| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent master: Add support at NB BGP Driver for exposing tenant IPs https://review.opendev.org/c/openstack/ovn-bgp-agent/+/899117 | 07:13 |
|---|---|---|
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent master: Add support at NB BGP Driver for exposing OVN LBs https://review.opendev.org/c/openstack/ovn-bgp-agent/+/899120 | 08:50 |
| opendevreview | Merged openstack/neutron stable/xena: Ensure ovn loadbalancer FIPs are centralized upon neutron restarts https://review.opendev.org/c/openstack/neutron/+/900982 | 09:09 |
| opendevreview | Merged openstack/neutron stable/wallaby: Ensure ovn loadbalancer FIPs are centralized upon neutron restarts https://review.opendev.org/c/openstack/neutron/+/900886 | 09:10 |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent stable/2023.2: Fix [EVPN-Driver] Cannot remove routes https://review.opendev.org/c/openstack/ovn-bgp-agent/+/901065 | 09:33 |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent stable/2023.2: Devstack plugin https://review.opendev.org/c/openstack/ovn-bgp-agent/+/901066 | 09:34 |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent stable/2023.2: Add initial support for local OVN cluster instead of kernel-networking https://review.opendev.org/c/openstack/ovn-bgp-agent/+/901067 | 09:34 |
| opendevreview | Merged openstack/ovn-bgp-agent master: Add support at NB BGP Driver for exposing subnets https://review.opendev.org/c/openstack/ovn-bgp-agent/+/899116 | 09:36 |
| opendevreview | Merged openstack/ovn-bgp-agent master: Add support at NB BGP Driver for exposing tenant IPs https://review.opendev.org/c/openstack/ovn-bgp-agent/+/899117 | 09:36 |
| opendevreview | Merged openstack/ovn-bgp-agent master: Add support at NB BGP Driver for exposing OVN LBs https://review.opendev.org/c/openstack/ovn-bgp-agent/+/899120 | 09:36 |
| opendevreview | Lajos Katona proposed openstack/neutron master: DNM: test bug/2042941 https://review.opendev.org/c/openstack/neutron/+/901125 | 10:41 |
| lajoskatona | ykarel: Hi, -----^ and https://review.opendev.org/c/openstack/devstack/+/901124 | 10:41 |
| lajoskatona | ykarel: I try to keep myself far from devstack code changes so please consider that ;) | 10:42 |
| ykarel | lajoskatona, looking | 10:46 |
| opendevreview | yatin proposed openstack/neutron master: DNM: test bug/2042941 https://review.opendev.org/c/openstack/neutron/+/901125 | 10:48 |
| ykarel | lajoskatona, looks good, just updated test patch to have results just from the related jobs | 10:49 |
| lajoskatona | ykarel: thanks | 12:07 |
| opendevreview | Lajos Katona proposed openstack/neutron master: DNM: test bug/2042941 https://review.opendev.org/c/openstack/neutron/+/901125 | 12:52 |
| zigo | haleyb: No need to do a backport, I'll just carry the patch in the Debian Bobcat version of neutron-lib. | 12:52 |
| zigo | Though if the team feels like it's nice to have in Bobcat, your call ... :) | 12:52 |
| lajoskatona | ykarel: there's some trick which is over my limits :-( | 13:44 |
| lajoskatona | ykarel: https://6652f590ce298f0dccf9-a12e67e84744b622e6a13b507a2faa27.ssl.cf5.rackcdn.com/901125/3/check/neutron-ovn-tempest-with-sqlalchemy-master/b777267/job-output.txt | 13:44 |
| lajoskatona | ykarel: die 611 'ERROR_ON_CLONE is set to True so cloning not allowed in this configuration' , even if I explicitly set it to false in zuul.yaml | 13:44 |
| ykarel | lajoskatona, ohkk seems need to set some vars to get it work | 13:49 |
| ykarel | i will check and update that patch | 13:49 |
| ykarel | with the vars i meant GITREPO, GITDIR, GITBRANCH etc | 13:51 |
| ykarel | but need to check if those repos are also copied to /opt/stack/ like other openstack repos | 13:52 |
| lajoskatona | ykarel: thanks | 14:02 |
| opendevreview | yatin proposed openstack/neutron master: DNM: test bug/2042941 https://review.opendev.org/c/openstack/neutron/+/901125 | 14:21 |
| haleyb | zigo: ack, and i don't think it's necessary since it was just a test | 14:32 |
