14:00:01 <abhishekk> #startmeeting glance
14:00:01 <opendevmeet> Meeting started Thu Aug 18 14:00:01 2022 UTC and is due to finish in 60 minutes.  The chair is abhishekk. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:01 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:01 <opendevmeet> The meeting name has been set to 'glance'
14:00:07 <abhishekk> #topic roll call
14:00:13 <abhishekk> #link https://etherpad.openstack.org/p/glance-team-meeting-agenda
14:00:26 <abhishekk> waiting for others to show
14:00:28 <abhishekk> o/
14:00:41 <mrjoshi_> o/
14:00:45 <dansmith> o/
14:00:51 <abhishekk> update, cyril is on leave for 3 weeks (including this)
14:01:36 <abhishekk> mrjoshi_, there are some comments on your patch, if you have any doubts then get it cleared here
14:02:39 <rosmaita> o/
14:02:57 <abhishekk> pdeore, will be late, lets start
14:03:09 <abhishekk> #topic Update
14:03:11 <rosmaita> (sorry, i am running a video meeting simultaneously)
14:03:19 <jokke_> o/
14:03:19 <abhishekk> np!
14:03:38 <abhishekk> The PTG will be virtual now
14:04:09 <abhishekk> We will follow the same time which we used to follow for previous PTG's
14:04:23 <abhishekk> that will be between 1400 UTC to 1700 UTC
14:04:52 <abhishekk> If anyone has any concerns related to timings then please let me know
14:04:56 <abhishekk> moving forward;
14:05:08 <abhishekk> #topic release/periodic jobs updates
14:05:21 <abhishekk> Milestone 3 two weeks away
14:05:29 <abhishekk> we have 2 features in waiting list
14:05:45 <abhishekk> lets discuss this in next topic
14:05:59 <abhishekk> glance-store final release will be next week
14:06:28 <abhishekk> We don't have anything pending for store, so good to tag the final release by next Wednesday
14:07:03 <abhishekk> I will suggest cores to go through list of patches of glance store and see it there is anything important we can get merged before next week
14:07:23 <jokke_> ack
14:07:51 <abhishekk> cool, periodic jobs all green except time out for fips related jobs
14:08:14 <abhishekk> moving to next topic
14:08:24 <abhishekk> #topic Pending features
14:08:31 <abhishekk> Store details API - mrjoshi
14:09:15 <abhishekk> there are few suggestions from rosmaita, I think one is related to using actual parameter names
14:09:35 <abhishekk> and other one is should we show computed values rather than in Bytes
14:10:08 <rosmaita> actually, the patch is showing the computed values in bytes, but the operator can only configure in MB
14:10:31 <rosmaita> so i guess the question is what is the primary goal of the discovery response?
14:10:47 * pdeore is back
14:11:05 <rosmaita> i was thinking it was for an operator to see how the store is configured
14:11:13 <abhishekk> correct
14:11:47 <rosmaita> ok, then we should probably not show the computed values ... or we could show them, but indicated separately somehow
14:12:05 <rosmaita> the big problem though is that we already implemented store discovery for rbd store
14:12:18 <abhishekk> rosmaita, I think if we decides to do it, do it separtely for both the stores
14:12:20 <rosmaita> and that currently shows the computed value, not the configured value
14:14:07 <abhishekk> at this moment I am good with computed values
14:14:22 <abhishekk> we can explain about it in release notes
14:14:51 <mrjoshi_> I think initially we were going with the computed values only when we had it for the rbd store
14:15:08 <abhishekk> right
14:15:44 <abhishekk> and if you are going to change it then you need to file a bug fix it for rbd first and then we can merge this patch
14:15:52 <abhishekk> that will be small change though
14:16:10 <jokke_> Will break the API
14:16:41 <abhishekk> how?
14:17:01 <jokke_> To change it for rbd it's published and released already
14:17:54 <abhishekk> in that case I think its better to explain it in docs and release note
14:18:01 <jokke_> so better option is to use another key name all together if we want to show it as it's configured rather than how the system sees it
14:18:03 <dansmith> it will change only in the respect that it will be correct instead of incorrect right?
