried to reproduce it | 14:10 |
| ralonsoh | haleyb, thanks a lot, I'll assign it to you | 14:10 |
| haleyb | already did :) | 14:11 |
| ralonsoh | thanks | 14:11 |
| ralonsoh | next one | 14:11 |
| ralonsoh | #link https://bugs.launchpad.net/neutron/+bug/2030747 | 14:11 |
| ralonsoh | I tried to replicate it but with no luck | 14:11 |
| ralonsoh | but I think I'm mixing admin and non-admin users | 14:11 |
| ralonsoh | so if you don't mind, I'll assign it to myself and I'll try to reproduce it | 14:12 |
| ralonsoh | maybe (maybe) is just a problem with the policy file | 14:12 |
| ralonsoh | next one | 14:13 |
| ralonsoh | #link https://bugs.launchpad.net/neutron/+bug/2030718 | 14:13 |
| ralonsoh | I really can't believe we can delete a network with a port created | 14:13 |
| ralonsoh | but of course we need to reproduce it | 14:13 |
| lajoskatona | I can check it | 14:13 |
| ralonsoh | if that is the case, this is a huge problem of course | 14:13 |
| ralonsoh | lajoskatona, thanks a lot, please assign it to you | 14:14 |
| lajoskatona | done | 14:14 |
| ralonsoh | lajoskatona, thanks! | 14:14 |
| ralonsoh | (that is going fast, thanks folks) | 14:14 |
| ralonsoh | next one | 14:14 |
| ralonsoh | #link https://bugs.launchpad.net/neutron/+bug/2031085 | 14:14 |
| ralonsoh | based in Train | 14:14 |
| ralonsoh | but could maybe be reproducible in newer versions | 14:15 |
| ralonsoh | also consider we have changed the callbacks during the last cycles | 14:15 |
| ralonsoh | changing the payloads | 14:15 |
| ralonsoh | in any case, I'm missing some information in the description | 14:15 |
| ralonsoh | for example, when this is happening, what operation | 14:15 |
| ralonsoh | so I'll ping the reported to provide more info, if you don't mind | 14:16 |
| ralonsoh | next one | 14:16 |
| ralonsoh | #link https://bugs.launchpad.net/neutron/+bug/2030741 | 14:16 |
| ralonsoh | sorry but I'm not sure what this bug is describing | 14:17 |
| ralonsoh | the point is that we have a hardcoded limit of HA GWs | 14:17 |
| ralonsoh | 5 | 14:17 |
| ralonsoh | and if the selected GWs, by the scheduler, belong to the same AZ (nothing prevents from that and it shouldn't) the 5 GWs will belong to the same AZ | 14:18 |
| ralonsoh | but this is correct | 14:18 |
| ralonsoh | because I've been "playing" recently with the OVN scheduler, I'll ask the reported what is wrong or what is he expecting | 14:19 |
| ralonsoh | because as I said, what is describing is a correct behaviour | 14:19 |
| ralonsoh | any comment? | 14:19 |
| mlavalle | yeah, seems correct bahavior | 14:20 |
| ralonsoh | agree | 14:20 |
| ralonsoh | so, again, I'll comment on the bug because maybe what he is expecting is having GWs from different AZs | 14:20 |
| ralonsoh | this is what I guess | 14:21 |
| ralonsoh | ok, let's jump to the next one | 14:21 |
| ralonsoh | #link https://bugs.launchpad.net/neutron/+bug/2030735 | 14:21 |
| ralonsoh | slaweq, proposed two possible solutions | 14:21 |
| ralonsoh | I think this could be a low hanging fruit | 14:22 |
| ralonsoh | please check it and consider taking it, if you have time | 14:22 |
| ralonsoh | and that's all I have!! | 14:23 |
| ralonsoh | any other bug I missed? | 14:23 |
| ralonsoh | ok | 14:23 |
| ralonsoh | This week obondarev is the deputy, next week will be isabek. | 14:23 |
| obondarev | on it | 14:23 |
| ralonsoh | perfect! | 14:24 |
| isabek | o/ | 14:24 |
| ralonsoh | thanks folks | 14:24 |
| ralonsoh | let's jump to the next topic | 14:24 |
| ralonsoh | #topic specs | 14:24 |
| ralonsoh | #link https://review.opendev.org/q/project:openstack%252Fneutron-specs+status:open | 14:24 |
| ralonsoh | we have a pending RFE spec | 14:24 |
| ralonsoh | but for 2024.1 | 14:24 |
| ralonsoh | of course, if you have time please check it | 14:25 |
| ralonsoh | this is the one related to the OVn interconnect and the DB sync tools | 14:25 |
| ralonsoh | and how to make the tool aware of the OVN interconnect resources | 14:25 |
| ralonsoh | ah, and the new folder for 2024.1 | 14:25 |
| ralonsoh | let's wait for https://review.opendev.org/c/openstack/neutron-specs/+/891342 until next week | 14:26 |
| ralonsoh | once we have the bobcat release, ok? | 14:26 |
| lajoskatona | agree | 14:26 |