| opendevreview | Fernando Royo proposed openstack/neutron stable/2023.2: Add constant to identify OVN LB HM ports https://review.opendev.org/c/openstack/neutron/+/901164 | 14:44 |
| haleyb | otherwiseguy: hey terry, let me know if anything else on https://review.opendev.org/q/project:openstack/ovsdbapp is ready to merge, i was going to update the release hash in https://review.opendev.org/c/openstack/releases/+/900769 later to get it updated for C-1. thanks :) | 14:45 |
| otherwiseguy | @haleyb: I'll take a look. Thanks for the heads-up. | 14:46 |
| haleyb | that was why i bugged you about that other change | 14:46 |
| otherwiseguy | @haleyb: on that ovn-ic spec, I really think that should just be a bug "ovn-db-sync deletes objects neutron doesn't own" and doesn't really have anything specifically to do with ovn-ic. | 14:49 |
| * otherwiseguy is splitting his attention with a meeting right now | 14:50 | |
| opendevreview | Fernando Royo proposed openstack/neutron stable/2023.1: Add constant to identify OVN LB HM ports https://review.opendev.org/c/openstack/neutron/+/901165 | 14:51 |
| haleyb | otherwiseguy: yeah, i keep meaning to comment in that change, i think you're probably right | 14:51 |
| otherwiseguy | but after the meeting, I'll make a pass through the outstanding ovsdbapp patches, approve what is approvable, and if there are minor tweaks left to do, just make the changes. | 14:51 |
| racosta | Hey Terry, I didn't know your IRC nickname was otherwiseguy. Thanks for proposing the discussion about ovn-ic RFE to make it generic and not use specific names to match external_ids register. | 14:54 |
| otherwiseguy | Neutron pretending it owns ports: https://getyarn.io/yarn-clip/2e7b879e-b3fa-476b-9053-c46446007e08/gif | 14:55 |
| racosta | I would be happy to change the spec if it's ok for everyone. | 14:55 |
| *** jlibosva is now known as Guest7233 | 15:00 | |
| otherwiseguy | @racosta: hi! I'm not in charge, but sounds good to me. :D Feedback kind of stalled. @haleyb is this something that we need to treat as a spec/blueprint, or is a LP bug sufficient? If it is a bug and not a feature, I presume it would theoretically make it backportable? Anyone have opinions on that? | 15:01 |
| otherwiseguy | Society seems to treat it as a bug when I set things I don't own on fire. | 15:02 |
| haleyb | I thought we were an autonomous collective, but we could be a narco-syndicous commune /s | 15:03 |
| otherwiseguy | Not a "let's plan to improve terry with a feature where he doesn't go around setting other people's things on fire" situation. | 15:03 |
| otherwiseguy | @haleyb: lol | 15:03 |
| racosta | It would be nice to get feedback from Neutron core members on Terry's comment: https://review.opendev.org/c/openstack/neutron-specs/+/891204/comments/590bda54_6812fc6c | 15:04 |
| otherwiseguy | There are plenty of people I can poke at Red Hat for comments, but that doesn't seem like the healthiest way to get input. | 15:07 |
| haleyb | racosta: it's been on my list it's just always out of reach... i did bring it up at the weekly meeting so others are aware | 15:08 |
| haleyb | as far as bug or feature, i'd have to look through the meeting notes when it was approved, assuming there was more than just making ovn-db-sync-util play nicer | 15:09 |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent stable/2023.2: Add initial support for local OVN cluster instead of kernel-networking https://review.opendev.org/c/openstack/ovn-bgp-agent/+/901067 | 15:09 |
| racosta | haleyb, it's not a bug, it's Neutron's current behavior. So changing this behavior is a feature, and that is why an RFE was proposed. | 15:11 |
| otherwiseguy | Neutron has lots of current behaviors that are bugs. :) | 15:12 |
| racosta | fair enough ;) let's get better then | 15:13 |
| otherwiseguy | "We didn't think of this edge case" can sometimes be treated as a bug, sometimes not. Sometimes it's hard to decide which. | 15:13 |
