15:00:54 <ttereshc> #startmeeting Pulp Triage 2021-02-02 go/no-go meeting for 3.10.0 15:00:54 <ttereshc> #info ttereshc has joined triage 15:00:54 <ttereshc> !start go/no-go meeting for 3.10.0 15:00:54 <pulpbot> Meeting started Tue Feb 2 15:00:54 2021 UTC. The chair is ttereshc. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:54 <pulpbot> Useful Commands: #action #agreed #help #info #idea #link #topic. 15:00:54 <pulpbot> The meeting name has been set to 'pulp_triage_2021-02-02_go/no-go_meeting_for_3.10.0' 15:00:54 <pulpbot> ttereshc: ttereshc has joined triage 15:01:36 <daviddavis> #info daviddavis has joined triage 15:01:36 <daviddavis> !here 15:01:36 <pulpbot> daviddavis: daviddavis has joined triage 15:02:07 <ttereshc> heh, you re the first one, daviddavis 15:02:12 <x9c4> #info x9c4 has joined triage 15:02:12 <x9c4> !here 15:02:12 <pulpbot> x9c4: x9c4 has joined triage 15:02:14 <daviddavis> hehe 15:02:44 <ttereshc> https://bit.ly/3pLHmg0 here is a list of open issues for 3.10 15:03:10 <daviddavis> the label filtering should be merged today or tomorrow. just have some feedback to address. 15:03:13 <ttereshc> and I believe x9c4 wanted to add one more 15:03:21 <bmbouter> #info bmbouter has joined triage 15:03:21 <bmbouter> !here 15:03:21 <pulpbot> bmbouter: bmbouter has joined triage 15:03:22 <ttereshc> great 15:03:23 <ggainey> #info ggainey has joined triage 15:03:23 <ggainey> !here 15:03:23 <pulpbot> ggainey: ggainey has joined triage 15:03:29 * bmbouter does the dance of 3.10 15:04:30 <ttereshc> we have one in the new state https://pulp.plan.io/issues/8182 15:04:32 <ggainey> I need a review (prob best from daviddavis ) of https://github.com/pulp/pulpcore/pull/1090 15:04:34 <x9c4> ttereshc, https://github.com/pulp/pulpcore/pull/1098 got merged. I can still add the tickets to the milestone. 15:04:39 <ggainey> we have 2 in assigned and one on new 15:04:45 <ttereshc> x9c4, please do 15:04:59 <daviddavis> ggainey: did you figure out the failing test? 15:05:06 <ggainey> I've been using this https://pulp.plan.io/issues?c%5B%5D=project&c%5B%5D=tracker&c%5B%5D=status&c%5B%5D=priority&c%5B%5D=cf_5&c%5B%5D=subject&c%5B%5D=author&c%5B%5D=assigned_to&c%5B%5D=cf_3&c%5B%5D=cf_7&f%5B%5D=fixed_version_id&f%5B%5D=&group_by=&op%5Bfixed_version_id%5D=%3D&set_filter=1&sort=status%2Cid%3Adesc&t%5B%5D=&utf8=%E2%9C%93&v%5Bfixed_version_id%5D%5B%5D=166 query 15:05:15 <bmbouter> I plan to work on 8182 next 15:05:29 <ggainey> daviddavis: the current failure has nothing to do with the code I checked in, it's CI being Bad (and I don't have perms to re-run) 15:05:31 <bmbouter> and post by EOB today which would be just to review tomorrow 15:05:53 <bmbouter> ggainey: I was thikning the CI would run against your fork 15:06:14 <daviddavis> ggainey: I meant the test that got commented out but I'll rerun the jobs 15:06:16 <ttereshc> daviddavis, I looked at the failing test for the import check pr today and it looks similar to the migration test failures, one of the tasks is cancelled 15:06:34 <ttereshc> so even if the re-run helps, I'm wondering if we have some deeper issue 15:06:49 <daviddavis> interesting 15:06:58 <ggainey> daviddavis: the test that is commented out shouldn't hold up the PR - I have no idea how long it'll take to figure out why CI behaves differently than expected 15:07:20 <ggainey> ttereshc: possibly, but not because of this PR 15:07:35 <ttereshc> no,not because of the changes 15:07:37 <ttereshc> I agree 15:07:44 <ggainey> kk 15:08:01 <ttereshc> bmbouter, great, ping for review when needed 15:08:20 <bmbouter> will do, ty 15:08:46 <ttereshc> did you all reach an agreement around the component name in the status API? (I missed the pulpcore meeting) 15:09:00 <x9c4> I belive yes: 15:09:03 <ttereshc> basically, are we blocked in anyway there or not? 15:09:23 <x9c4> We want to make the requrement to the plugins now. 15:09:37 <bmbouter> we did 15:09:45 <x9c4> But defer the breaking change in the status response in 3.11 15:09:54 <bmbouter> I'm going to split my PR into two, the parts that will require more coordiantion will come with 3.11 15:10:04 <bmbouter> the parts that can go into 3.10 will. I'm splitting it now 15:10:11 <ggainey> +1 , cool 15:10:13 <x9c4> early in the cycle to have time to fix issues with e.g. the CLI. 15:10:14 <ttereshc> great, thanks ! 15:10:26 <bmbouter> I will split the issue into two also to separate the scope 15:10:31 <x9c4> +1 15:10:42 <dkliban> #info dkliban has joined triage 15:10:42 <dkliban> !here 15:10:42 <pulpbot> dkliban: dkliban has joined triage 15:10:45 <ipanova> #info ipanova has joined triage 15:10:45 <ipanova> !here 15:10:45 <pulpbot> ipanova: ipanova has joined triage 15:10:54 <ttereshc> any concerns from anyone to release 3.10 this Thursday? 15:11:13 <daviddavis> not me 15:11:21 <ipanova> i have one thing 15:11:22 <bmbouter> just to make sure the PR from ewoud merges 15:11:48 <ipanova> since we have a day of learning on friday and we will have troubles to release 3.10 on thursday we could do it on monday 15:11:59 <ipanova> and if we * 15:12:14 <bmbouter> that would push back the pulp_ansible 0.7 release timeline 15:12:21 <ggainey> I'm just concerned that we have 2 assigned and one new, and are running out of time :( 15:12:22 <ipanova> bmbouter: that's ok 15:12:26 <bmbouter> which is slated for feb 9 15:12:28 <ipanova> AH is also having day of learning 15:13:11 <dkliban> bmbouter: ewoud's PR is going to merge in the next hour 15:13:21 <ttereshc> FWIW, I'm releasing and I believe I won't have a learning day regardless, so I can release on Friday as well if there are folks to review 15:13:24 <bmbouter> can we revisit if we are feeling good about thursday 15:13:40 <ttereshc> sure 15:13:47 <bmbouter> ggainey: I hear your concerns so I want to not have them pass unnoticed 15:13:56 <ipanova> ttereshc: sounds good 15:14:10 <dkliban> ggainey: one assigned to me is a 3 line change. i'll open a PR for it after this meeting. 15:14:20 <daviddavis> my concern is that the CI is broken. does anyone have any idea why? 15:14:33 <ggainey> coolio 15:15:29 <ttereshc> daviddavis, is it? it worked fine in the morning after merging pulp-fixtures port fix 15:15:32 <x9c4> which CI? The last commit merged 22' ago is green. 15:15:57 <daviddavis> oh, I thought you all said there was some deeper issue causing the CI to fail 15:16:12 <dalley> daviddavis, it's no longer broken. it was yesterday though 15:16:22 <daviddavis> oh ok 15:16:23 <ttereshc> daviddavis, it's for task groups from what I can see and reprodicuble every now and then 15:16:24 <x9c4> Thats an issue with the status api change. that is postponed. 