Monday, 2026-01-12

ykarelThanks for confirming and fixing the unstack issue04:34
ykareland hopefully you also didn't reproduced your original issue on fresh vm and stack.sh, right?04:34
opendevreviewyatin proposed openstack/neutron master: [DNM] Check migration ovn  https://review.opendev.org/c/openstack/neutron/+/97282005:10
ralonsohhello folks, please check https://review.opendev.org/c/openstack/neutron/+/97266908:22
ralonsohthis is for 2025.1 only08:22
ralonsohit is tested in https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/97226608:22
ralonsohslaweq, lajoskatona ^^ thanks!!08:22
daidv_1haleyb, lajoskatona Happy new year ^^ I updated patch for DNS forwarder extendsion, hope you guys can spent sometime to review it for me, thanks alot: https://review.opendev.org/c/openstack/neutron/+/95139008:28
lajoskatonadaidv_1, ralonsoh: Hi, on my list08:30
daidv_1thank you in advance :D 08:31
opendevreviewBrian Haley proposed openstack/networking-sfc master: Change the vxlan vni_range to be 1:1000  https://review.opendev.org/c/openstack/networking-sfc/+/96778308:58
*** srelf_ is now known as ContinuitySR09:25
*** ContinuitySR is now known as Continuity09:25
opendevreviewEduardo Olivares proposed x/whitebox-neutron-tempest-plugin master: [multicast] Only use vlan_transparent option when really required  https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/97289910:08
opendevreviewElvira García Ruiz proposed openstack/tap-as-a-service master: Add neutron-tempest-plugin ovn tests to zuul  https://review.opendev.org/c/openstack/tap-as-a-service/+/97305710:19
opendevreviewBrian Haley proposed openstack/networking-bgpvpn master: Remove py312 jobs from check and gate queue  https://review.opendev.org/c/openstack/networking-bgpvpn/+/97278610:33
*** dmellado4 is now known as dmellado11:41
opendevreviewJames Bhattarai proposed openstack/neutron master: Docs Changes in CLI FQDN of Neutron BGP Speaker Create  https://review.opendev.org/c/openstack/neutron/+/97307312:04
opendevreviewLajos Katona proposed openstack/neutron-fwaas master: CI: Add required_projects to unit test jobs  https://review.opendev.org/c/openstack/neutron-fwaas/+/97307912:33
opendevreviewEduardo Olivares proposed x/whitebox-neutron-tempest-plugin master: Remove try/except from VlanTransparencyWhiteboxBaseTest resource_setup  https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/97308012:34
opendevreviewMerged openstack/networking-sfc master: Change the vxlan vni_range to be 1:1000  https://review.opendev.org/c/openstack/networking-sfc/+/96778312:55
jlibosvaralonsoh: hi, I tried to play a bit with the functional tests setup/teardown resources and if we do not build a new DB for every test but only per test worker - the speedup is in the order of magnitude - from 2 hours down to 18 minutes. But there are still some issues there. I think there is another bug in the OVN base class. It's only OVN tests that have problems with the13:33
jlibosvaDB13:33
jlibosvaralonsoh: speaking about https://review.opendev.org/c/openstack/neutron/+/97298513:33
ralonsohjlibosva, let me check (let me finish one quick thing)13:42
opendevreviewEduardo Olivares proposed x/whitebox-neutron-tempest-plugin master: [multicast] Only use vlan_transparent option when really required  https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/97289914:21
opendevreviewEduardo Olivares proposed x/whitebox-neutron-tempest-plugin master: Remove try/except from VlanTransparencyWhiteboxBaseTest resource_setup  https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/97308014:21
ralonsohjlibosva, I don't think this is strictly related to the OVN tests14:26
ralonsohsome workers had initially problems with the DB:14:27
ralonsohsqlalchemy.exc.OperationalError: (pymysql.err.OperationalError) (1050, "Table 'agents' already exists")14:27
ralonsohthen the following tests using MySQL failed too14:27
jlibosvathe DB migrations tests make sense since they assume empty DB14:28
ralonsohjlibosva, so you can maybe check without `test_walk_versions`, `test_db_upgrade_is_idempotent` and `TestModelsMigrations`14:29
ralonsohjlibosva, what about Terry's suggestion about creating the DDL files and load them at the beginning of every MySQL test?14:30
jlibosvaralonsoh:here https://review.opendev.org/c/openstack/neutron/+/97296114:30
ralonsohcool! let me check it14:31
jlibosvaralonsoh: didn't help much though in terms of overall execution time14:31
