Thursday, 2026-08-06

« Wednesday, 2026-08-05 Index Friday, 2026-08-07 »
opendevreviewMerged openstack/watcher-dashboard master: Improvements to Playwright artifacts  https://review.opendev.org/c/openstack/watcher-dashboard/+/99606210:04
opendevreviewMerged openstack/watcher-dashboard master: Register test cleanup before assertions can fail  https://review.opendev.org/c/openstack/watcher-dashboard/+/99942911:52
opendevreviewDavid proposed openstack/watcher-dashboard master: Add Playwright tests for filter functionality  https://review.opendev.org/c/openstack/watcher-dashboard/+/99813112:36
sean-k-mooneydviroel: amoralej_ so regarding the playwright tests12:38
sean-k-mooneywe are currently still capturing thet videos in the jobs12:39
sean-k-mooneyhttps://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_d06/openstack/d06bc0576ca84ca1b7eaef4933a03f4b/controller/logs/playwright/videos/12:39
sean-k-mooneywe shoudl only do that for failing tests.12:39
sean-k-mooneynot all tests12:39
sean-k-mooneyi think we shoudl be doign that for the traces as well12:39
sean-k-mooneyhttps://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_d06/openstack/d06bc0576ca84ca1b7eaef4933a03f4b/controller/logs/playwright/traces/index.html12:39
sean-k-mooneythey are pretty large12:40
sean-k-mooneythe screenshot we proably can keep if we want 12:40
sean-k-mooneybtu the selenium tests for horizon for example only keep the detailed artifact when a test fails to save log space in the ci12:40
dviroelhum, yeah you are right, they are large and not so useful when tests are passing12:41
sean-k-mooneyso this is configurable in playwright12:41
sean-k-mooneyand we may have the envar in place to contol it12:41
dviroelscreenshot for passing jobs should be enough, at least I am only looking at screenshots recently12:41
dviroelack12:41
sean-k-mooneybut the job is not configred to do this12:41
sean-k-mooneyhttps://playwright.dev/docs/videos#record-video12:43
sean-k-mooneyso we can pass retain-on-failure12:43
sean-k-mooneyi think12:43
dviroelif we can control that with a playwright var, should be easier then12:43
sean-k-mooneyhttps://github.com/openstack/watcher-dashboard/blob/master/tox.ini#L6512:43
sean-k-mooneyi thik we are pulling the value form the env var12:44
sean-k-mooneybut we shoudl confirm12:44
chandankumarhttps://github.com/openstack/watcher-dashboard/blob/master/.zuul.yaml#L3412:44
sean-k-mooneythat not working then12:45
chandankumaryes, that part might be broken for video as well as traces12:45
sean-k-mooneyhttps://github.com/openstack/watcher-dashboard/blob/e0d689b84480c3b1c1777cd26fc7c2a2ec1022ab/watcher_dashboard/test/integration/playwright_config.py#L192-L20512:46
dviroelhttps://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_d06/openstack/d06bc0576ca84ca1b7eaef4933a03f4b/tox/integration-playwright/2-commands%5B0%5D.log12:46
sean-k-mooneyso the code looks correct12:46
sean-k-mooneybut i wonder fi the envirotnm varibale is not defiend12:46
sean-k-mooney oh weird12:47
sean-k-mooneyso it is there12:47
sean-k-mooneyok well we shoudl look into this before adding more tests12:48
sean-k-mooneydavid does not seam to be on irc12:48
sean-k-mooneyoh chandankumar  this is the only use of get_video_mode12:49
chandankumarI will spin up a env and try to reproduce it12:49
sean-k-mooneyhttps://github.com/openstack/watcher-dashboard/blob/master/watcher_dashboard/test/integration/playwright_base.py#L107-L11312:49
sean-k-mooneyso we are not passing it to playright properly12:49
chandankumarhttps://github.com/openstack/watcher-dashboard/blob/master/watcher_dashboard/test/integration/playwright_base.py#L12212:52
chandankumarHere we are only passing trace mode no video mode12:52
sean-k-mooney well tracemode is not beeing used properly either12:53
chandankumaryes 12:53
sean-k-mooneyso apprently the fucntionalty is not nativly part of playwrite-pyton12:54
sean-k-mooneyits in the pytest wrapper12:54
sean-k-mooneyor in typescript12:54
sean-k-mooneychatgpt suggests https://paste.opendev.org/show/brFzvR9mN5YUkO0HpYnc/12:55
sean-k-mooneybasiclly we will need to do it ourslevs12:55
chandankumarhttps://playwright.dev/python/docs/videos12:55
chandankumaryes, I will dig into that12:55
sean-k-mooneyright there is no documetion fo this in the python veriosn12:55
sean-k-mooneyjust in typescript12:55
