Monday, 2025-11-17

opendevreviewDavid proposed openstack/watcher master: [DNM] Testing nodeset with three nodes (two computes + 1 controller)  https://review.opendev.org/c/openstack/watcher/+/96733110:36
opendevreviewDavid proposed openstack/watcher-tempest-plugin master: Add test for skipped actions  https://review.opendev.org/c/openstack/watcher-tempest-plugin/+/96686010:59
sean-k-mooneyjgilaber: would you mind taking a look at https://review.opendev.org/c/openstack/watcher/+/966710 again when you have time11:36
jgilabersure, I +wed it, I thought I already did last week 11:39
sean-k-mooneyno worries i just saw that takashi responede to my question so i said i woudl ping you to take a look again11:45
sean-k-mooneyjgilaber: while you here can you appove https://review.opendev.org/c/openstack/watcher/+/966849 to unblock the ci11:46
sean-k-mooneynot super urgent but i need to back port that to all stabel brnahces11:46
jgilabersure, done11:48
sean-k-mooneythanks, the implemention of the requirements check job changes about a week ago. im not entirly sure if/when it will break stable but for such a trivial change there is no reason to wait to backport it IMO11:50
opendevreviewDouglas Viroel proposed openstack/watcher master: Consolidate and improve Zuul CI job definitions  https://review.opendev.org/c/openstack/watcher/+/96694211:58
opendevreviewMerged openstack/watcher master: ruff: Use more specific name to enable pyupgrade rule  https://review.opendev.org/c/openstack/watcher/+/96671012:12
opendevreviewMerged openstack/watcher stable/2025.2: use the correct package name for microversion-parse  https://review.opendev.org/c/openstack/watcher/+/96684912:26
opendevreviewsean mooney proposed openstack/watcher stable/2025.1: use the correct package name for microversion-parse  https://review.opendev.org/c/openstack/watcher/+/96736013:05
dviroelsean-k-mooney: hey, not sure what to do with this patch after: https://review.opendev.org/c/openstack/watcher/+/966226/4#message-84a20635258bb9b696a84145700784e0553957c413:33
dviroelmaybe investigate a bit more, or proceed with mergind and considering this threading job as non-voting for now13:33
opendevreviewDavid proposed openstack/watcher master: [DNM] Testing nodeset with three nodes (two computes + 1 controller)  https://review.opendev.org/c/openstack/watcher/+/96733113:41
opendevreviewDavid proposed openstack/watcher-tempest-plugin master: Add test for skipped actions  https://review.opendev.org/c/openstack/watcher-tempest-plugin/+/96686013:46
sean-k-mooney dviroel  the threading job seams stable13:50
sean-k-mooneydviroel: the greade job is the one that is failing13:50
opendevreviewDavid proposed openstack/watcher-tempest-plugin master: Add test for skipped actions  https://review.opendev.org/c/openstack/watcher-tempest-plugin/+/96686013:50
dviroelsean-k-mooney: i think that you are mixing, I am talking about the eventlet patch13:51
dviroelopenstack-tox-py312-threading https://zuul.opendev.org/t/openstack/build/2165a83bfad74dd79aecf53291329721 : FAILURE in 5m 23s13:51
sean-k-mooney oh ya i clicked on the wrong link13:52
sean-k-mooneyso test_execute_with_action_plan_cancel is faling is it alwasy that tes?t13:53
sean-k-mooneylooks like you had at least one clean run but that looks like the same nested db connection issue13:54
dviroelacually this is the first time that it fails after we moved to serial engine in taskflow13:55
sean-k-mooneysqlite3.OperationalError: cannot commit - no transaction is active13:56
sean-k-mooneyso we had to many and now to few13:56
sean-k-mooneythis is coming form the test fixture13:57
sean-k-mooneywell maybe 13:58
sean-k-mooneyoslo_db.exception.DBNonExistentTable: (sqlite3.OperationalError) no such table: action_plans13:58
dviroelhum slightly different13:58
sean-k-mooneyso it looks like the test fixture tried to use the cached scheme13:58