15:16:38 <daviddavis> ttereshc: I see, thank you 15:16:54 <ttereshc> daviddavis, I'm trying to look at it 15:17:07 <ttereshc> it blocks migration plugin 15:17:31 <daviddavis> +1 15:17:47 <bmbouter> I have a concern/item 15:17:57 <bmbouter> if its ok to move on 15:18:04 <ttereshc> +1 from me 15:18:20 <dkliban> let's move on to your concern bmbouter 15:18:21 <bmbouter> I was looking at 8048 https://pulp.plan.io/issues/8048 but it's not going to go into 3.10 15:18:42 <bmbouter> I still plan to continue (I've solved about half of it I think) so I'd like to keep assigned and move out of 3.10 15:19:02 <dkliban> is AH cool with that? 15:19:07 <ipanova> will this block ansible folks to set their own perms for container ansible plugins? 15:19:08 <bmbouter> galaxy_ng is affected but are ok to not have in 3.10 (my understanding), but what about pulp_container? 15:19:32 <ipanova> bmbouter: i am not sure they will be able to start the integration if this is not fixed 15:19:42 <ipanova> integration with pulp_container 15:19:44 <bmbouter> yeah ths is what I was concerned about also 15:20:08 <x9c4> For pulp_container itself, there is no object without a NamedViewset. 15:20:20 <bmbouter> agreed (and I wish galaxy_ng would adopt this as well) 15:20:36 <ipanova> x9c4: https://pulp.plan.io/issues/8048#note-4 15:20:38 <bmbouter> but anyway we can continue on to fix it, we can make it more general and usable 15:20:51 <dkliban> yes, but there are other plugins that don't have it ... and having that plugin installed breaks it for all permissions 15:21:00 <ipanova> yep 15:21:21 <bmbouter> let me check my understanding 15:21:37 <ttereshc> (time check 8 mins left) 15:21:47 <bmbouter> by having 8048 not fixed, the issue is that a plugin cannot have a viewset that is not NamedModelViewSet that uses an AccessPolicy is that right? 15:22:13 <bmbouter> and if that's correct, that is what they need fixed in 3.10.0? 15:22:47 <dkliban> yes, i believe that is correct 15:22:57 <x9c4> If there is a NamedViewset, i think there can be another one additionally. 15:23:14 <bmbouter> ok so let's keep this on 3.10 and keep it a release blocker 15:23:41 <bmbouter> is someone able to take this work from me or should I try to fix tomorrow? 15:23:50 <x9c4> I can 15:24:06 <bmbouter> x9c4: can we meet for a few min for me to show you what I've learned 15:24:12 <bmbouter> x9c4: feel free to mark as reassigned to you 15:24:15 <x9c4> sure. 15:24:26 <bmbouter> ok we can coordinate maybe after open/floor triage 15:24:28 <ipanova> bmbouter: x9c4 i have asked davidn about this issue 15:24:42 <ipanova> he said this won't block them from integration with pulp_container 15:24:54 <bmbouter> ok well then we're safe in two ways 15:24:59 <bmbouter> x9c4: are you still good to try to fix it? 15:25:09 <x9c4> yes 15:25:18 <bmbouter> ipanova: and should we remove its blocker status from 3.10? 15:25:24 <bmbouter> x9c4: ty 15:25:26 <x9c4> Let's still meet later. 15:25:29 <ipanova> i think so 15:25:34 <bmbouter> x9c4: agreed 15:25:43 <bmbouter> ok then this addresses my concerns I think we're on track for thrusday 15:25:49 <ggainey> woot 15:25:50 <ipanova> yes 15:26:09 <x9c4> +1 15:26:15 <ttereshc> alright, let me know if something changes in the next 1.5 days 15:26:21 <ggainey> heh 15:26:44 <ttereshc> thanks everyone for the discussion 15:27:02 <ttereshc> #endmeeting 15:27:02 <ttereshc> !end