| mlavalle | https://review.opendev.org/c/openstack/neutron-specs/+/891204 still failing the docs job | 14:27 |
| ralonsoh | because we has pushed the spec on 2024.1 folder | 14:28 |
| ralonsoh | instead of rebasing the spec on top of the other patch | 14:28 |
| ralonsoh | but that's a minor error | 14:28 |
| mlavalle | ok | 14:28 |
| ralonsoh | ok, let's jump to the next topic | 14:28 |
| ralonsoh | #topic community_goals | 14:28 |
| ralonsoh | I'll skip the sRBAC section, because slaweq is on PTO | 14:29 |
| ralonsoh | 2) Neutron client deprecation | 14:29 |
| ralonsoh | lajoskatona, any update? | 14:29 |
| lajoskatona | not much | 14:29 |
| lajoskatona | the usual etherpad: https://etherpad.opendev.org/p/python-neutronclient_deprecation | 14:30 |
| lajoskatona | the sfc sdk patch is ready I would say (https://review.opendev.org/c/openstack/openstacksdk/+/887387 ) | 14:30 |
| ralonsoh | ah yes | 14:30 |
| ralonsoh | I'll check it tomorrow this one | 14:30 |
| lajoskatona | there's a doc patch als ofor it: https://review.opendev.org/c/openstack/neutron-lib/+/887193 | 14:30 |
| lajoskatona | please check it, it just adds the api-ref | 14:30 |
| ralonsoh | ah, I think this one is ready | 14:31 |
| ralonsoh | I need to make a last review | 14:31 |
| ralonsoh | and https://review.opendev.org/c/openstack/python-neutronclient/+/880629? | 14:31 |
| ralonsoh | is the SDK patch released? | 14:31 |
| lajoskatona | as a wip or PoC I checked Horizon: https://review.opendev.org/c/openstack/horizon/+/891205 | 14:31 |
| lajoskatona | yes for https://review.opendev.org/c/openstack/python-neutronclient/+/880629 sdk is released (1.3.0, but since than we have 1.4.0 I think) | 14:32 |
| ralonsoh | so perfect, we can merge it | 14:32 |
| ralonsoh | please, any other core review it ^^ | 14:32 |
| ralonsoh | --> https://review.opendev.org/c/openstack/python-neutronclient/+/880629 | 14:32 |
| mlavalle | ok, I'll review it after the meeting | 14:33 |
| lajoskatona | For sfc any comment are welcome (https://etherpad.opendev.org/p/python-neutronclient_deprecation#L40 ) For that the client patch is ready but first the sdk one must be merged and released) | 14:33 |
| lajoskatona | thanks for the attention | 14:33 |
| ralonsoh | I'll check the SFC part | 14:34 |
| ralonsoh | lajoskatona, again, thanks for your efforts on this topic | 14:34 |
| obondarev | +1 | 14:34 |
| ralonsoh | so let's jump to the last topic | 14:35 |
| ralonsoh | #topic on_demand | 14:35 |
| ralonsoh | we have two topics but i don't know if these were discussed last week | 14:35 |
| ralonsoh | #link https://lists.openstack.org/pipermail/openstack-discuss/2023-August/034608.html | 14:35 |
| ralonsoh | (slaweq) Do we want to add neutron-core to the openstacksdk/OSC "service-core" group? | 14:36 |
| ralonsoh | did you talk about it? | 14:36 |
| obondarev | it was discussed | 14:36 |
| frickler | there is a patch waiting for your approval | 14:36 |
| ralonsoh | sorry, I was on medical leave | 14:36 |
| ralonsoh | and I still need to catch them all | 14:36 |
| frickler | everyone else was +1 on it afaict | 14:36 |
| ralonsoh | do you have the link? | 14:36 |
| haleyb | https://review.opendev.org/c/openstack/project-config/+/890914 | 14:36 |
| ralonsoh | so perfect, you agreed on this, correct? | 14:37 |
| frickler | yes, just slaweq wanted PTL vote | 14:37 |
| opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/2023.1: [OVN] ovn-db-sync check for router port differences https://review.opendev.org/c/openstack/neutron/+/891501 | 14:37 |
| obondarev | yes AFAIR | 14:37 |
| ralonsoh | then I'll +1 it right now | 14:37 |
| ralonsoh | thanks for the update! | 14:38 |
| ralonsoh | and now we have more responsibilities!! | 14:38 |
| ralonsoh | so that's all I have | 14:38 |
| ralonsoh | anything else you want to add? | 14:38 |
| frickler | survey questions | 14:39 |
| frickler | slaweg made an etherpad, let me look for the link | 14:39 |
| opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/zed: [OVN] ovn-db-sync check for router port differences https://review.opendev.org/c/openstack/neutron/+/891502 | 14:39 |
| frickler | https://etherpad.opendev.org/p/neutron-user-survey-questions-2023 | 14:39 |
