Thursday, 2026-03-12

opendevreviewchandan kumar proposed openstack/watcher master: Add watcher-grenade-skip-level-always job  https://review.opendev.org/c/openstack/watcher/+/97913704:02
opendevreviewchandan kumar proposed openstack/watcher master: Add watcher-grenade-skip-level-always job  https://review.opendev.org/c/openstack/watcher/+/97913705:05
jgilaberHi everyone! IRC meeting will start in ~30 minutes, feel free to add any topics to the agenda https://etherpad.opendev.org/p/openstack-watcher-irc-meeting#L2411:29
jgilaber#startmeeting watcher12:00
opendevmeetMeeting started Thu Mar 12 12:00:15 2026 UTC and is due to finish in 60 minutes.  The chair is jgilaber. Information about MeetBot at http://wiki.debian.org/MeetBot.12:00
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.12:00
opendevmeetThe meeting name has been set to 'watcher'12:00
jgilaberhi o/ - who's around for today's meeting?12:00
opendevreviewAlfredo Moralejo proposed openstack/watcher master: DNM - PoC nova and prometheus standalone emulators  https://review.opendev.org/c/openstack/watcher/+/98025712:01
jgilabercourtesy ping: dviroel amoralej sean-k-mooney chandankumar morenod rlandy12:01
amoralejo/12:01
chandankumaro/12:02
dviroelo/12:02
jgilaberas usual, feel free to add your own topics to the agenda12:02
jgilaber#link https://etherpad.opendev.org/p/openstack-watcher-irc-meeting#L2412:02
jgilaberwe'll give a couple more minutes to see if more people will join12:03
jgilaberotherwise we can start12:03
rlandyo/12:04
opendevreviewAlfredo Moralejo proposed openstack/watcher master: DNM - PoC nova and prometheus standalone emulators  https://review.opendev.org/c/openstack/watcher/+/98025712:04
jgilaberok, I think we can get started with the first topic12:04
jgilaber#topic Anouncements12:05
jgilaberwe're in the RC1 target week12:05
jgilabertomorrow RC1 should be cut12:05
jgilaberand it looks like we've merged all the pending bug fixes (except for one that is currently in gate)12:06
jgilabergreat job everybody! and thanks for the reviews12:06
dviroelyeah, i saw this last one missing one vote12:06
dviroelI think that should be our last one12:06
jgilaberyes, I checked and I did not see any other bug fix proposed12:07
jgilaberthat's all I wanted to announce, any comment on this topic? any other announcement?12:07
sean-k-mooneyo/12:08
jgilaberif not, we can move to the next topic12:08
jgilaber#topic Reviews12:09
jgilaberI added to the list a couple of patches I have for the tempest plugin12:09
jgilaber#link https://review.opendev.org/c/openstack/watcher-tempest-plugin/+/96270212:09
jgilaber#link https://review.opendev.org/c/openstack/watcher-tempest-plugin/+/96355912:09
jgilaberthey add more tempest coverage for volume migrations in zone migration12:09
jgilabernot urgent but I think they are nice to have12:10
jgilaberany question/comment on these patches?12:10
dviroeldont think so, just need to take a look again12:11
jgilaberack thanks dviroel 12:11
jgilaberlet's move to the next one from chandankumar then12:11
dviroelare they protected with config options12:11
dviroelright? because of the fixes that still need to merge in stable branches?12:12
jgilaberyes, they are under the 'run_zone_migration_storage_tests' flag12:12
jgilabercurrently only set on master12:12
sean-k-mooneyright we shoudl not have test that require backport to pass enabeld by default12:12
sean-k-mooneyjgilaber: is that new12:12
jgilaberthe flag? not really, we added it during this cycle12:13
jgilaberdon't remember when exactly12:13
sean-k-mooneywhen we have feature gates like this we cant reuese it12:13
sean-k-mooneyif it was this cycle its ok12:13
sean-k-mooneybut we cant use that same feature flag for this next cycle12:13
sean-k-mooneyi.e. if we have anohter storage fix12:13
jgilaberack I did not know that12:14
sean-k-mooneytempest test are ment to work with alsl stable branches at all times12:14
sean-k-mooneyif they need specififc featuer to work the entierh need to check the api microverison or have a flag12:14
jgilaberI double checked we added it on December12:14
jgilaber#link https://review.opendev.org/c/openstack/watcher-tempest-plugin/+/96924412:14
dviroeloh ywah, the extra flags12:15
sean-k-mooneyack we likely will want to have a set of bug flags at some point that we will eventuly deprecate and remove or have a better long term solution for this12:15
sean-k-mooneyi guess we coudl ungate tests our of extra over time12:17
sean-k-mooneythat woudl be an option12:17
sean-k-mooneyi.e. after the final stable branch is patched move the option but when to do that is a bit tricky12:17