jlibosvaralonsoh: that's why I started digging deeper and came up with not creating the schema per every test but reuse since even the DDL takes some time14:31
ralonsohso still 2 hours 14:31
jlibosvaralonsoh: I measured it manually and the DDL was significantly faster than create_all()14:32
jlibosvaralonsoh: https://bugs.launchpad.net/neutron/+bug/2136827/comments/414:32
ralonsohjlibosva, but if https://review.opendev.org/c/openstack/neutron/+/972961 is using the DDL files, why does it take 2 hours?14:34
jlibosvaralonsoh: I don't have a good answer to that yet14:37
jlibosvabut14:37
jlibosvaRan: 1147 tests in 1651.9493 sec. <- this is a significant improvement14:37
jlibosvaah no, it's not14:37
ralonsohlet me execute your patch locally14:38
ykarel#startmeeting neutron_ci15:01
opendevmeetMeeting started Mon Jan 12 15:01:14 2026 UTC and is due to finish in 60 minutes.  The chair is ykarel. Information about MeetBot at http://wiki.debian.org/MeetBot.15:01
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.15:01
opendevmeetThe meeting name has been set to 'neutron_ci'15:01
ykarelPing list: bcafarel, lajoskatona, slawek, mlavalle, mtomaska, ralonsoh, ykarel, jlibosva, elvira15:01
ykarelHi everyone, it's on IRC today15:01
ralonsohhello15:01
mtomaskao/15:01
elvirahi!15:01
lajoskatonao/15:02
slaweqo/15:02
haleybo/15:03
bcafarelo/15:03
ykarellet's start with the topics15:03
ykarel    lajoskatona to check vpnaas and bgpvpn UT failures15:03
ykarel#link https://bugs.launchpad.net/neutron/+bug/213748915:03
ykareli see both green now, thx lajoskatona 15:04
ykarelralonsoh to check sfc failures15:04
ralonsohfixed and merged15:04
ykarel#link https://bugs.launchpad.net/neutron/+bug/213763315:04
ykarelthx ralonsoh 15:04
ykarelralonsoh to check https://bugs.launchpad.net/neutron/+bug/2137467 this with jakub15:04
opendevreviewLajos Katona proposed openstack/neutron-fwaas master: CI: Add required_projects to unit test jobs  https://review.opendev.org/c/openstack/neutron-fwaas/+/97307915:04
ralonsohstill in progress15:05
ralonsohbut the issue has been identified 15:05
ykarelthx jlibosva ralonsoh 15:05
ykarel#link slawek to check fwaas functional job failure15:05
ykarel#link https://bugs.launchpad.net/neutron/+bug/213748015:05
ykarelthis too fixed, but have a new issue but can see next15:06
ykarelthx slaweq 15:06
ykarel#topic Stable branches15:06
ykarelall good as per periodic and not much activity with reviews15:06
ykarelbcafarel, anything else to add?15:06
bcafarelyes, it was mostly quiet (and OK on the few ones that went in)15:06
ykarels/reviews/backports15:06
ykarelok thx15:07
ykarel#topic Stadium projects15:07
ykareldynamic-routing still broken https://zuul.openstack.org/buildset/15ea550c13584730a51dc3576568702a15:07
lajoskatonaquite good compared to last weeks15:07
ykarelKnown one https://bugs.launchpad.net/neutron/+bug/213063115:07
lajoskatonayes that one is is red15:07
lajoskatonafor fwaas I think it is only the required-projects missing, see my patch: https://review.opendev.org/c/openstack/neutron-fwaas/+/97307915:07
ykarelyes better this week with all those fixes merged15:07
ykarelneutron-fwaas tox jobs https://zuul.openstack.org/buildset/068ba76c76da4d9185ac8635a80154a415:07
ykarelbroken with https://review.opendev.org/c/openstack/neutron-fwaas/+/97219515:07
lajoskatonaI hope now I have the yaml correct15:07
ykarel    either need neutron release or job to use unreleased neutron15:08
ralonsohwe have a patch to release beta 28.0.0.b115:08
ralonsohhttps://review.opendev.org/c/openstack/releases/+/97245715:08
lajoskatonayeah, that can be ok also, or my patch with required-projects15:08
ralonsoh^^ this one has the project_id changes and should be used15:08
ykarelyes some issue there15:08
lajoskatonaralonsoh: thanks15:08
ykarelbut i couldn't get how the job was working on check but not on periodic15:09
ykarelin periodic neutron was not included as required-projects while in check it does but couldn't spot where thats included 15:09
lajoskatonaI didn't check but my theory is that the template works differently15:09
ralonsohykarel, that should be the reason, if included in required-projects, should use master by defaultr15:10
lajoskatonaI mean using openstack-python3-jobs-neutron in template list, or just listing jobs like openstack-tox-py312-with-oslo-master15:10
ykarelralonsoh, but i couldn't find required-projects: neutron at all in check jobs running against neutron-fwaas15:11
ralonsohthat's weird, I'll check it later15:11
ralonsohin any case, we should bump the tag of neutron in the stadium projects15:11
