15:30:08 <asmacdo> #startmeeting Pulp Triage 2019-06-18
15:30:08 <asmacdo> #info asmacdo has joined triage
15:30:08 <asmacdo> !start
15:30:08 <pulpbot> Meeting started Tue Jun 18 15:30:08 2019 UTC.  The chair is asmacdo. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:30:08 <pulpbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:30:08 <pulpbot> The meeting name has been set to 'pulp_triage_2019-06-18'
15:30:08 <pulpbot> asmacdo: asmacdo has joined triage
15:30:16 <daviddavis> #info daviddavis has joined triage
15:30:16 <daviddavis> !here
15:30:16 <pulpbot> daviddavis: daviddavis has joined triage
15:30:19 <bmbouter> #info bmbouter has joined triage
15:30:19 <bmbouter> !here
15:30:19 <pulpbot> bmbouter: bmbouter has joined triage
15:30:21 <asmacdo> !next
15:30:22 <pulpbot> asmacdo: 7 issues left to triage: 4920, 4939, 4947, 4959, 4970, 4979, 4987
15:30:22 <asmacdo> #topic https://pulp.plan.io/issues/4920
15:30:23 <pulpbot> RM 4920 - kersom - ASSIGNED - Collection - Repository versions not being update after successive syncs
15:30:24 <pulpbot> https://pulp.plan.io/issues/4920
15:30:26 <daviddavis> skip
15:30:32 <ttereshc> #info ttereshc has joined triage
15:30:32 <ttereshc> !here
15:30:32 <pulpbot> ttereshc: ttereshc has joined triage
15:30:38 <asmacdo> dkliban: i guess we have more to do
15:30:45 * daviddavis weeps
15:30:46 <asmacdo> !skip
15:30:47 <asmacdo> #topic https://pulp.plan.io/issues/4939
15:30:47 <pulpbot> asmacdo: 6 issues left to triage: 4939, 4947, 4959, 4970, 4979, 4987
15:30:48 <pulpbot> RM 4939 - kersom - NEW - Docs - Collections upload workflows is using role endpoints
15:30:48 * daviddavis weeps
15:30:49 <pulpbot> https://pulp.plan.io/issues/4939
15:30:54 <daviddavis> skip
15:30:56 <asmacdo> !skip
15:30:57 <pulpbot> asmacdo: 5 issues left to triage: 4947, 4959, 4970, 4979, 4987
15:30:57 <asmacdo> #topic https://pulp.plan.io/issues/4947
15:30:58 <pulpbot> RM 4947 - amacdona@redhat.com - NEW -  As a user I can add tags to a repository by name.
15:30:59 <daviddavis> skip
15:30:59 <pulpbot> https://pulp.plan.io/issues/4947
15:31:05 <asmacdo> !skip
15:31:06 <ttereshc> story?
15:31:06 <asmacdo> #topic https://pulp.plan.io/issues/4959
15:31:07 <pulpbot> asmacdo: 4 issues left to triage: 4959, 4970, 4979, 4987
15:31:08 <pulpbot> RM 4959 - amacdona@redhat.com - NEW - Do not log that a download fails with 401 (INFO) unless it fails again after token refresh
15:31:09 <pulpbot> https://pulp.plan.io/issues/4959
15:31:19 <asmacdo> !skip
15:31:20 <pulpbot> asmacdo: 3 issues left to triage: 4970, 4979, 4987
15:31:20 <asmacdo> #topic https://pulp.plan.io/issues/4970
15:31:21 <pulpbot> RM 4970 - kersom - NEW - Collection - RepositoryDistribution does not provide url to consume to cosume content from Pulp
15:31:22 <pulpbot> https://pulp.plan.io/issues/4970
15:31:36 <asmacdo> !skip
15:31:37 <pulpbot> asmacdo: 2 issues left to triage: 4979, 4987
15:31:37 <asmacdo> #topic https://pulp.plan.io/issues/4979
15:31:38 <pulpbot> RM 4979 - dgoetz - NEW - Katello complains during reposync about being unable to update existing errate
15:31:39 <pulpbot> https://pulp.plan.io/issues/4979
15:31:46 <asmacdo> !skip
15:31:47 <pulpbot> asmacdo: 1 issues left to triage: 4987
15:31:47 <asmacdo> #topic https://pulp.plan.io/issues/4987
15:31:48 <pulpbot> RM 4987 - jcabrera - NEW - Pulp 3 is not validating that all required fields are being supplied to the RPM create API
15:31:48 <ggainey> #info ggainey has joined triage
15:31:48 <ggainey> !here
15:31:49 <pulpbot> https://pulp.plan.io/issues/4987
15:31:50 <pulpbot> ggainey: ggainey has joined triage
15:31:57 <dalley> #info dalley has joined triage
15:31:57 <dalley> !here
15:31:57 <pulpbot> dalley: dalley has joined triage
15:32:13 <dawalker> #info dawalker has joined triage
15:32:13 <dawalker> !here
15:32:13 <pulpbot> dawalker: dawalker has joined triage
15:32:14 <dkliban> i am writing a comment about this one
15:32:17 <daviddavis> skip
15:32:19 <daviddavis> it's rpm
15:32:20 <dkliban> but we should accept and add to sprint
15:32:26 <dkliban> okk
15:32:29 <dkliban> sounds good
15:32:31 <asmacdo> !skip
15:32:32 <pulpbot> asmacdo: No issues to triage.