| ralonsoh | ah yes, I need to review this yet | 14:40 |
| lajoskatona | another hot topic is the unmaintained resolution: https://review.opendev.org/c/openstack/governance/+/888771, which will affect all projects | 14:40 |
| ralonsoh | one sec | 14:40 |
| ralonsoh | who is going to receive this survey? | 14:40 |
| opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/yoga: [OVN] ovn-db-sync check for router port differences https://review.opendev.org/c/openstack/neutron/+/891503 | 14:40 |
| frickler | all openstack deployers are asked to take the survey once per year | 14:40 |
| ralonsoh | ok so this is the Neutron part of the survey | 14:41 |
| frickler | and the foundation collects the results | 14:41 |
| ralonsoh | perfect | 14:41 |
| frickler | yes | 14:41 |
| opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/xena: [OVN] ovn-db-sync check for router port differences https://review.opendev.org/c/openstack/neutron/+/891504 | 14:41 |
| ralonsoh | I'll check it tomorrow morning, for sure | 14:41 |
| opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/wallaby: [OVN] ovn-db-sync check for router port differences https://review.opendev.org/c/openstack/neutron/+/891506 | 14:41 |
| ralonsoh | lajoskatona, https://review.opendev.org/c/openstack/governance/+/888771 | 14:41 |
| ralonsoh | yeah, to be honest there are as much opinions as people in the community | 14:42 |
| ralonsoh | IMO, the documents as is now, is correct | 14:42 |
| lajoskatona | exactly, it is under review, but I suppose before the new release it will be finalized | 14:42 |
| ralonsoh | IMO, we can approve this document as is now and the improve these areas with flaws | 14:43 |
| ralonsoh | we don't need to release a perfect "rosetta stone" document | 14:43 |
| ralonsoh | I'll make a final review of it, but I'll +1 it | 14:44 |
| ralonsoh | do you have any comment about it? | 14:44 |
| ralonsoh | something that doesn't work for Neutron, for example | 14:44 |
| ralonsoh | ok, you can also review the ongoing patch too | 14:45 |
| ralonsoh | anything else you want to add? | 14:45 |
| haleyb | i have a comment on something else | 14:46 |
| ralonsoh | please | 14:46 |
| haleyb | argh, cut/paste not working, but just noticed our check queue is broken | 14:46 |
| haleyb | so maybe a topic for CI meeting, but neutron-tempest-plugin-openvswitch* jobs not happy | 14:47 |
| frickler | devstack made some changes, I can take a look later | 14:47 |
| frickler | so the global_venv is at work now, also support for bookworm added | 14:48 |
| haleyb | i think it started late last week, and it's only those jobs for some reason, thanks frickler for taking a look! | 14:48 |
| ralonsoh | haleyb, please open a LP bug | 14:48 |
| ralonsoh | with critical state | 14:48 |
| lajoskatona | good catch, by this on Friday: https://zuul.opendev.org/t/openstack/builds?job_name=neutron-tempest-plugin-openvswitch | 14:49 |
| frickler | I have another question: does anyone know how to compile OVN 23.06 in devstack? I always get compilation errors | 14:49 |
| ralonsoh | frickler, what errors? | 14:49 |
| ralonsoh | you need to have the correct OVS version | 14:50 |
| ralonsoh | let me check which one | 14:50 |
| ralonsoh | ok, we can discuss this out of this meeting | 14:50 |
| ralonsoh | anything else? | 14:50 |
| ralonsoh | ok, thank you all for attending | 14:51 |
| ralonsoh | #endmeeting | 14:51 |
| opendevmeet | Meeting ended Tue Aug 15 14:51:28 2023 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 14:51 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/networking/2023/networking.2023-08-15-14.00.html | 14:51 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/networking/2023/networking.2023-08-15-14.00.txt | 14:51 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/networking/2023/networking.2023-08-15-14.00.log.html | 14:51 |
| obondarev | o/ | 14:51 |
| lajoskatona | o/ | 14:51 |
| rubasov | o/ | 14:51 |
| mlavalle | o/ | 14:51 |
| ralonsoh | frickler, one sec, I'm checking the code now | 14:51 |
| frickler | ralonsoh: I tried with latest OVS, I don't have the logs right now, will need to retry | 14:51 |
| haleyb | i created https://bugs.launchpad.net/neutron/+bug/2031415 for the gate failure | 14:55 |
