| opendevreview | yatin proposed openstack/neutron master: [DNM] check revert with cs10 stream fix https://review.opendev.org/c/openstack/neutron/+/1002013 | 06:42 |
|---|---|---|
| opendevreview | yatin proposed openstack/neutron stable/2026.1: [CI][FT] Tempoarary skip failing tests centos 9 stream https://review.opendev.org/c/openstack/neutron/+/1002014 | 06:48 |
| opendevreview | yatin proposed openstack/neutron stable/2025.1: [CI][FT] Tempoarary skip failing tests centos 9 stream https://review.opendev.org/c/openstack/neutron/+/1002015 | 06:49 |
| *** elodilles_pto is now known as elodilles | 07:26 | |
| opendevreview | Merged openstack/neutron-tempest-plugin master: [CI][2025.1] Skip broken network segment range test in 2025.1 jobs https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/1001746 | 07:59 |
| opendevreview | Renjing Xiao proposed x/whitebox-neutron-tempest-plugin master: Add retry for traffic flow check after OVS failback https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/1002025 | 08:42 |
| opendevreview | Harald Jensås proposed openstack/neutron master: ml2: Filter dynamic segments at bind entry point https://review.opendev.org/c/openstack/neutron/+/1002026 | 08:48 |
| opendevreview | Harald Jensås proposed openstack/neutron master: ml2: Remove DB fallback from _expand_segment https://review.opendev.org/c/openstack/neutron/+/1002027 | 08:48 |
| hjensas | cardoe, tbachman: standing up lab to test the revert, I think with the reverse order on unbind change it might be ok -> (https://review.opendev.org/c/openstack/neutron/+/989141). Also proposed ^^ as possible alternative that moves the filtering of dynamic segments. | 09:28 |
| opendevreview | Merged openstack/neutron stable/2025.1: Allow shrinking a segment range to an allocated boundary https://review.opendev.org/c/openstack/neutron/+/1001168 | 10:10 |
| opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider master: Document SOURCE_IP lb-algorithm support in OVN provider guides https://review.opendev.org/c/openstack/ovn-octavia-provider/+/1002039 | 10:24 |
| opendevreview | Fernando Royo proposed openstack/neutron master: Doc: OVN driver supports SOURCE_IP lb-algorithm https://review.opendev.org/c/openstack/neutron/+/1002044 | 10:47 |
| opendevreview | Lajos Katona proposed openstack/neutron-vpnaas master: Rally scenario plugin API changed https://review.opendev.org/c/openstack/neutron-vpnaas/+/1000417 | 11:03 |
| opendevreview | Eduardo Olivares proposed openstack/neutron-tempest-plugin master: [evpn] Fix router cleanup https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/1000968 | 11:03 |
| mohsen______ | ralonsoh: slaweq | 11:24 |
| mohsen______ | ralonsoh: slaweq | 11:25 |
| mohsen______ | hi friends. I added a release note. would you please check this patch again? | 11:25 |
| mohsen______ | https://review.opendev.org/c/openstack/neutron-fwaas/+/989857 | 11:25 |
| mohsen______ | thank you in advance | 11:25 |
| mohsen______ | slaweq: | 11:26 |
| mohsen______ | ralonsoh: | 11:26 |
| mohsen______ | Hi friends. I added a release note. Would you please check this patch again? | 11:26 |
| mohsen______ | https://review.opendev.org/c/openstack/neutron-fwaas/+/989857 | 11:26 |
| mohsen______ | Thank you in advance | 11:26 |
| ralonsoh | I'll check it later today | 11:29 |
| mohsen______ | ralonsoh: Thank you | 11:31 |
| opendevreview | Harald Jensås proposed openstack/neutron master: ml2: Filter dynamic segments at bind entry point https://review.opendev.org/c/openstack/neutron/+/1002026 | 11:39 |
| opendevreview | Eduardo Olivares proposed openstack/neutron master: [bgp] Bump ovn and ovn versions https://review.opendev.org/c/openstack/neutron/+/1002065 | 11:43 |
| tbachman | hjensas: Thanks! Just curious - what are the MDs you have for HPB? | 12:00 |
| opendevreview | Eduardo Olivares proposed openstack/neutron master: [bgp] Bump ovn and ovn versions https://review.opendev.org/c/openstack/neutron/+/1002065 | 12:49 |
| hjensas | tbachman_: I test with this combination: ovn,baremetal-l2vni,genericswitch | 13:11 |
