14:00:07 <pdeore> #startmeeting glance 14:00:07 <opendevmeet> Meeting started Thu Jun 30 14:00:07 2022 UTC and is due to finish in 60 minutes. The chair is pdeore. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:07 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:07 <opendevmeet> The meeting name has been set to 'glance' 14:00:08 <pdeore> #topic roll call 14:00:08 <pdeore> #link https://etherpad.openstack.org/p/glance-team-meeting-agenda 14:00:12 <pdeore> o/ 14:00:16 <mrjoshi> o/ 14:00:57 <abhishekk> o/ 14:01:27 <pdeore> abhishekk, how are feeling now? I thought you will not around today 14:01:34 * abhishekk down with fever so probably will leave after the meeting 14:02:00 <abhishekk> not much well 14:02:10 <abhishekk> will signoff after the meeting 14:02:19 <dansmith> o/ 14:02:26 <pdeore> ack 14:02:34 <abhishekk> I think rosmaita is not around 14:02:45 <pdeore> yeah he is on PTO 14:02:56 <pdeore> lets wait few minutes for everyone to join 14:04:29 <abhishekk> ack 14:05:29 <pdeore> I think jokke_ and Cyril are not around, 14:05:54 <pdeore> shall we start or wait for few more mints ? 14:06:05 <abhishekk> we should start 14:06:11 <pdeore> ok 14:06:13 <pdeore> #topic release/periodic jobs 14:06:20 <pdeore> Milestone 2 is 2 weeks away from now 14:06:35 <pdeore> JFYI, I have updated the release schedule doc as well with the glance spec freeze date for Zed schedule 14:06:41 <pdeore> #link https://releases.openstack.org/zed/schedule.html 14:07:04 <pdeore> So next week would be review party week, as we have decided during PTG to have this 10 days before milestone release 14:07:06 <abhishekk> ack, thank you for updating the schedule 14:07:24 <abhishekk> ok 14:07:28 <pdeore> we will discuss about finalizing the date in the dedicated topic today 14:07:30 <croelandt> so we'd need 3,4 cores :) 14:07:38 <dansmith> I'll be out all next week except for tuesday, FYI :) 14:07:47 <abhishekk> I will be around 14:08:01 <pdeore> I will also be around 14:08:20 <abhishekk> we can make an exception and have this review party one week later after the release or during release week 14:08:48 <abhishekk> 11th or 12th July 14:08:55 <pdeore> +1 14:09:51 <abhishekk> we can spend one hour and go through some important patches 14:10:29 <abhishekk> or on scheduled time we may have jokke_ brian me and cyril as core and pdeore as well 14:10:57 <jokke_> Brian isback next week, right? 14:11:13 <croelandt> 5th I believe 14:11:16 <abhishekk> yes, he will be there on 5th 14:11:25 <croelandt> so mayeb the 7th would be alright after all 14:11:35 <abhishekk> sounds good 14:11:43 <abhishekk> 7th same as meeting time 14:12:15 <jokke_> sounds like a plan 14:12:48 <pdeore> cool, 14:12:49 <jokke_> unless dansmith has objection us doing it without him 14:12:54 <dansmith> lol 14:12:58 <abhishekk> cool, pdeore make a note of it 14:13:05 <abhishekk> we will miss you dansmith 14:13:17 <dansmith> heh 14:13:25 <pdeore> abhishekk, ack 14:13:44 <pdeore> moving ahead :) 14:13:54 <pdeore> we have released glance-store stable/wallaby with some important bug fixes 14:13:56 <jokke_> dansmith: maybe we call ya an hour before, just to make sure you don't want to join us? :P 14:14:41 <pdeore> Periodic Jobs are all green except fips jobs failing with RETRY_LIMIT 14:14:41 <dansmith> you go right ahead.. I'll be far from cell range (and even pavement) 14:14:56 <abhishekk> wow 14:14:56 <jokke_> dansmith: sounds gr8, enjoy 14:15:11 <dansmith> :) 14:15:24 <jokke_> pdeore: any idea what it's retrying? 14:17:02 <pdeore> seems 'nslookup_target' is undefined 14:17:13 <dansmith> that's supposed to be resolved 14:17:21 <dansmith> we should poke ade_lee 14:17:46 <pdeore> ok 14:19:07 <pdeore> ok, moving to next topic 14:19:14 <pdeore> #topic Spec for review 14:19:25 <pdeore> [APIImpact] Add DELETE api for metadef resource types #link https://review.opendev.org/c/openstack/glance-specs/+/818192 14:19:46 <pdeore> This is the only spec remaining from our review list, which is a low priority but good to have I think. 14:20:12 <pdeore> I assume that everyone is agreed on having the separate Delete api for resource type rather than auto-deleting with some periodic operation. 14:20:53 <abhishekk> I will have a look at it, I am pro for having API for it 14:21:12 <dansmith> tbh, 14:21:33 <dansmith> I haven't really gotten a clear understanding of why we can't delete when the last use is deleted 14:21:56 <dansmith> I think both rosmaita and I preferred that approach, but I'm not sure I ever got clarity on why it wasn't okay 14:22:37 <abhishekk> i didn't aware about that approach, will think on it 14:22:55 <pdeore> Bcz I think it can be used later by someone if available.. 14:23:04 <dansmith> but what does that matter? 14:23:12 <abhishekk> not much 14:23:21 <dansmith> they're auto-created in the first place, why not auto-delete them when they're no longer used? 14:24:30 <abhishekk> if they are auto-created then I guess we can delete it if it is not used on last delete 14:24:41 <dansmith> that's what we were thinking 14:25:28 <jokke_> I tend ot agree with Brian and Dan here. I think the fact that there is no create api at all for the resource type is good indication that we likely don't need user action to delete them either 14:25:32 <pdeore> yeah right, that's why we had a discussion during PTG to have create & delete both similar to other metadef apis 14:26:09 <dansmith> right, it's just an extra step for the user (or client or whatever) to have to delete them to clean up 14:27:41 <jokke_> crazy question. Does the resource type contain any valuable information for it to exists as it's own resource in the db in first place? 