| otherwiseguy | In this case, I think it'd come down to whether it was something we think would be bad to backport. | 15:14 |
| otherwiseguy | I can see arguments either way. | 15:15 |
| otherwiseguy | But ultimately, I'm having trouble imagining a scenario where someone is relying on neutron to delete objects that it doesn't own in the OVN DBs. | 15:15 |
| otherwiseguy | But sometimes my imagination isn't that good. So I invite other people to imagine as well. :D | 15:16 |
| haleyb | racosta: i guess it depends on perspective as terry said. if the change is to make the tool better detect neutron things, as opposed to ignore ovn-ic things, it solves the same problem. it's almost an implementation detail so feels like i'm getting ahead of myself | 15:17 |
| otherwiseguy | also neutron: https://imgur.com/a/q0HGJgY | 15:20 |
| racosta | yah, regardless of the way of treating resources not managed by neutron, it solves the ovn-ic problem, but if it is implemented in a generic way to only look at what has a neutron signature, 'it could solve the problem for many other people'... | 15:20 |
| opendevreview | Brian Haley proposed openstack/neutron stable/zed: Catch non-existent entry failures better in ip_lib https://review.opendev.org/c/openstack/neutron/+/901073 | 15:20 |
| otherwiseguy | @racosta: I'm also interested into looking at using ovn-ic with neutron. I haven't played with it at all though. Could be interesting to work on actual support in neutron for AZs. | 15:22 |
| otherwiseguy | I don't know much about it, but I'm trying to learn. Your spec actually helps me understand it better. :) | 15:24 |
| haleyb | racosta: right, it's opt-in versus opt-out, for when ovn-xyz comes along and we don't have to add another filter | 15:25 |
| otherwiseguy | racosta: Do you plan eventually to work actual integration of ovn-ic into neutron? Or is it better to handle that stuff outside of neutron? | 15:26 |
| * otherwiseguy notices he has another meeting in 3 mins | 15:28 | |
| racosta | Felix is planning to integrate TS into Neutron to use the existing gw_port chassis scheduler, etc. | 15:28 |
| racosta | But that's for the future, for now the implementation basis would be to not remove OVN interconnection resources | 15:29 |
| racosta | By the way, canonical is investing in LXD - microcloud uses LXD for Virtualisation and orchestration... the cool part is that the base is OVN, so it would be technically viable to interconnect this cloud with openstack ;) | 15:31 |
| otherwiseguy | +1 | 15:31 |
| *** jlibosva is now known as Guest7236 | 15:32 | |
| otherwiseguy | racosta: sounds cool. If anyone wants any extra hands working on the ovn-ic stuff in the future, I'm available. ;) | 15:33 |
| otherwiseguy | I at least plan on knowing more about it very shortly. \ | 15:33 |
| racosta | otherwiseguy, good to know! I will ping you | 15:37 |
| opendevreview | Merged openstack/ovn-bgp-agent stable/2023.2: Devstack plugin https://review.opendev.org/c/openstack/ovn-bgp-agent/+/901066 | 15:38 |
| opendevreview | Merged openstack/ovn-bgp-agent stable/2023.2: Fix [EVPN-Driver] Cannot remove routes https://review.opendev.org/c/openstack/ovn-bgp-agent/+/901065 | 15:38 |
| opendevreview | Merged openstack/ovn-bgp-agent stable/2023.2: Add initial support for local OVN cluster instead of kernel-networking https://review.opendev.org/c/openstack/ovn-bgp-agent/+/901067 | 15:38 |
| opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider stable/2023.2: Change device_owner used for OVN LB HM port https://review.opendev.org/c/openstack/ovn-octavia-provider/+/901178 | 16:10 |
| opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider stable/2023.2: Add maintenance task to update entities on component start https://review.opendev.org/c/openstack/ovn-octavia-provider/+/901179 | 16:11 |
| opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider stable/2023.1: Change device_owner used for OVN LB HM port https://review.opendev.org/c/openstack/ovn-octavia-provider/+/901180 | 16:17 |
| opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider stable/2023.1: Add maintenance task to update entities on component start https://review.opendev.org/c/openstack/ovn-octavia-provider/+/901181 | 16:19 |
| opendevreview | Brian Haley proposed openstack/neutron stable/zed: Catch non-existent entry failures better in ip_lib https://review.opendev.org/c/openstack/neutron/+/901073 | 17:21 |
| opendevreview | Merged openstack/neutron-tempest-plugin master: Fix importing security group tests https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/900772 | 17:23 |
| *** Continuity__ is now known as Continuity | 17:25 | |
| opendevreview | Merged openstack/ovsdbapp master: nb: allow to add route even if learned route exists https://review.opendev.org/c/openstack/ovsdbapp/+/896686 | 19:18 |
| haleyb | otherwiseguy: so is https://review.opendev.org/c/openstack/ovsdbapp/+/897090 ready to go? the comments are unclear, unless you want to refactor later | 19:39 |
| otherwiseguy | @jlibosva: and I are currently discussing the finer details of readability. :D | 19:39 |
| jlibosva | I'm not opposing, it's a very minor nit. I +2ed current patch as is | 19:40 |
| otherwiseguy | but it's so easy to be like "but we could make it even better!" Even I'm jumping on little changes. :p | 19:41 |
| otherwiseguy | haleyb: but yeah, from a functionality standpoint it's good. | 19:41 |
| haleyb | shave that yak! | 19:42 |
| otherwiseguy | Going through another test cycle to move where we define the timeout is probably not worth it with release stuff waiting. | 19:42 |
| otherwiseguy | for a class that's only used once in a test. :D | 19:42 |
| haleyb | well, that patch probably isn't critical, there's always C-2 | 19:42 |
| otherwiseguy | I WILL HAVE THE SMOOTHEST YAK. | 19:42 |
| * otherwiseguy is sometimes exhausting | 19:44 | |
| haleyb | mlavalle: i'm going to cancel the drivers meeting as i see nothing on the agenda and no new rfe bugs | 19:47 |
| opendevreview | Brian Haley proposed openstack/neutron stable/yoga: Catch non-existent entry failures better in ip_lib https://review.opendev.org/c/openstack/neutron/+/901074 | 19:52 |
| opendevreview | Brian Haley proposed openstack/neutron stable/yoga: Catch non-existent entry failures better in ip_lib https://review.opendev.org/c/openstack/neutron/+/901074 | 19:58 |
| otherwiseguy | haleyb: I'm looking at https://review.opendev.org/c/openstack/ovsdbapp/+/896682. It's the last one that I think could be ready for being included. I hate sleep() calls with a passion and never saw vtep tests failing as it was locally or in CI, but I don't ever do anything with vtep myself so... | 20:00 |
| otherwiseguy | I've run w/ and w/o the patch and same tests pass. And I've run with the patch, but with the wat_db_rows method commented out, and it still passes. | 20:02 |
| haleyb | otherwiseguy: let me know and i can look later, taking off early today and off tomorrow | 20:02 |
| otherwiseguy | +1 | 20:02 |
| otherwiseguy | have a good weekend! | 20:03 |
| haleyb | thanks, you too! | 20:05 |
| mlavalle | haleyb: LOL, I was about to ping you. Thanks for letting me know. Enjoy your day off | 20:09 |
| opendevreview | Bodo Petermann proposed openstack/neutron-vpnaas master: VPNaaS support for OVN https://review.opendev.org/c/openstack/neutron-vpnaas/+/765353 | 20:09 |
| opendevreview | Bodo Petermann proposed openstack/neutron-vpnaas master: VPNaaS support for OVN: documentation https://review.opendev.org/c/openstack/neutron-vpnaas/+/895651 | 20:09 |
| opendevreview | Merged openstack/ovsdbapp master: TAAS: Add commands for creating Mirrors https://review.opendev.org/c/openstack/ovsdbapp/+/890328 | 20:37 |
| opendevreview | Merged openstack/ovsdbapp master: Handle events with conditions and match_fn https://review.opendev.org/c/openstack/ovsdbapp/+/897090 | 20:43 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!