| tbachman_ | hjensas: thx! Which MD is the one that has the issue with the missing segment? | 13:14 |
| hjensas | tbachman_: genericswitch - see: https://paste.opendev.org/show/bbi7npPpJFCsH2SuGObm/ | 13:17 |
| hjensas | tbachman_: NOTE, that is with both your revert + this change by cardoe reverted: https://review.opendev.org/c/openstack/neutron/+/989141 | 13:18 |
| hjensas | tbachman_: afict from testing, the reversing the order of MD's on delete alone also works. | 13:19 |
| tbachman_ | hjensas: let me have a look at that MD | 13:19 |
| tbachman_ | I'm wondering if the baremetal-l2vni MD is incorrectly releasing the dynamic segment. In this case, is the baremetal VNIC bound to this segment? | 13:25 |
| opendevreview | Eduardo Olivares proposed openstack/neutron master: Add upstream EVPN multinode Zuul job https://review.opendev.org/c/openstack/neutron/+/994257 | 13:27 |
| tbachman_ | I guess one question with HPB is: who is responsible for releasing the dynamically allocated segment? Is it the allocating MD? Or is it the MD for the last port unbound to that segment? | 13:27 |
| hjensas | baremetal-l2vni is creating the segment, and it is also releaseing it when there are no more ports https://opendev.org/openstack/networking-baremetal/src/branch/master/networking_baremetal/plugins/ml2/baremetal_l2vni_mapping.py#L389 | 13:29 |
| tbachman_ | Hmmm - but then why is the networking-generic-switch considering it? | 13:29 |
| ralonsoh | tbachman_, hjensas what solution should we take: | 13:31 |
| ralonsoh | https://review.opendev.org/c/openstack/neutron/+/1002026 | 13:31 |
| ralonsoh | https://review.opendev.org/c/openstack/neutron/+/994936 | 13:31 |
| ralonsoh | because both cannot be merged | 13:32 |
| hjensas | generic switch is the responsible for the switch config, it need to read the segmentation ID of the segment. https://opendev.org/openstack/networking-generic-switch/src/branch/master/networking_generic_switch/generic_switch_mech.py#L1117 | 13:33 |
| tbachman_ | ralonsoh: I'm wondering if neither of these should be used. Instead a change might be made in networking-baremetal (and/or networking-generic-switch) | 13:33 |
| tbachman_ | In my mind, the question is: who is responsible for releasing the dynamically allocated segment? | 13:33 |
| tbachman_ | If it's the last user of that segment, then all MDs need to be HPB aware, in that they need to check if the segment being unbound is dynamically allocated, and if there are no more users, then free it | 13:34 |
| tbachman_ | (in which case, I'm wondering if there should just be logic in driver_context.py that does this for all MDs) | 13:34 |
| tbachman_ | If it's the allocating MD, then we need to figure out a way so that MDs that bound to the dynamically allocated segment don't have an issue with it. | 13:35 |
| tbachman_ | (which is what I think hjensas is reporting as the issue) | 13:37 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron-vpnaas master: zuul: add LibreSwan jobs to check and periodic-weekly https://review.opendev.org/c/openstack/neutron-vpnaas/+/1002112 | 13:38 |
| tbachman_ | One subtlety about HPB that I hadn't noticed before - the original architecture allows for the same network segment being used at different levels. I guess one example would be that a VLAN type segment on a compute host, bound by OVN, and that same segment (VLAN ID/physnet) could be used on a different compute host,which has more layers in thei hierarchy | 13:40 |
| tbachman_ | (e.g. spine => leaf => ToR => compute for one case, and spine => leaf => compute for another.... but same segment) | 13:41 |
| tbachman_ | I don't have a good reason for this, but my gut says that we should have the MD that has bound the segment do the free. | 13:42 |
| tbachman_ | (but again, only after ensuring that no other ports are bound to this segment) | 13:43 |
| tbachman_ | It would also be good to get cardoe's thoughts on this, as I know he has looked into this area (and is the author of the patch that merged to reverse order for deletes) | 13:47 |
| cardoe | Define MD for me? | 13:48 |
| tbachman_ | Mechanism Driver (sorry - should have avoided the shorthand) | 13:48 |