15:32:41 <asmacdo> well, i swear there *was* an issue a minute ago
15:32:50 <dkliban> Open Floor .....
15:32:51 <daviddavis> yea, it was a story actually
15:32:55 <daviddavis> I updated it
15:33:17 <asmacdo> oh ok
15:33:25 <asmacdo> yeah open floo
15:33:46 <ttereshc> :)
15:34:01 <daviddavis> *deafening silence*
15:34:15 <asmacdo> daviddavis: did you want to talk about https://pulp.plan.io/issues/4981
15:34:46 <daviddavis> we can if people want to
15:34:59 <bmbouter> I think I originally brought up that requirement, but I don't think it's 3.0 requirement par-se
15:35:13 <daviddavis> I marked it as 3.0 just in case
15:35:28 <daviddavis> I figured it was easier to remove it later but I am fine with it not being 3.0
15:35:32 <asmacdo> i'd like to have an endpoint for this, not a scheduled job
15:36:58 <asmacdo> and users could hit the endpoint with cron if they want
15:36:58 <bmbouter> asmacdo: that would make sense to me
15:36:59 <daviddavis> should it be part of the orphan namespace in our api?
15:36:59 <bmbouter> overall though I think it's too early since we don't have a user or stakeholder driving it's needs
15:37:03 <bmbouter> just me saying 'maybe we should do this'
15:37:11 <daviddavis> you are the stakeholder bmbouter
15:37:13 <daviddavis> jk
15:37:30 <asmacdo> well, if we "leak" bits, we should provide a way to clean them up
15:37:38 <dkliban> bmbouter is the dude with a steak!
15:37:47 <bmbouter> true but with that argument could be applied to a lot of areas of the 3.0 launch
15:37:53 <bmbouter> asmacdo: ^
15:38:12 <asmacdo> bmbouter: fair enough
15:38:27 <dkliban> i am still curious about the cut of steak bmbouter is holding
15:38:45 <daviddavis> it's a stake (a wooden post) and not a steak
15:38:48 <daviddavis> he's hunting vampires
15:38:49 <bmbouter> asmacdo: are you saying that we should just ensure DELETE works w/ single uploads or as a filterset
15:38:49 <asmacdo> bmbouter: vampire hunter, stake holder
15:38:56 <asmacdo> dang daviddavis you beat me
15:39:04 * bmbouter gets the Chimichurri out
15:39:09 <dkliban> lol
15:39:13 <daviddavis> mmm chimichurri
15:39:22 <daviddavis> it must be almost lunch time
15:39:23 * bmbouter pours chimichurri over pulp
15:39:32 <bmbouter> indeed
15:39:39 <asmacdo> bmbouter: i is there a way to get a list of incomplete uploads?
15:39:53 <asmacdo> if so, then the single upload DELETE would be sufficient for 3.0 IMO
15:39:55 <bmbouter> daviddavis: what did you imagine w.r.t asmacod's quesiton?
15:39:58 <ttereshc> if we don't introduce this cleanup now, would it be easy/possible to remove those incomplete chunked uploads later?
15:40:17 <bmbouter> asmacdo: I think I agree that there needs to be some way for the user to cleanup otherwise it's probably not ok
15:40:36 <daviddavis> asmacdo bmbouter we could add filtering to get a list of incomplete uplaods but currently I don't think there is a way
15:40:37 <bmbouter> ttereshc: I think not and I believe that is asmacdo's concern
15:41:15 <ttereshc> yeah, then I agree with the concern
15:41:23 <bmbouter> me too
15:41:49 <bmbouter> can ew just ensure that the actual upload story allows for a DELETE of an upload?
15:41:53 <bmbouter> as a reoslution
15:42:01 <daviddavis> whcih upload story?
15:42:02 <dkliban> yeah ... i think that's needed at the minimum
15:42:13 <dkliban> daviddavis: that chunked uploads story you wrote yesterday
15:42:22 <daviddavis> dkliban: the parallel upload one?
15:42:32 <dkliban> https://pulp.plan.io/issues/4488
15:42:38 <dkliban> yes
15:42:39 <daviddavis> https://pulp.plan.io/issues/4488
15:42:49 <daviddavis> I don't think it should be part of this story
15:42:57 <dkliban> the delete?