sean-k-mooneyso we will have to add clean up of the traces and videos to our basclass tearDown function when the test passes12:56
dviroelfolks, the watcher meetings will start in 3 minutes, please add your topics to the agenda https://etherpad.opendev.org/p/openstack-watcher-irc-meeting12:56
dviroels/3/4 :)12:57
sean-k-mooneyi think that will be simple enough we are creating a directory per test anyway12:57
sean-k-mooneyso we jsut delete that directory if it passes 12:57
sean-k-mooneycan we file a bug for that12:58
dviroel#startmeeting watcher13:00
opendevmeetMeeting started Thu Aug  6 13:00:36 2026 UTC and is due to finish in 60 minutes.  The chair is dviroel. Information about MeetBot at http://wiki.debian.org/MeetBot.13:00
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.13:00
opendevmeetThe meeting name has been set to 'watcher'13:00
dviroelhello all o/13:00
sean-k-mooneyo/13:01
amoralej_o/13:01
dviroelcourtesy ping: chandankumar morenod rlandy13:01
morenodo/13:02
dviroelhey o/13:02
dviroellet's start with today's meeting agenda13:02
dviroel#link https://etherpad.opendev.org/p/openstack-watcher-irc-meeting#L33 (Meeting agenda)13:02
dviroelfeel free to add your own topics to the agenda13:02
chandankumaro/13:02
dviroel#topic Announcements13:02
dviroelfirst announcement13:02
dviroelOpenStack Election13:03
dviroelthe combined  PTL/TC 2027.1 cycle Election started and nominations period will end at Aug 19, 2026 23:45 UTC13:03
dviroel#link https://governance.openstack.org/election/13:03
dviroelif you plan to submit a nomination for PTL or TC, pay attention to the deadline13:04
dviroelif  you want to know more, there are some links in the election page that explain the process13:05
dviroelso moving on..13:05
dviroelsecond announcement in the list13:05
dviroel2027.1 PTG Etherpad is up13:05
dviroeli just created it today, so we can start populating it with topic proposals13:06
dviroel#link https://etherpad.opendev.org/p/watcher-2027.1-ptg13:06
dviroelfeel free to start adding topics that you would like to discuss in the next PTG13:06
dviroelnext PTG is schedule for Oct 12-16 week13:07
dviroelregistrations are open in13:07
dviroel#link https://openinfra.org/ptg/13:07
dviroelI also updated out meeting etherpad with the new planning/ptg etherpad13:08
dviroelany other announcement folks?13:08
dviroelok, next 13:08
dviroel#topic Reviews13:08
dviroellets go through some of the changes listed in our review etherpad13:09
dviroelif you have any that is not listed there, feel free to add it13:09
dviroelyou can also create a topic there to group all related changes13:09
dviroeleasier for us to track when reviewing them13:09
dviroelfrom freatures list13:10
dviroel1. Openstacksdk migration13:10
dviroelI think that we merged most of the patches there, there as just a few related patches open13:10
dviroeljoan propose ironic migration too, but it is still in progress i guess13:10
dviroelwe will probably circle back for the next cycle, since Joan is on pto 13:11
amoralej_yes https://review.opendev.org/c/openstack/watcher/+/975498?usp=search is missing, i didn't prioritize it tbh13:11
amoralej_given that current retries work and i want to test that carefully given the issues in the past13:12
amoralej_although it'd be nice to leverage openstacksdk feature13:12
dviroelyeah, i need to recap everything again during the review13:12
dviroelack, it is a good change to review, not the top prio atm13:12
dviroelbut we could get it merged when possible13:13
dviroelalright13:13
dviroelnext feature is13:13
dviroel2. Audite Pipeline13:13
dviroels/Audite/Audit13:13
dviroelalso fixed in the etherpad :)13:13
dviroeli still need to get back to tempest change in audit template extension  - will do once i push the remaining work from audit pipeline feature13:14
dviroelfor the audit pipeline there are 4 changes ready to review, which already have been receiving feedbacks13:15
dviroelthanks amoralej_ and jgilaber13:15
dviroeltrying to update them as I also push the new changes13:15
dviroelI expect 2 to 3 new changes yet13:15
amoralej_dviroel, wrt https://review.opendev.org/c/openstack/watcher/+/995119 i sent some comments after your last update, i think13:16
dviroelone for the audit pipeline handle and planner(which may be 1 or 2), another for the API and docs13:16
sean-k-mooneyso 7-8 ish patches total13:16
sean-k-mooney+ teh tempest tests13:17
dviroelamoralej_: hum,yeah i missed, i will take a look afterwards, thanks13:17