| cardoe | okay just making sure I was on the same page. | 13:49 |
| cardoe | So I've got a driver internally that's like baremetal-l2vni... in fact baremetal-l2vni started out as a patch I wrote extracting our driver. I'm working to drop our internal version and migrate to baremetal-l2vni. | 13:50 |
| hjensas | tbachman_: I agree that it makes sense for MD (Mechanism Driver) to own both bind and release. But your idea of moving it out of the mech driver is interesting, if the framework could do it without per Mech Driver involvement. | 13:50 |
| cardoe | tbachman_: that's one of the reasons why I'm doing the reversing. our baremetal-l2vni is the accounting MD if you will. It allocates the dynamic segment and releases the dynamic segment. | 13:51 |
| cardoe | So it's the first thing called in a bind/create/update operation and then last thing called in an unbind/delete operation. | 13:51 |
| cardoe | ralonsoh: tbachman_: hjensas: in fact since we're talking about design items... https://review.opendev.org/c/openstack/neutron/+/1000482 and https://review.opendev.org/c/openstack/neutron/+/1000485 are two different ways I've implemented the same fix... I'd appreciate feedback on which way to take... this is for VLAN -> VNI binding. | 13:56 |
| cardoe | tbachman_: hjensas: if we take a path on how it should flow, I'll update https://review.opendev.org/c/openstack/neutron/+/965415 I'm gonna move this over to networking-baremetal and maybe we can agree to document the correct convention of operations there? Then we can work on adding tests to neutron and tempest to enforce that flow? | 13:57 |
| ykarel | #startmeeting neutron_ci | 14:00 |
| opendevmeet | Meeting started Mon Aug 24 14:00:50 2026 UTC and is due to finish in 60 minutes. The chair is ykarel. 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 'neutron_ci' | 14:00 |
| tbachman_ | hjensas: the check for this would ideally live in the precommiit, as that's within a DB transaction, which would give protection for concurrent operations to avoid cases where multiple ports are unbinding from the same segment (e.g. last 2 users) | 14:00 |
| ralonsoh | hello | 14:00 |
| ykarel | Ping list: bcafarel, lajoskatona, slawek, mlavalle, mtomaska, ralonsoh, ykarel, jlibosva, elvira | 14:01 |
| ralonsoh | ykarel, irc or video? | 14:01 |
| ykarel | Hi all, it's IRC today | 14:01 |
| ralonsoh | cool | 14:01 |
| ykarel | let's give a few minutes for others to join | 14:02 |
| slaweq | o/ | 14:02 |
| lajoskatona | o/ | 14:04 |
| hjensas | tbachman_: I need to go in about 30 min, but I guess we should pause to not interfere the meeting. | 14:04 |
| ykarel | thx hjensas | 14:04 |
| ykarel | Hi all, Let's start with the topics. meanwhile if others around can join | 14:04 |
| ykarel | #topic Actions from previous meetings | 14:05 |
| ykarel | ykarel to look into cs9 failures and open a centos bug for it and skip tests in CI | 14:05 |
| ykarel | Reported https://redhat.atlassian.net/browse/RHEL-246486 | 14:05 |
| ykarel | Skipping https://review.opendev.org/c/openstack/neutron/+/1002014 and https://review.opendev.org/c/openstack/neutron/+/1002015 | 14:05 |
| ykarel | please do check ^ patches | 14:05 |
| ykarel | #topic Stable branches | 14:05 |
| ykarel | all good with stable, except ^ periodic failures | 14:06 |
| cardoe | 2025.1 seems like its got problems again with neutron-tempest-plugin | 14:06 |
| cardoe | https://zuul.opendev.org/t/openstack/build/456ab4b2c0dc42ccb2490670fbeb4c48 | 14:06 |
| ykarel | review wise stable patches mostly handled | 14:07 |
| ykarel | thx cardoe , seems to be new issue started today | 14:07 |
| ykarel | https://zuul.opendev.org/t/openstack/builds?job_name=neutron-tempest-plugin-ovn-ubuntu-jammy-2025-1&project=openstack/neutron-tempest-plugin | 14:07 |
| ralonsoh | something again with setuptools dependencies? | 14:07 |
| ralonsoh | I think we had this problem before | 14:08 |
| ralonsoh | I'll check it | 14:08 |
| ykarel | new pbr release few hours back https://pypi.org/project/pbr/#history | 14:08 |
| ralonsoh | that makes sense | 14:08 |