15:43:01 <daviddavis> yea
15:43:09 <daviddavis> 4488 = support for parallel uploads
15:43:14 <asmacdo> it might be worth making the delete story blocked by 4488
15:43:28 <dkliban> if it's not part of it, then yes
15:43:34 <dkliban> i don't mind not making it part of it
15:43:54 <dkliban> 4891 can be follow up work for 4488
15:44:09 <bmbouter> it makes more sense to me to do it together because it sounds like we'll be in a state that people are uncomfortable with if say we release in between these two stories
15:44:28 <daviddavis> I'm not sure I understand
15:44:41 <bmbouter> ttereshc and asmacdo can you restate your concerns
15:44:52 <daviddavis> if people do chunked uploads currently, they will have incomplete uploads
15:45:11 <daviddavis> it seems like deleting uploads should be implemented regardless of whether we do parallel uploads
15:45:24 <asmacdo> daviddavis: i'm just assuming that the current chunked uploads are going to change
15:45:51 <daviddavis> they will but I think of parallel uploads and deleting uploads as two separate pieces of work
15:46:00 <daviddavis> I'm just having trouble imaging them as one I guess
15:46:31 <asmacdo> especially if we are going to use a different library, i figure theres a decent chance that the deletion could be changed too
15:46:53 <daviddavis> I think we're just removing the library
15:47:09 <asmacdo> ok, i dont feel strongly about blocking relationship
15:47:24 <ttereshc> I think it's fine if they are separate stories as long as both go into ga
15:47:30 <asmacdo> +1
15:47:37 <daviddavis> that's my feeling too
15:47:43 <dkliban> what about RC 3?
15:47:49 <bmbouter> that was the concern I heard was rc3
15:47:59 <bmbouter> or interpreted, actually no one said rc3
15:48:23 <asmacdo> it sounds like if we do parallel uploads, that wont change the nature of the cleanup problem
15:48:25 <bmbouter> I'm good w/ whatever you all agree to w.r.t to the landing of these various things
15:48:45 <daviddavis> asmacdo: correct
15:48:52 <daviddavis> it sounds like delete should go into rc3
15:49:06 <daviddavis> when is rc3?
15:49:07 <ttereshc> since we don't provide an upgrade path from rc to ga, I think it's a non-blcoking issue for rc3 but nice-to-have one
15:49:16 <asmacdo> +1 ttereshc
15:49:35 <ttereshc> rc3 is in a week
15:49:39 <ttereshc> right bmbouter ?
15:49:46 <dkliban> sounds good to me
15:49:48 <x9c4> I have issues with the python api bindings: `sync_response = fileremotes.remotes_file_file_sync(file_remote.href, repository_sync_data)`
15:50:00 <dkliban> what's the issue?
15:50:26 <bmbouter> ttereshc: yes rc3 is slated for monday but is movable within next week
15:50:45 <x9c4> fileremotes is already an object, that is closely related to the file/remotes endpoint.
15:50:51 <daviddavis> does someone have time to work on https://pulp.plan.io/issues/4981 by rc3?
15:51:04 <daviddavis> I might be able to do it by monday. not sure.
15:51:14 <x9c4> it seems too much to prefix the `sync` action with `remotes_file_file`
15:51:23 <dkliban> daviddavis: it depends on when 4488 gets done
15:51:33 <daviddavis> oh sorry I meant 4488
15:51:40 <dkliban> lol
15:52:01 <dkliban> daviddavis: i was hoping you would do it ... but if you can't, i will definitely work on it
15:52:11 <dkliban> x9c4: i agree that it is very verbose
15:52:18 <x9c4> Also it prevents generic programming when creating / updating resources (That is the real problem).
15:52:29 <x9c4> ^ ...,deleting
15:52:31 <dkliban> x9c4: what do you mean?
15:52:47 <daviddavis> dkliban: I think I can do it by monday if I can get some help (eg PR reviews)
15:53:06 <ttereshc> daviddavis, dkliban I will be glad to dig in, maybe tomorrow (but if it's in new state, then I'm not on it)
15:53:08 <daviddavis> I'll file the upload delete story. I think it's different than https://pulp.plan.io/issues/4981
15:53:09 <dkliban> daviddavis: i plan on providing PR review ... but i plan to be out all day friday
15:53:27 <asmacdo> thanks daviddavis
15:53:33 <daviddavis> ok, hopefully rc3 gets moved back a day or two
15:53:47 <asmacdo> anything else for open floor?
15:53:51 <daviddavis> if I could get someone to groom https://pulp.plan.io/issues/4488 that'd be great
15:54:13 <ttereshc> daviddavis, I struggle to understand how do you get old incomplete chunked uploads to remove later?
15:54:46 <dkliban> ttereshc: by date of creation
15:54:51 <daviddavis> ttereshc: you can list the uploads and see if they are complete or not
15:54:54 <daviddavis> and then delete them
15:54:55 <x9c4> dkliban: I cannot write a function, that accepts an api-object (like FileRemote, Repository, ...) and let it decide whether to call 'update' or 'create' on it.
15:55:49 <x9c4> Instead, i need to have a copy of that code for every single api-object.
15:56:01 <dkliban> x9c4: i see what you are saying
15:56:34 <dkliban> x9c4: the name of that function is derived from the Operation Id from the OpenAPI schema definition of that endpoint
15:57:09 <asmacdo> #endmeeting
15:57:09 <asmacdo> !end