jgilaberyes, we have quite a few patches in the stable branches needed for that 12:18
jgilaberok, anything else on the topic?12:18
sean-k-mooneynot for now we can liekly dicuss this on the review12:19
jgilaberthanks! let's move on12:19
jgilaber#link https://review.opendev.org/c/openstack/watcher/+/97913712:19
jgilaberchandankumar, do you want to highlight something about the patch?12:19
chandankumaryes12:19
chandankumarIt adds watcher-grenade-skip-level-always job which will test upgrade from 2025.1 to 2026.1 (N-2 to N) release12:20
chandankumarIt is an additional upgrade job apart from normal watcher-grenade job which tests N-1 to N release12:20
chandankumarIt would be good to have it in our pipeline.12:21
chandankumarPlease have a look, thank you!12:21
dviroelthis is very importanto to have yes12:21
jgilaberthanks, that's great to have, I'll add it to my review list12:21
chandankumarwe can see the logs here https://c50701f61b79b2f93421-0d0b23b7f62d479822a1b52b605369e3.ssl.cf2.rackcdn.com/openstack/5f7f8c97ef5e4b77be2f352bac1f81cd/controller/logs/grenade.sh_log.txt12:21
chandankumarthat's it from my side on this review12:22
jgilaberthanks! any question?12:22
jgilaberif not we can move on to triage some bugs12:23
sean-k-mooneyfor what its worht im ok with the curernt patch12:23
sean-k-mooneyi was +2 on it yesterday and i resproed it just now after the description update12:23
sean-k-mooneyit would be good to include this in rc112:23
sean-k-mooneyso that its there when the stabel branch is created12:24
sean-k-mooneyalthoguh we can cherry pick it if needed12:24
chandankumarthanks sean-k-mooney!12:24
sean-k-mooneyill also note tha we technially shoudl eb doing this on the watcher-dashboard too12:24
jgilaber+1, I'll try to review today12:24
sean-k-mooneybut that is less imporant12:24
jgilaberack, let's move to the next topic 12:26
jgilaber#topic Bugs12:26
dviroelyeah, we lack testing in dashboard, we should solve that in the next cycle :) 12:26
jgilaberI did not see any untriaged bugs in watcher, but there were two in the tempest plugin12:26
jgilaber#link Add more tempest coverage for volume migrations in zone migration12:26
jgilabersorry about that12:26
jgilaber#undo12:27
opendevmeetRemoving item from minutes: #link Add12:27
jgilaber#link https://bugs.launchpad.net/watcher-tempest-plugin/+bug/213934412:27
jgilaberthat's the one :)12:27
jgilaberon first glance that looks more like a watcher feature than a tempest plugin bug 12:27
amoralej_i reported that as part of a previous bug discussion iirc but we ended up agreen that it's  feature12:28
amoralej_as it was documented in the spec12:28
jgilaberok to move it to the watcher tracker then? or close it altogether and discuss the possible change in PTG?12:28
amoralej_we can tag it as a feature ans set as triaged, yes12:29
amoralej_and move it to watcher12:29
amoralej_or close if we prefer not to track features as bugs, both wfm12:29
sean-k-mooneyya we found the orgianal spec and unfortuenly the non intuitive bevhior si what was desgined12:30
sean-k-mooneyso we need to decieed how to adress that as a fetarue rather then a bug12:30
jgilaberso we keep the bug as wishlist for watcher?12:31
sean-k-mooneywe can12:31
amoralej_given that we are changing the state machine i think we should treat it as a spec feature ?12:31
sean-k-mooneybut we will likely close it when we figure out how to impove it12:31
sean-k-mooneyamoralej_: that my inclination but im ok with a wishlist bug to just track it in the backlog until it becomes a priority12:32
amoralej_i think that's a good approach, keep it open as whishlist or rfe until to not forget it until we start the feature process12:32
sean-k-mooneyamoralej_: unless you plan to start workign on it now 12:32
amoralej_nop12:32
jgilaberack, sounds good I'll do that12:32
sean-k-mooneyya so let add an rfe tag too12:32
sean-k-mooneyand ya we can dicuss at the ptg or a future meeting if it makes sense to expidite but it not super urgent to fix this as long as it clearly docuemented12:33
sean-k-mooneyjgilaber: ithink you said there were two bugs? what is the ohter12:33
jgilaber#link https://bugs.launchpad.net/watcher-tempest-plugin/+bug/213448712:34
jgilaberthis one comes from a ci failure12:34
jgilaberI'm not sure how often that happends12:34
jgilaberI don't remember seeing that test failing too much12:34
jgilaberdviroel, do you remember any details on this bug since you reported it?12:35
dviroeltrying to remember but, i don't think I see that recently in ci too12:35
chandankumarmay be we close it? If we see again, reopen it?12:37