| ykarel | so likely need to pin pbr if only stable 2025.1 impacted | 14:09 |
| ykarel | thx ralonsoh will add action for you to check this | 14:09 |
| ykarel | #action ralonsoh to check issue with latest pbr release in 2025.1 + jammy | 14:10 |
| ykarel | so noble looks good only jammy + 2025.1 impacted | 14:10 |
| ykarel | #topic Stadium projects | 14:10 |
| ykarel | Mostly green. neutron-vpnaas had one periodic-weekly job failure https://zuul.openstack.org/build/36979791a4f24e46b9e9a8183f487c25 | 14:11 |
| ykarel | Known one https://launchpad.net/bugs/2163189 | 14:11 |
| ykarel | lajoskatona looking into this | 14:11 |
| ykarel | anything else to share for stadiums lajoskatona ? | 14:11 |
| lajoskatona | for vpnaas only | 14:11 |
| lajoskatona | the patch for rally: https://review.opendev.org/c/openstack/neutron-vpnaas/+/1000417 | 14:12 |
| lajoskatona | but the rally job is still failing | 14:12 |
| lajoskatona | so ongoing | 14:12 |
| ykarel | thx lajoskatona | 14:12 |
| ykarel | #topic Rechecks | 14:13 |
| ykarel | not much activity this week, there were few rechecks | 14:13 |
| ykarel | bare recheck wise it was 3/5 bare recheck, all 3 were in a test patch | 14:13 |
| ykarel | now let's check on failures | 14:15 |
| ykarel | #topic Unit tests/pep8 | 14:15 |
| ykarel | all good with non voting py314 | 14:15 |
| ykarel | #topic Periodic | 14:16 |
| ykarel | https://zuul.openstack.org/builds?job_name=neutron-functional-fips&project=openstack%2Fneutron&branch=stable%2F2025.1&branch=stable%2F2026.1&skip=0 | 14:16 |
| ykarel | this we already discussed just need to get the skip patches merged | 14:16 |
| ykarel | ovn bgp still not healthy https://zuul.opendev.org/t/openstack/builds?job_name=neutron-ovn-bgp-tempest-multinode&project=openstack%2Fneutron&skip=0 | 14:16 |
| ralonsoh | I think eolivare_ is working on it | 14:16 |
| ykarel | ok good | 14:17 |
| ralonsoh | something related to deleting admin routers | 14:17 |
| ykarel | ack | 14:17 |
| eolivare_ | admin routers issue is for evpn (not merged yet) | 14:17 |
| eolivare_ | for the bgp job timeouts -> https://review.opendev.org/c/openstack/neutron/+/1001108 | 14:17 |
| ykarel | seems still need more work ? https://zuul.opendev.org/t/openstack/builds?job_name=neutron-ovn-bgp-tempest-multinode&project=openstack%2Fneutron&change=1001108&skip=0 | 14:19 |
| ykarel | ok let's move | 14:20 |
| ykarel | #topic Grafana | 14:20 |
| ykarel | https://grafana.opendev.org/d/f913631585/neutron-failure-rate | 14:21 |
| ykarel | let's have quick look here as well | 14:21 |
| ykarel | looks all good, check spikes looks patches specific | 14:22 |
| ykarel | anything to add? | 14:22 |
| ralonsoh | all good | 14:22 |
| ykarel | #topic On Demand | 14:23 |
| ykarel | anything else you would like to raise? | 14:23 |
| ralonsoh | no thanks | 14:23 |
| ykarel | i have a small one, pushed a revert https://review.opendev.org/c/openstack/neutron/+/1001808, so do check that | 14:24 |
| ralonsoh | nice | 14:24 |
| lajoskatona | nothing more from me, thanks | 14:24 |
| ykarel | ok thx, in that case let's close early and have everyone almost 35 minutes back | 14:25 |
| ykarel | #endmeeting | 14:25 |
| opendevmeet | Meeting ended Mon Aug 24 14:25:30 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 14:25 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/neutron_ci/2026/neutron_ci.2026-08-24-14.00.html | 14:25 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/neutron_ci/2026/neutron_ci.2026-08-24-14.00.txt | 14:25 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/neutron_ci/2026/neutron_ci.2026-08-24-14.00.log.html | 14:25 |
| ralonsoh | bye | 14:25 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron-vpnaas master: rally: fix ``create_server`` passing duplicate argument https://review.opendev.org/c/openstack/neutron-vpnaas/+/1002125 | 14:25 |
| ralonsoh | lajoskatona, check this ^^ | 14:25 |
| ralonsoh | push your rally patch on top | 14:25 |
| lajoskatona | ralonsoh: thanks, checking | 14:26 |
| lajoskatona | I just started to fix the sec-group create calls | 14:26 |
| lajoskatona | and added a mental note that rally is also using neutronclient :-) | 14:26 |