14:18:32 <abhishekk> that will increase the confusion
14:18:35 <dansmith> a user doesn't know that it's wrong right now and the units are no different once we show the right value yeah?
14:20:22 <abhishekk> I think its better to show computed value and explain it in reno
14:20:38 <jokke_> it's not wrong, the code uses bytes, the config otoh is expecting Mi. So both are correct and it does not have unit with it on the field
14:21:13 <dansmith> right, we expose bytes, but where they will only ever be a multiple of 1MiB today correct?
14:21:49 <dansmith> there's no reasonable client behavior that would be broken by reading a byte value that is not MiB-aligned in the future, that I can think of :)
14:23:54 <abhishekk> Ok, we need an agreement here otherwise this will get dropped from the cycle
14:24:16 <jokke_> So there is no usecase for that info in the first place. But what will be confusing AF, is that if we change it now, and you look the output from older system showing you bytes (without unit) and new version that shows you integer without units too but the value just happens to be MiB rather than bytes
14:24:52 <abhishekk> I am still at opinion of implementing it same way as implemented in past
14:25:04 <jokke_> ++
14:25:46 <jokke_> I'm sure we have something like API ref that explains what the value is and helptext in the config explaining what is expected
14:25:55 <abhishekk> mrjoshi_, just change the parameter names of swift store at this moment
14:26:07 <mrjoshi_> abhishekk, ack
14:26:18 <abhishekk> moving ahead
14:26:29 <abhishekk> 2nd pending work is Glance download plugin
14:27:12 <abhishekk> I am afraid both alistarl1 and Pierre-Samuel Le Stang are not around today
14:27:51 <abhishekk> the patch is in good shape with few changes, should we go forward and work on that
14:28:08 <abhishekk> or we should use FFE for the same?
14:28:38 <dansmith> yeah I was going to ask..seems very close but not sure if they're working on those feedback items or not
14:29:26 <abhishekk> not sure either, didn't heard from them since last meeting
14:29:43 <abhishekk> and I also don't want to duplicate the efforts
14:30:22 <abhishekk> I will wait till Monday to get update from them,else will start working on the patch
14:30:34 <abhishekk> sounds good?
14:30:39 <dansmith> sure
14:30:44 <jokke_> for the sake of keeping the patch moving, I'd say if there's no response from them by next week, feel free to propose a new rev ... I'm sure they'd appreciate having it in this cycle too
14:30:44 <abhishekk> cool
14:30:54 <abhishekk> ++
14:31:18 <abhishekk> moving forward
14:31:29 <abhishekk> #topic Glance PTG planning etherpad
14:31:35 <abhishekk> PTG planning etherpad is up
14:31:44 <abhishekk> #link https://etherpad.opendev.org/p/antelope-ptg-glance-planning
14:32:00 <abhishekk> Feel free to add topics which you want to discuss during this PTG
14:32:12 <abhishekk> also this will be my last cycle as PTL
14:32:32 <abhishekk> and hopefully pdeore is up for taking the responsibility
14:32:38 <jokke_> PTL is gone, long live PTL!
14:32:48 <jokke_> :)
14:32:49 <abhishekk> :P
14:33:37 <abhishekk> I will help pdeore to organize this PTG
14:34:01 <abhishekk> that is all from me
14:34:09 <abhishekk> any questions, suggestions?
14:34:59 <abhishekk> Just to inform you, I will be going on leave from 10th of September (just before rc1 week)
14:35:21 <abhishekk> if everything is alright and we manage to tag Zed release before that then it will be good
14:35:35 <abhishekk> otherwise I will work part time during rc1 week
14:36:32 <jokke_> Working holiday? Aren' you getting old for that :P
14:36:57 <abhishekk> :D if you have my back then I don't need to do that
14:37:22 <abhishekk> I think we should wrap up now
14:37:45 <abhishekk> thank you all
14:37:51 <abhishekk> have a nice weekend ahead
14:37:59 <jokke_> thanks
14:38:07 <abhishekk> #endmeeting