ykarel+115:12
ykarelok let's move15:12
ykarel#topic Rechecks15:12
ykarelnot much activity, and recheck wise not bad15:12
ykarelwe had zuul outage last week and that resulted in couple of rechecks along with some intermittent known issues15:12
ykarelbare recheck wise 1/28 rechecks, so that good too15:13
ykarel#topic fullstack/functional15:13
ykarelfor this we seeing timeouts, but i see ralonsoh jlibosva looking into optimizing the tests15:13
ykarelso hopefully that will help15:14
ykarel#topic Periodic15:15
ykarel    broken https://zuul.openstack.org/builds?job_name=propose-translation-update&project=openstack/neutron15:15
ykarelRecent zuul upgrade/secret rotate broke this, https://review.opendev.org/c/openstack/project-config/+/973029 should probably fix, tomorrows run will confirm it15:15
ykarel#topic Grafana15:15
ykarelhttps://grafana.opendev.org/d/f913631585/neutron-failure-rate15:15
ykarellet's have quick look here too15:16
ykarelthe spikes on 8/9 are related to the zuul/secret issue15:18
ykarelwhich were unblocked by https://review.opendev.org/c/openstack/project-config/+/97282815:18
ykarelapart from that looks fine, some other spikes likely patch specific15:19
ykarelanyting to add?15:19
ralonsohno thanks15:19
lajoskatonanothing from me15:19
mtomaska.15:19
bcafarelall good :)15:19
ykarel#topic On Demand15:20
ykarelanything else you would like to raise?15:20
ralonsohnope15:20
haleybi had a qustion about a patch15:20
haleyb#link https://review.opendev.org/c/openstack/neutron/+/97273415:20
haleybi didn't quite understand your comment15:20
ykarelsure looking15:20
haleybor feel free to look after meeting, the series was all blocked with the zuul issues last week15:21
ykarelhaleyb, for periodic jobs irrelevant-files are irrelevant15:21
ykarelie. jobs always runs15:22
haleyband for experimental?15:22
ykarelfor experimental they are relevant, but considering experimental as on-demand so we should be fine even if irrelevant-files not set15:23
ykarelbut can continue over patch itself15:23
ykarelthx everyone for joining15:24
haleybsure15:24
ykarellet's close early and have everyone almost 35 minutes back15:24
ykarel#endmeeting15:24
opendevmeetMeeting ended Mon Jan 12 15:24:25 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:24
opendevmeetMinutes:        https://meetings.opendev.org/meetings/neutron_ci/2026/neutron_ci.2026-01-12-15.01.html15:24
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/neutron_ci/2026/neutron_ci.2026-01-12-15.01.txt15:24
opendevmeetLog:            https://meetings.opendev.org/meetings/neutron_ci/2026/neutron_ci.2026-01-12-15.01.log.html15:24
mtomaskao/15:24
ralonsohbye!15:24
lajoskatonao/15:24
elviraI realized tap as a service did not have it's ovn tests added on the .zuul yaml so I uploaded a tiny patch fixing that in case anyone wants to check :) https://review.opendev.org/c/openstack/tap-as-a-service/+/97305715:24
ralonsoh+115:25
slaweqo/15:25
slaweqthx for the patch elvira 15:25
elviranp!15:25
ralonsohelvira, btw, if you want, you can also remove the wsgi job15:26
slaweqI added comment there15:26
ralonsohnow we only have wsgi job, so one of them is duplicated (neutron-tempest-plugin-tap-as-a-service, neutron-tempest-plugin-tap-as-a-service-wsgi)15:26
slaweqmaybe we should add that job to gate queue also, wdyt?15:26
ykarel+115:27
ykarelto add to gate15:27
elviraoh sure, I will propose a new patchset then15:28
elviraAlthough I just saw gate jobs are starting because of the +215:28
ykareli removed that already after reading slawek comment15:28
opendevreviewLajos Katona proposed openstack/neutron-fwaas master: CI: Add required_projects to unit test jobs  https://review.opendev.org/c/openstack/neutron-fwaas/+/97307915:28
ralonsohelvira, don't worry, push a new one15:29
elviraoh ok awesome, I will make the change right now15:29
elvirathanks everyone!!15:29
ralonsohpffff and we also have p312 jobs still running there with master branch (openstack-tox-py312-with-oslo-master)15:29
ralonsohhaleyb, ^ we would need a document not to loose any project when migrating the jobs15:30
haleybi have a patch for neutron at least for that, maybe others i;ll have to look15:30
ralonsohfor example, when changing the python versions15:30
haleybralonsoh: yes, agreed15:30
ralonsohI'll open a LP, just to have a reference15:30
jlibosvaralonsoh: I forgot how the functional tests are supposed to work :) What is the openstack_citest database for?15:31
ralonsohjlibosva, just for FTs, it is created only for that15:31