14:27:45 <pdeore> yeah so if we are not having create then let's not have delete as well 14:28:18 <abhishekk> pdeore, I guess this approach will reduce your work as well 14:28:21 <pdeore> not really I think 14:28:28 <pdeore> yeah :) 14:28:49 <jokke_> Like is there a reason why it wouldn't be extra field on the namespace entry rather than being it's own thing in general 14:29:14 <dansmith> it's used by multiple things when it's used, 14:29:25 <dansmith> so I think that's why it is its own entity, for the relation 14:29:30 <jokke_> ok, so makes sense to keep it as it's own entity 14:29:33 <dansmith> but it's auto-created when you create a namespace that uses it 14:29:35 <jokke_> yeah 14:32:11 <jokke_> Next crazy question I have then is, if we decide to take the route of auto-delete instead of new API call, do we even need full spec for this? 14:32:36 <jokke_> Or can that be converted to -lite and done as best effort refactoring as convenient? 14:33:00 <abhishekk> sounds good to me 14:33:10 <dansmith> not sure we even need a lite spec 14:33:11 <dansmith> could be a bug 14:33:15 <dansmith> but sure :) 14:33:42 <abhishekk> if we convert it to bug then we don't need to worry about spec freeze :D 14:33:52 <abhishekk> but I think lite spec will be good 14:33:53 <dansmith> \o/ 14:34:00 <dansmith> /o\ 14:34:02 <pdeore> cool :) 14:34:03 <abhishekk> :D 14:34:59 <abhishekk> I will discuss this offline with pdeore later 14:35:03 <abhishekk> lets move ahead 14:35:10 <pdeore> yeah, sure 14:35:22 <pdeore> #topic Columbus PTG Team signup 14:35:27 <abhishekk> this is me 14:35:50 <abhishekk> I need to fill a survey for Infra about head counts for October PTG 14:36:02 <abhishekk> who is planning to attend this time? 14:36:17 <abhishekk> I am thinking to give approx 7/8 head count 14:36:41 <jokke_> I have no idea as I haven't heard anything about our travel planning for it yet 14:36:51 <abhishekk> we will soon 14:38:29 <abhishekk> ok, I will give tentative count for this and we can modify it later 14:38:40 <croelandt> Reminded that air travel is not compatible with our CO2 emissions goals for the end of the century 14:38:43 <croelandt> Reminder* 14:39:00 <abhishekk> ++ 14:39:05 <croelandt> and we should probably start by stopping intercontinental travel 14:39:15 <croelandt> and therefore downsizing meetings :/ 14:39:16 <jokke_> croelandt: fortunately I'm on the coast line, so they can hire a yacht to take me over the pond :P 14:39:28 <abhishekk> :D 14:39:33 * croelandt checks whether yachts are worse than planes 14:39:47 <dansmith> depends on the yacht 14:40:21 <abhishekk> jokke_, if you get one then take a small detour and pick me up as well 14:41:18 <jokke_> abhishekk: then we can meet at Dan's house and take road trip from there 14:41:32 <dansmith> it's a long road trip from here :) 14:41:34 <jokke_> cause that's so carbon neutral plan as any 14:41:50 <abhishekk> I am in :D 14:41:59 <dansmith> and especially not carbon-neutral in any of my cars :) 14:42:00 <jokke_> dansmith: other option is that you hop on, it's longer but nicr trip if we get around via Panama 14:42:17 <croelandt> nothing is truly carbon-neutral 14:43:40 <abhishekk> lets move to open discussion, we can discuss the possible road trip a week after once dansmith is back 14:43:55 <dansmith> heh 14:44:10 <pdeore> moving to open discussions :D 14:44:11 <pdeore> #topic Open Discussions 14:44:34 <abhishekk> there are few glance patches open,please have a look at those 14:44:49 <abhishekk> important is glance-download and immediate caching 14:46:15 <jokke_> oh speaking of which 14:46:31 <jokke_> the fips job fix is submitted by Ada, just not reviewed 14:46:41 <jokke_> 847089 14:46:43 <abhishekk> ack 14:46:46 <dansmith> if you're looking at the glance patch, 14:46:58 <jokke_> yup 14:47:00 <dansmith> he decided to fix it globally, so I thought that glance patch was not required anymore 14:47:13 <jokke_> specially mentioning the nslookup_target 14:47:31 <dansmith> right, 14:47:40 <dansmith> although I see he depends on the base patch, so maybe both are required, 14:47:44 <dansmith> but from the discussion I thought not 14:47:49 <jokke_> well it's depending on the zuul job patch 14:48:00 <jokke_> so I'd expect it being still needed 14:48:21 <dansmith> yeah, okay maybe so 14:49:44 <abhishekk> if get time I will have a look at that patch 14:50:37 <dansmith> it's just zuul config, so if he says it's required we might as well just slam it in 14:50:50 <abhishekk> ack 14:51:01 <dansmith> I just rechecked the experimental job so we'll see 14:51:12 <abhishekk> cool 14:51:16 <abhishekk> I think we are good for today, nothing else from me other than some review requests 14:51:37 <pdeore> ack 14:51:37 <jokke_> nothing from me 14:51:54 <pdeore> let's wrap up then .. 14:52:04 <abhishekk> thank you all o/ 14:52:06 <pdeore> Thanks everyone for joining !! 14:52:09 <abhishekk> have a nice weekend 14:52:27 <pdeore> #endmeeting