sean-k-mooneyi hae not really spent much time looking at it since the spec, im not sure ill have much time to review it but if i can make some time i will13:17
dviroelyeah, and + watcherclient13:17
amoralej_wrt https://review.opendev.org/c/openstack/watcher/+/996849/ i'd like to hear other opinions about where to implement constraints, i'm not sure tbh13:17
dviroelamoralej_: sure, I added my version of it, based on what I think that is better. We can get some feebacks and decide yeah13:18
sean-k-mooneyso13:18
sean-k-mooneywe normally specifi foreignkey and unique constratis13:19
amoralej_my feeling is that doing it only in the object may be too high13:19
sean-k-mooneybtu we dont thend to use db triggers to reject things13:19
sean-k-mooneyif that is waht your asking13:19
sean-k-mooneywhich specifi cconstratit are you asking about13:19
amoralej_iiuc there are three levels, db , create_audit_template in dbapi and object13:19
amoralej_we are making audit_id and strategy_id optional in the actionplan objects13:20
amoralej_also audit_pipeline_id is optional13:20
amoralej_but there should be always one of them13:20
dviroelyeah to guarantee that in action_plan, that we always set a audit_id or a audit_pipeline_id13:20
amoralej_audit_pipelihne_id or goal_id13:20
amoralej_yeah, sorry audit_id13:21
sean-k-mooneyso we coudl have a db triger to enforce this but tha twoudl not be typical in openstack13:21
sean-k-mooneynormally we woudl do such a check in save or the db method not the the schema13:21
sean-k-mooney(save on the object)13:21
amoralej_yep, i think create_action_plan may be adequate13:21
dviroeli avoid adding logic to dbapi13:22
amoralej_dviroel, proposed it in the object https://review.opendev.org/c/openstack/watcher/+/996849/5/watcher/objects/action_plan.py13:22
sean-k-mooneyya it s really a object constratit13:22
amoralej_all action/changes on the actionplan go through the object, right?13:23
sean-k-mooneyyes13:23
sean-k-mooneythe db api is internal and behind the object interface13:24
dviroelti should13:24
amoralej_btw, there is also save method13:24
amoralej_we may implement the check there, in case something updates the object?, i think save can be used directly13:24
dviroeli may need additional check in save, since the object could be updated.. need to check13:24
sean-k-mooneysave is for updates to the object that has already been created in the db13:25
dviroelit is used to update the state13:25
amoralej_one last question about this13:25
amoralej_in the past both audit_id and strategy_id were mandatory13:25
sean-k-mooneyyes13:25
amoralej_so i understand the constraints should be audit_template_id or (audit_id AND strategy_id)13:25
sean-k-mooneyno13:25
sean-k-mooneyi think the pipeline id13:26
sean-k-mooneynot the audit_template_id13:26
amoralej_sorry13:26
dviroelyeah13:26
amoralej_ /forget13:26
dviroelI can additionally check the strategy tooi13:26
amoralej_ audit_pipeline_id or (audit_id AND strategy_id)13:27
sean-k-mooneydviroel: i woudl factgor out a validation helper method13:27
sean-k-mooneyand just reuse that in all the places we need to chack the state13:27
dviroelyep13:27
dviroelcan be improved13:27
sean-k-mooneyi think that is just save and create but its a good patthtern to get into13:28
dviroelyeah, thanks for the feedback 13:28
sean-k-mooneywe coudl acutlly have a valildate method on WatcherObject that is a noop13:28
amoralej_that was my main concern in that one13:28
sean-k-mooneythen just implemtn that on the objects as requried13:28
dviroelyeah it could be13:29
sean-k-mooneyi have a related but oftiopc comment fi we are dont with this?13:30
dviroelat least for the audit pipeline I think that we covered13:30
dviroelsean-k-mooney: go on13:30
sean-k-mooneyso we are never goign support plugabel ORMs in relaity 13:31
dviroelvery unlikely i think13:31
sean-k-mooneyso we can get rid of the indirection vai watcher/db/api.py and promote watcher/db/sqlachmey/api.py13:31
dviroelwe could I think13:32
sean-k-mooneythat just existing tech debt to my eyes13:32
sean-k-mooneyso we may want to tack that as a bug and do it next cycle13:32
dviroelnoted13:32
sean-k-mooneyit not super imporant13:32
dviroeli can create e LP to track13:32
dviroelyeap13:32
sean-k-mooneyjsut when you are doing db change like this it make it obvious13:33
sean-k-mooneywe are not realy getting an benifit form it13:33