ralonsohjlibosva, if you have a devstack env, you can change the mysql user and database15:32
jlibosvaralonsoh: right but then we create a new database per each test15:32
ralonsohyes, that's also the user15:32
jlibosvanot user, database15:32
ralonsohahh, just for checking15:32
ralonsoh    if mysql openstack_citest > /dev/null 2>&1 < /dev/null; then15:33
ralonsoh        echo_summary "DB config appears to be complete, skipping."15:33
ralonsoh        return 015:33
ralonsoh    fi15:33
jlibosvaso the DB itself is not used at alL?15:34
ralonsohnope15:34
jlibosvaand it's been always like that or it's changed?15:34
opendevreviewElvira García Ruiz proposed openstack/tap-as-a-service master: Update zuul jobs and gates definitions  https://review.opendev.org/c/openstack/tap-as-a-service/+/97305715:34
ralonsohjlibosva, I think that was the only goal of this DB15:35
ralonsohas you mentioned, each test creates its own DB15:35
ralonsohjlibosva, the point is that the loop https://review.opendev.org/c/openstack/neutron/+/972961/1/neutron/tests/unit/testlib_api.py#169 is taking around 8 seconds in my testing VM15:46
ralonsohsame as the `create_all` call15:46
ralonsoh(at least in my env)15:46
jlibosvaralonsoh: yeah, that's why I thought the DDL file might be a dead end15:46
ralonsohjlibosva, I've tested also using shell commands (mysqldump, mysql). The time is similar. The issue is the engine loading the DDL statements, regardless of the method (mysql cli, sqlalchemy)16:16
ralonsohjlibosva, I need to leave now but the fastest alternative seems to backup directly the file16:16
ralonsohthat would be the fastest way16:16
opendevreviewMichel Nederlof proposed openstack/neutron master: Keep virtual ports unbound when binding  https://review.opendev.org/c/openstack/neutron/+/97311516:24
opendevreviewMerged x/whitebox-neutron-tempest-plugin master: Configurable setup hosts information  https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/97203316:56
opendevreviewMerged openstack/neutron master: Add ``ModuleNotFoundError`` to ``get_api_worker_id``  https://review.opendev.org/c/openstack/neutron/+/97268717:18
opendevreviewJakub Libosvar proposed openstack/neutron master: func: Do not teardown DB if not dirty  https://review.opendev.org/c/openstack/neutron/+/97298517:43
opendevreviewJakub Libosvar proposed openstack/neutron master: DNM: Trigger functional tests  https://review.opendev.org/c/openstack/neutron/+/97298617:43
opendevreviewsean mooney proposed openstack/os-vif master: Add TAP device pre-creation support for OVS/OVN  https://review.opendev.org/c/openstack/os-vif/+/97123118:06
sean-k-mooneyralonsoh: https://review.opendev.org/c/openstack/nova/+/973149 works for booting a vm but i still need to test live migration with this chagne18:31
sean-k-mooneyill add a secodn patch that depens on the neutron change and updates one of the jobs tomorrow after is add a second compute and test this locally.18:32
opendevreviewJakub Libosvar proposed openstack/neutron master: func: Do not teardown DB if not dirty  https://review.opendev.org/c/openstack/neutron/+/97298518:40
opendevreviewJakub Libosvar proposed openstack/neutron master: DNM: Trigger functional tests  https://review.opendev.org/c/openstack/neutron/+/97298618:40
opendevreviewMerged openstack/neutron master: bgp: Introduce BGP chassis commands  https://review.opendev.org/c/openstack/neutron/+/96042219:13
opendevreviewJakub Libosvar proposed openstack/neutron master: func: Do not teardown DB if not dirty  https://review.opendev.org/c/openstack/neutron/+/97298519:50
opendevreviewJakub Libosvar proposed openstack/neutron master: DNM: Trigger functional tests  https://review.opendev.org/c/openstack/neutron/+/97298619:50
opendevreviewBrian Haley proposed openstack/neutron master: Update jobs to use the periodic zuul job templates  https://review.opendev.org/c/openstack/neutron/+/97273421:30
opendevreviewBrian Haley proposed openstack/neutron master: Update jobs based on testing runtime for 2026.1  https://review.opendev.org/c/openstack/neutron/+/97273421:38
*** jlibosva is now known as Guest3571621:50
opendevreviewMerged openstack/neutron stable/2025.1: [stable-only 2025.1] Initialize segment_network_range in non-WSGI server  https://review.opendev.org/c/openstack/neutron/+/97266922:45
opendevreviewBrian Haley proposed openstack/neutron master: Use project_id key in port plugin code  https://review.opendev.org/c/openstack/neutron/+/97298223:25

Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!