jgilaberI would either do that or set it as low and revisit in the future, any thoughts?12:38
dviroelyeah, maybe set to incomplete and wait for more logs/failures12:39
amoralej_yes, i think setting it as incomplete is good12:39
jgilaberack, done12:40
sean-k-mooneyis this realtede to the previous issue12:40
sean-k-mooneyor is "CANCELLED instead of SUCCEEDED" a coincidnece12:41
jgilaberI don't think so, this is refering to the state of an audit12:41
jgilaberthe other was about the action plan of a continous audit12:41
amoralej_yes, it's audit, not action plan12:41
sean-k-mooneyok but is the audit a continuous one or a onehost12:42
amoralej_i was also remembering another issue related with cancelled state and continuous but that was also for actionplans12:42
sean-k-mooneyim wondierign if we cancel runing autis when the next interval triggers12:42
sean-k-mooneybut superceed compelted ones12:43
chandankumarhttps://github.com/openstack/watcher-tempest-plugin/blob/49e68f7556c35248d3c960e67ecbb927183031bc/watcher_tempest_plugin/tests/api/admin/test_action.py#L7412:43
jgilaberI think by default it creates a oneshot audit12:44
jgilaberin https://github.com/openstack/watcher-tempest-plugin/blob/49e68f7556c35248d3c960e67ecbb927183031bc/watcher_tempest_plugin/tests/api/admin/test_action.py#L3712:44
jgilaberlooking at https://github.com/openstack/watcher-tempest-plugin/blob/49e68f7556c35248d3c960e67ecbb927183031bc/watcher_tempest_plugin/tests/common/base.py#L15412:44
dviroelyeah, when no set, it defaults to ONESHOT12:45
chandankumaryes it is oneshot audit12:45
sean-k-mooneyok thats defaulted at the api level?12:46
sean-k-mooneyi was not seeign it really clearly 12:46
sean-k-mooneyin teh code12:46
amoralej_not in the api12:46
amoralej_it's jgilaber link12:47
amoralej_setting ONESHOT12:47
amoralej_in create_audit method12:47
sean-k-mooney there are 2 implemation of create_audit12:47
sean-k-mooneyhttps://github.com/openstack/watcher-tempest-plugin/blob/master/watcher_tempest_plugin/services/infra_optim/v1/json/client.py#L140-L15112:48
sean-k-mooneythat one does nto default anything12:48
sean-k-mooneyand that is the one we are using12:48
amoralej_that's called by https://github.com/openstack/watcher-tempest-plugin/blob/49e68f7556c35248d3c960e67ecbb927183031bc/watcher_tempest_plugin/tests/common/base.py#L154 iiuc12:48
dviroelright12:49
sean-k-mooneyok12:50
sean-k-mooneyits coming in via the leper mixing12:50
amoralej_yes12:50
sean-k-mooneythen ya12:50
sean-k-mooneyi dont know of anything that woudl cause it to be canceld12:50
sean-k-mooneyother then a direct api request for one shot12:50
sean-k-mooneybeyond a restart12:50
dviroelyeah, the WatcherHelperMixin was added in the commit that I mentioned, but I remember that the job started failing after we merged that change12:51
dviroelbut we need more failures and logs to debug12:51
amoralej_yep12:51
amoralej_restarting the decision engine i.e. can cancel audits, but i guess this is not the case12:51
sean-k-mooneyof ya we need to have it fail then determin how it when to canceled12:51
sean-k-mooneyamoralej_: well its possible there was a restart but we woudl see that in the logs12:52
amoralej_yes, we need logs12:52
sean-k-mooneywe print the config optiosn at start up so there coudl have been an OOM event12:52
sean-k-mooneyor somethign like that12:52
sean-k-mooneyi guess for now lets mark it incomplete12:53
sean-k-mooneysince we dont have teh data or a theory of how it can happen outside fo the descion engine restarting12:54
jgilaberack, done12:54
jgilaberwe have one last topic12:54
jgilaber#topic PoC on nova and prometheus fixtures12:54
jgilaber#link https://review.opendev.org/c/openstack/watcher/+/98025712:54
amoralej_yes, i added last minute one :)12:55
amoralej_when working in the scale bugs i ended up creating simple fake nova and prometheus services12:55
amoralej_otherwise i was working blind12:55
amoralej_i've thought those could be part also of the functional test12:56
amoralej_so that we can use them both integrated in the functional tests or as standalone servers12:56
sean-k-mooneyamoralej_: oh you made a start12:56
amoralej_but before going beyong i wanted to check if i'm going in the right direction12:56
sean-k-mooneyi can try and take a look befor i go on pto12:57
amoralej_in that early state it has already be useful for me to replicate scalability issues locally12:57
amoralej_i used flask to create them12:57
sean-k-mooneyill be way monday->wednesday next week 12:57
amoralej_it gives features as ssl, etc...12:57
sean-k-mooneyamoralej_: im not sure we want to dothat12:57