dviroelyeha, we just duplicate things everytime13:33
sean-k-mooneyexactly13:33
dviroelack, thanks sean-k-mooney 13:33
sean-k-mooneyso out of scope of this work13:33
sean-k-mooneybut good to fix evcentulaly13:33
dviroelyes13:33
dviroel3. Functional Tests Infrastructure13:34
dviroelamoralej_: want to comment about?13:34
dviroelI know that I have to get back to your changes13:34
amoralej_yes13:34
dviroelI believe that the functional infra can be very interesting even for audit-pipeline and other tests13:35
dviroelamoralej_: sure13:35
amoralej_i was about to propose that13:35
sean-k-mooneyi think we were convering on the desgin13:35
sean-k-mooneyamoralej  already adress much of my feedback13:35
dviroelyes13:36
sean-k-mooneyi have not looked in a few days howver13:36
amoralej_so https://review.opendev.org/c/openstack/watcher/+/994845/4 and https://review.opendev.org/c/openstack/watcher/+/988389/11 have possitivie feedback13:36
amoralej_merging those will allow to start adding functional tests for api, etc...13:36
amoralej_not only api, also running the entire workflow13:36
amoralej_but with dummy i.e.13:36
amoralej_aboaut https://review.opendev.org/c/openstack/watcher/+/993352/ i moved out of flask13:37
sean-k-mooney+113:37
dviroelack13:37
amoralej_and was waiting for your feedback, i understand the ongoing approach is acceptable13:37
dviroeli think that is ready, i just need to review in details13:38
dviroeland vote13:38
sean-k-mooneyso right now you dont really ahve these all in one serise13:38
amoralej_once i get those approved, i will work in new versions of https://review.opendev.org/c/openstack/watcher/+/993543/ and https://review.opendev.org/c/openstack/watcher/+/993634/ to apply the same approach13:38
amoralej_will need to create objects for cinder, do helpers, etc...13:38
sean-k-mooneyas in in gerrit they are not all rembased into a singel stack13:38
amoralej_i cut in the nova one13:39
amoralej_the three i mentioned are13:39
sean-k-mooneyack but your goign to reabse the rest later right13:40
amoralej_yes13:40
sean-k-mooneyi think we can try and get the first 3 merged before the next meeting13:40
amoralej_i didn't want to start doing big refactor and adding features until getting approved the previous ones13:40
dviroelack13:40
sean-k-mooneyill see if i can review the first 3 today13:41
amoralej_also, those provide good value as allow to cover a good amount of the features13:41
amoralej_but yeah, as soon as i get them approved or merged, i will work on them13:41
dviroelack, thanks amoralej_  i will try to review them, together with your other fixes13:41
dviroelamoralej_: anything else?13:42
amoralej_i don't think so13:42
dviroelthe 4. is Preemtible Instances13:42
dviroelwe are waiting winicius to update the patch13:42
dviroeli will check with him later13:43
dviroelto see if he has any questions or need help with something13:43
dviroelit is all for features13:44
dviroelfrom bug fixes i think that worth mentioning13:44
dviroelAdd allocation-based capacity checks to vm_workload_consolidation13:44
dviroel#link https://review.opendev.org/c/openstack/watcher/+/993088/13:44
dviroeland its parent change13:44
dviroel998883: Add node resource allocation cache to ModelRoot 13:45
dviroel#link https://review.opendev.org/c/openstack/watcher/+/99888313:45
dviroeli want to review with time, the resource allocation cache13:45
dviroelit already has an approval from joan13:46
dviroelwhich is great13:46
dviroelany comment in these ones?13:46
amoralej_not much13:47
dviroelnext in the list are dashboard improvements13:47
dviroel976594: Add Playwright integration test for skip action workflow | https://review.opendev.org/c/openstack/watcher-dashboard/+/97659413:47
dviroel997716: Add Playwright tests for audit template scope validation | https://review.opendev.org/c/openstack/watcher-dashboard/+/99771613:47
dviroel998324: Fix babel configs and add translations tox target | https://review.opendev.org/c/openstack/watcher-dashboard/+/99832413:47
sean-k-mooneyamoralej_: i left comment on https://review.opendev.org/c/openstack/watcher/+/994845 fyi we shoudl not be reloadign the config on ever audit trigerr that is a seperate prexistign bug13:48
sean-k-mooneyanyway back to dashbaord13:48
sean-k-mooneythe babel supprot is intersting13:49
sean-k-mooneywe shoudl fix it13:49
sean-k-mooneybut i honestly dont think this will ever really be translated13:49