| ralonsoh | that is a typing problem | 14:26 |
| ralonsoh | I think we can just pass the SG name and that's all | 14:27 |
| lajoskatona | ok, I check it | 14:28 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron-lib master: placement: Raise ``PlacementResourceProviderNotFound`` on delete https://review.opendev.org/c/openstack/neutron-lib/+/1002126 | 14:36 |
| opendevreview | Eduardo Olivares proposed openstack/neutron-tempest-plugin master: Fix admin router cleanup https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/1000968 | 14:52 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: Reduce ``test_get_and_delete_gateway`` privsep load https://review.opendev.org/c/openstack/neutron/+/1002135 | 15:01 |
| opendevreview | Eduardo Olivares proposed openstack/neutron master: [bgp] bump cirros version to 0.6.2 https://review.opendev.org/c/openstack/neutron/+/1002137 | 15:10 |
| opendevreview | Slawek Kaplonski proposed openstack/neutron master: Use RBAC API policy to create default security group https://review.opendev.org/c/openstack/neutron/+/1002138 | 15:10 |
| opendevreview | Lajos Katona proposed openstack/neutron-vpnaas master: Rally scenario plugin API changed https://review.opendev.org/c/openstack/neutron-vpnaas/+/1000417 | 15:15 |
| cardoe | ykarel: ralonsoh: https://review.opendev.org/c/openstack/pbr/+/1002140 might fix that issue with 2025.1 | 15:33 |
| opendevreview | Miro Tomaska proposed openstack/neutron master: Raise error when adding another `advertise-host` subnet to a evpn router https://review.opendev.org/c/openstack/neutron/+/1000789 | 15:43 |
| opendevreview | Takashi Kajinami proposed openstack/os-vif master: zuul: Drop redundant override of configure_swap_size https://review.opendev.org/c/openstack/os-vif/+/1002161 | 16:15 |
| opendevreview | Merged openstack/neutron master: Revert "[CI][FT] Tempoarary skip failing tests centos 10 stream" https://review.opendev.org/c/openstack/neutron/+/1001808 | 16:55 |
| opendevreview | Lajos Katona proposed openstack/neutron-vpnaas master: Introduce OVO objects for VPNaaS resources https://review.opendev.org/c/openstack/neutron-vpnaas/+/1002170 | 16:57 |
| opendevreview | Lajos Katona proposed openstack/neutron-vpnaas master: Rally scenario plugin API changed https://review.opendev.org/c/openstack/neutron-vpnaas/+/1000417 | 17:03 |
| opendevreview | Merged openstack/neutron stable/2026.1: [CI][FT] Tempoarary skip failing tests centos 9 stream https://review.opendev.org/c/openstack/neutron/+/1002014 | 17:42 |
| opendevreview | Miro Tomaska proposed x/whitebox-neutron-tempest-plugin master: [DNM] POC whitebox test which will allow inspecting state of chassis https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/1000366 | 17:47 |
| opendevreview | Merged openstack/neutron stable/2025.1: [CI][FT] Tempoarary skip failing tests centos 9 stream https://review.opendev.org/c/openstack/neutron/+/1002015 | 17:52 |
| opendevreview | Miro Tomaska proposed openstack/neutron-tempest-plugin master: Test that only one subnet with `advertise-host` is allowed per network https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/1000973 | 18:34 |
| opendevreview | Lajos Katona proposed openstack/neutron-vpnaas master: Rally scenario plugin API changed https://review.opendev.org/c/openstack/neutron-vpnaas/+/1000417 | 19:11 |
| opendevreview | Miro Tomaska proposed x/whitebox-neutron-tempest-plugin master: [DNM] POC whitebox test which will allow inspecting state of chassis https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/1000366 | 19:25 |
| opendevreview | Miro Tomaska proposed openstack/neutron master: Migrate vtysh commands from using rootwrap to oslo_privsep https://review.opendev.org/c/openstack/neutron/+/1002197 | 20:48 |
| *** Uggla is now known as Guest16245 | 21:34 | |
| *** Uggla7 is now known as Uggla | 21:34 | |
| opendevreview | Jakub Libosvar proposed openstack/neutron master: bgp: Make spine and leaf ansible roles generic https://review.opendev.org/c/openstack/neutron/+/1002215 | 22:55 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!