| ralonsoh | haleyb, thanks | 14:55 |
| frickler | I don't see anything obvious for the OVS failures so far, but the timing of when they start looks suspicious | 14:57 |
| frickler | https://review.opendev.org/c/openstack/devstack/+/558930 merged 18 UTC on the 11th | 14:58 |
| haleyb | and i see there is an opt-out in grenade | 14:59 |
| haleyb | so i suppose we can play with GLOBAL_VENV=false or ??? | 15:01 |
| frickler | yes, upgrade testing would need further work | 15:01 |
| frickler | we can try whether that solves the issue, good idea, yes | 15:01 |
| ralonsoh | frickler, please use "git submodule" in the OVN repository | 15:03 |
| ralonsoh | e.g.: | 15:03 |
| ralonsoh | stack@u22ovn:/opt/stack/ovn$ git submodule | 15:03 |
| ralonsoh | -0187eadfce4505d502e57c0e688b830f0a1ec728 ovs | 15:03 |
| haleyb | i can throw up a neutron patch as a test i suppose, just need coffee first | 15:03 |
| ralonsoh | btw, I'll catch all this tomorrow | 15:03 |
| zigo | git-submodule sucks, it's always a real pain... :/ | 15:04 |
| mlavalle | ralonsoh: do we have ci meeting today? | 15:04 |
| ralonsoh | no if slaweq didn't send an update | 15:04 |
| ralonsoh | and I didn't prepare it | 15:04 |
| mlavalle | ack, thanks! He didn't | 15:04 |
| ralonsoh | I was in medical leave until yesterday | 15:04 |
| lajoskatona | ack, I just wanted to ask the same | 15:05 |
| mlavalle | now I remember that today is a Holiday in Poland | 15:05 |
| ralonsoh | and in Spain | 15:05 |
| ralonsoh | so I'm leaving in 10 seconds | 15:05 |
| mlavalle | yeah, that's true, Good Catholic countries | 15:05 |
| ralonsoh | exactly | 15:05 |
| mlavalle | enjioy | 15:05 |
| ralonsoh | thanks | 15:05 |
| zigo | Not in Swiss ... :/ | 15:05 |
| mlavalle | nor the US | 15:06 |
| -opendevstatus- NOTICE: Zuul job execution is temporarily paused while we rearrange local storage on the servers | 16:54 | |
| opendevreview | Brian Haley proposed openstack/neutron master: DNM: Test if GLOBAL_VENV broke tempest jobs https://review.opendev.org/c/openstack/neutron/+/891519 | 17:10 |
| opendevreview | Brian Haley proposed openstack/neutron master: DNM: Test if GLOBAL_VENV broke tempest jobs https://review.opendev.org/c/openstack/neutron/+/891519 | 17:13 |
| opendevreview | Miro Tomaska proposed openstack/neutron master: OVN Metadata Agent check if haproxy is already running https://review.opendev.org/c/openstack/neutron/+/890986 | 17:21 |
| -opendevstatus- NOTICE: Zuul job execution has resumed with additional disk space on the servers | 17:44 | |
| opendevreview | Miro Tomaska proposed openstack/neutron master: Fix ovn-metadata agent sync of unused namespaces https://review.opendev.org/c/openstack/neutron/+/891232 | 19:00 |
| haleyb | frickler: so my test patch for GLOBAL_VENV=false didn't seem to work, https://review.opendev.org/c/openstack/neutron/+/891519 - or should I have changed this in the tempest repo? From what I can tell the tests are just running really slow in some cases, is that a side-effect of this change? | 20:40 |
| frickler | haleyb: I think this is the correct location https://opendev.org/openstack/neutron-tempest-plugin/src/branch/master/zuul.d/master_jobs.yaml#L25 | 21:18 |
| haleyb | frickler: ah, that makes sense, let me update things before I leave for the day | 21:18 |
| frickler | I wouldn't know how things would get slower when running from a venv, but since I didn't test for that, I also cannot confirm that this does not happen | 21:19 |
| frickler | haleyb: you may also consider setting the MYSQL_* vars for tempest plugin | 21:22 |
| frickler | though again I'm not sure whether they may help or make things worse | 21:22 |
| opendevreview | Brian Haley proposed openstack/neutron-tempest-plugin master: DNM: Test if GLOVAL_VENV broke tempest jobs https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/891550 | 21:26 |
| opendevreview | Brian Haley proposed openstack/neutron master: DNM: Test if GLOBAL_VENV broke tempest jobs https://review.opendev.org/c/openstack/neutron/+/891519 | 21:26 |
| haleyb | so the n-t-p change seems to have helped, at least one of the jobs has passed, will look closer tomorrow but I guess we do what grenade did? | 22:45 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!