dviroelyeap13:58
sean-k-mooneyand that failed in setup but we still tried to use the db connection13:58
sean-k-mooneyso we ened up trying to use an empty db13:59
sean-k-mooneyso this to me looks liek we have a bug in how we are creatin ght db in the fixture in threaded mode13:59
sean-k-mooneylets see of we can repoduce that reliable locally if so we can likely fix that14:00
dviroelack, maybe this time is actually from unit test side, let me try it locally 14:04
sean-k-mooneyit passed intially but i have it running in a loop with tox -e py3-threading -- --until-failure14:18
sean-k-mooneyok it failed on loop 414:19
sean-k-mooneythis time in a diffent test watcher.tests.applier.workflow_engine.test_default_workflow_engine.TestDefaultWorkFlowEngine.test_execute_with_two_actions14:20
sean-k-mooneysqlite3.OperationalError: not an error ... ok then why did you raise :)14:20
dviroellol14:21
dviroelyeah, in a loop, it failed on test_execute_with_one_action for me14:21
dviroelbut not fixture setUp error mention14:21
sean-k-mooneyso its the same root cause npot table found sqlite3.OperationalError: no such table: action_plans14:22
dviroelyes14:22
sean-k-mooneyi have not seen the transaction error yt14:22
sean-k-mooneybut its proably there too14:22
sean-k-mooneyso all our unit test that need db access (which make htenm not really unit tests...) use https://github.com/openstack/watcher/blob/master/watcher/tests/db/base.py14:26
sean-k-mooneynova has its onw fixtre because of legacy but looking at placmente ther eare once form oslo deb that we can use https://github.com/openstack/placement/blob/master/placement/tests/fixtures.py#L3014:28
sean-k-mooneyhttps://github.com/openstack/oslo.db/blob/master/oslo_db/sqlalchemy/test_fixtures.py#L35014:30
sean-k-mooneydviroel: it looks liek that supprot schea caching nativly 14:31
sean-k-mooneyim going to quickly jsut try swappign to that and see fi that is enough14:31
*** haleyb|out is now known as haleyb14:32
* dviroel looking14:33
sean-k-mooneyim going to leave that running while i get coffee but i think its working. this woudl be a nice tech debt reduction anyway so i my rebase this on master and submit it sepreeatly14:51
sean-k-mooneydviroel: i cant prove a nevitive but its still runing after 10 mins so ill give it 20-30 and push that if it is still good15:00
dviroelsean-k-mooney: ack, thanks, I will take a look then15:00
* dviroel lunch in meantime15:01
opendevreviewDavid proposed openstack/watcher master: [DNM] Testing nodeset with three nodes (two computes + 1 controller)  https://review.opendev.org/c/openstack/watcher/+/96733115:19
opendevreviewsean mooney proposed openstack/watcher master: Adds support for threading mode in applier  https://review.opendev.org/c/openstack/watcher/+/96622615:21
opendevreviewsean mooney proposed openstack/watcher master: [DB] adopt the AdHocDbFixture from oslo  https://review.opendev.org/c/openstack/watcher/+/96739015:21
sean-k-mooneyim going to continue leaving that run in the background during my next meeting but no failure after 20 mins so looking good. we can see how it does in ci15:22
opendevreviewDavid proposed openstack/watcher-tempest-plugin master: Add test for skipped actions  https://review.opendev.org/c/openstack/watcher-tempest-plugin/+/96686015:32
opendevreviewsean mooney proposed openstack/watcher master: [DB] adopt the AdHocDbFixture from oslo  https://review.opendev.org/c/openstack/watcher/+/96739016:47
opendevreviewsean mooney proposed openstack/watcher master: Adds support for threading mode in applier  https://review.opendev.org/c/openstack/watcher/+/96622616:47
sean-k-mooneyfixed the pep8 issue16:48
dviroelack16:50
opendevreviewMerged openstack/watcher-specs master: Add functional test infrastructure spec  https://review.opendev.org/c/openstack/watcher-specs/+/96329919:39

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