dviroelyeah, chandan proposed but i didn't have time to properly review it13:49
sean-k-mooneybut sure we shoudl enabel it to be13:49
amoralej_i think translation will require time13:49
amoralej_dunno if we can take care of it now13:50
amoralej_i mean, the entire functionality, not only that patch13:50
chandankumartranslation review ill help to publish the pot files to translation server13:50
chandankumaronly releasenotes pot file exist for watcher dashboard, not user facing string13:50
amoralej_yes, that's correct, it's required step13:50
sean-k-mooneyi didn tthink we trasnlated release notes13:51
sean-k-mooneybut ok13:51
amoralej_but will need to invest some time in checking that code is fine, and after that, getting the actual translations13:51
sean-k-mooneyright but i dotn think the later will happen13:51
sean-k-mooneyunless ye also decied to start doign translations yoruself13:51
dviroelnot sure how much effort to get a translation there13:52
sean-k-mooneyi mean redhat does not even do that in our downstream product13:52
dviroelwe could have spanish and portuguese :)13:52
sean-k-mooneybut ya i dont really have an issue with this supprot being added as long as it not a maintnace burden logn term13:52
chandankumardoing translation is hard13:52
dviroelyeah, lets follow up that later, we should get translation enabled at least13:53
dviroelthanks again for starting the effort chandankumar 13:53
morenodmost of the modules, only menu items are translated. for us, it is about 10-15 items13:53
chandankumarif you want to translate you can use lokalize, https://apps.kde.org/lokalize/ i used it long time back13:54
dviroelinteresting13:55
sean-k-mooneywell technial all exctpiopn can be translated13:55
sean-k-mooneybut if we are doing this13:55
sean-k-mooneyit means all our playright test shoudl only use the ids13:56
sean-k-mooneynever the names of buttons13:56
sean-k-mooneyectra13:56
sean-k-mooneywe shodul prefer the ids in general13:56
amoralej_i think it's doable, it's just a matter of time and priority13:56
dviroelyeah13:56
sean-k-mooneybut the playwrithgt test shoudl work without change even if you have a non default lanaugage in use13:56
amoralej_there are not AI bots creating translations ? :) 13:56
sean-k-mooneythat a contoversial topic13:57
dviroelalright, any other review that you folks want to highlight here?13:57
amoralej_i guessed so13:57
sean-k-mooneywe coudl do that13:57
amoralej_don't want to open the discussion :) 13:57
sean-k-mooneybut ther eis push back13:57
dviroel#topic Bugs13:57
* dviroel is moving on13:57
dviroel#link https://bugs.launchpad.net/watcher/+bug/2162057 13:57
dviroelCannot combine --detail parameter on openstack optimize audit list13:58
dviroelthis one was already in progress so I assigned to morenod :)13:58
morenodyep, I created a fix https://review.opendev.org/c/openstack/watcher/+/99909413:58
dviroelthere is a patch for it13:58
dviroelyeah that one13:58
morenodfail is not on the cli, it is on the api, so it affects api and horizon13:59
dviroeli see13:59
dviroelhttps://docs.openstack.org/api-ref/resource-optimization/#list-audit-detailed13:59
dviroelok, last minute14:00
dviroelthanks for proposing a fix morenod 14:00
dviroelwe will get our eyes on it14:00
dviroeli will add to the review etherpad too14:00
dviroelnext topic14:00
dviroel#topic Volunteers to chair next week14:00
dviroelI can go, unless someone want to chair14:00
dviroelok, i will chair next week14:01
dviroel:)14:01
dviroellet's wrap up for today14:01
dviroelwe will meet again next week14:01
dviroelthank you all for participating14:01
dviroel#endmeeting14:01
opendevmeetMeeting ended Thu Aug  6 14:01:54 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)14:01
opendevmeetMinutes:        https://meetings.opendev.org/meetings/watcher/2026/watcher.2026-08-06-13.00.html14:01
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/watcher/2026/watcher.2026-08-06-13.00.txt14:01
opendevmeetLog:            https://meetings.opendev.org/meetings/watcher/2026/watcher.2026-08-06-13.00.log.html14:01
opendevreviewDouglas Viroel proposed openstack/watcher master: Add AuditPipeline execution handler and cascade execution mode  https://review.opendev.org/c/openstack/watcher/+/100003119:13
« Wednesday, 2026-08-05 Index Friday, 2026-08-07 »

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!