sean-k-mooneyi.e. create a webserver12:57
sean-k-mooneywe dont wnat to test ssl ectrs in the fucntional tests12:57
amoralej_yes, that's my question :)12:58
amoralej_yes, but as said before, i think it'd be good to have a tool also useful beyond automated functional tests12:58
amoralej_i.e. replacing actual services in a local deployment12:58
sean-k-mooneyperhaps12:58
amoralej_for devel, replicate issues, etc...12:58
sean-k-mooneybut at that point it likely shoudl not be in the watcher repo12:58
amoralej_other option would be that, to create two totally different things12:59
amoralej_and not being part of watcher tests12:59
sean-k-mooneywe dont currently use falks by the way12:59
sean-k-mooney*flask12:59
amoralej_going to contrib folder or to even other repo12:59
amoralej_yes, i know12:59
amoralej_we don't have much time now13:00
amoralej_but i wanted to open the discussion13:00
sean-k-mooneyyep maybe a good topic for the ptg13:00
sean-k-mooneyim not really afan of the idea of having giant xml files either by the way13:00
sean-k-mooneyi dont like howe we are curertny using the them in our unit tests13:01
sean-k-mooneybut it felt to risky to just rip them out13:01
sean-k-mooneybut i was hoping we coudl entirly get away form that in the funtional tests13:01
sean-k-mooneyand instead compute the moduels programaticly13:01
amoralej_i need to join another mtg now13:02
sean-k-mooneycool lets loop back to this 13:02
jgilaberthanks amoralej_ looks interesting but we're out of time, we can move discussion to the patch or defer to PTG13:02
amoralej_but i think having the option to use different models or metrics as data files is good :)13:02
dviroelack, we can also discuss more in the patch13:02
jgilaberany last minute topic for opens discussion?13:02
jgilaberI see dviroel++ has voluntereed for next meeting so we can wrap up here13:03
jgilaberthanks everyone!13:03
jgilaber#endmeeting13:03
opendevmeetMeeting ended Thu Mar 12 13:03:47 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)13:03
opendevmeetMinutes:        https://meetings.opendev.org/meetings/watcher/2026/watcher.2026-03-12-12.00.html13:03
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/watcher/2026/watcher.2026-03-12-12.00.txt13:03
opendevmeetLog:            https://meetings.opendev.org/meetings/watcher/2026/watcher.2026-03-12-12.00.log.html13:03
dviroeljgilaber++13:03
chandank`jgilaber++13:03
amoralej_thanks jgilaber++13:04
opendevreviewMerged openstack/watcher master: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/96026513:18
dviroeljgilaber ^ so our last fix merged, we can update the release patch13:27
dviroeljgilaber: please check/review the 2 release patches:14:14
dviroelhttps://review.opendev.org/c/openstack/releases/+/97962214:14
dviroelhttps://review.opendev.org/c/openstack/releases/+/97962314:14
chandankumardviroel: Do we want to include https://review.opendev.org/c/openstack/watcher/+/979137 grenade patch in rc1 patch or will backport later?14:19
jgilaberwill do dviroel, thanks14:21
dviroelchandankumar: we can backport later if needed, but not a fix that we need to include in the release imho14:22
jgilaber+1, since it's a CI fix14:23
jgilaberI think we can merge today anyway, I'm +2 on the new grenade job14:23
jgilaberdviroel, did you plan to review it or should I +W as well?14:24
dviroelack, i may review today too14:24
jgilaberack14:24
dviroelreviewing now14:24
chandankumarack14:25
opendevreviewJoan Gilabert proposed openstack/watcher stable/2025.2: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/98029814:42
opendevreviewJoan Gilabert proposed openstack/watcher stable/2025.1: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/98029914:43
opendevreviewJoan Gilabert proposed openstack/watcher stable/2024.2: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/98030014:43
opendevreviewJoan Gilabert proposed openstack/watcher stable/2024.2: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/98030016:11
opendevreviewJoan Gilabert proposed openstack/watcher stable/2025.2: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/98029816:12
opendevreviewJoan Gilabert proposed openstack/watcher stable/2025.1: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/98029916:13
opendevreviewJoan Gilabert proposed openstack/watcher stable/2025.1: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/98029917:54
opendevreviewJoan Gilabert proposed openstack/watcher stable/2024.2: Handle missing fields building storage model XML or list  https://review.opendev.org/c/openstack/watcher/+/98030017:54

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