14:30:01 <ttereshc> #startmeeting Pulp Triage 2017-10-06 14:30:01 <ttereshc> #info ttereshc has joined triage 14:30:02 <pulpbot> Meeting started Fri Oct 6 14:30:01 2017 UTC and is due to finish in 60 minutes. The chair is ttereshc. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:30:02 <pulpbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:30:02 <pulpbot> The meeting name has been set to 'pulp_triage_2017_10_06' 14:30:02 <pulpbot> ttereshc: ttereshc has joined triage 14:30:08 <dalley> !here 14:30:08 <dalley> #info dalley has joined triage 14:30:08 <pulpbot> dalley: dalley has joined triage 14:30:11 <ipanova> !here 14:30:11 <ipanova> #info ipanova has joined triage 14:30:11 <pulpbot> ipanova: ipanova has joined triage 14:30:13 <daviddavis> !here 14:30:13 <daviddavis> #info daviddavis has joined triage 14:30:13 <pulpbot> daviddavis: daviddavis has joined triage 14:30:17 <ttereshc> !next 14:30:18 <pulpbot> ttereshc: 5 issues left to triage: 3042, 3049, 3052, 3053, 3056 14:30:18 <ttereshc> #topic Scheduled tasks break GET calls to /pulp/api/v2/tasks/ - http://pulp.plan.io/issues/3042 14:30:19 <pulpbot> Issue #3042 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:30:20 <pulpbot> Scheduled tasks break GET calls to /pulp/api/v2/tasks/ - http://pulp.plan.io/issues/3042 14:30:24 <jortel> !here 14:30:24 <jortel> #info jortel has joined triage 14:30:25 <pulpbot> jortel: jortel has joined triage 14:31:04 <dkliban> !here 14:31:04 <dkliban> #info dkliban has joined triage 14:31:04 <pulpbot> dkliban: dkliban has joined triage 14:31:19 <ipanova> i think we should re-define the understaning of what 'blocker' is :) anyone tried to reproduce this? 14:31:30 <bmbouter> !here 14:31:30 <bmbouter> #info bmbouter has joined triage 14:31:30 <pulpbot> bmbouter: bmbouter has joined triage 14:31:49 <mhrivnak> !here 14:31:49 <mhrivnak> #info mhrivnak has joined triage 14:31:49 <pulpbot> mhrivnak: mhrivnak has joined triage 14:32:11 <mansari> !here 14:32:11 <mansari> #info mansari has joined triage 14:32:11 <pulpbot> mansari: mansari has joined triage 14:32:24 <dkliban> so i am not sure that this is an area of pulp is worth improving 14:32:57 <mhrivnak> Agreed. 14:32:57 <ttereshc> +1 14:33:02 <bmbouter> I also agree 14:33:24 <bmbouter> I also acknowledge what QE is saying here 14:33:35 <ttereshc> Should we discuss priorities in general with QE at some point? So we are on the same page 14:33:41 <mhrivnak> We can take this as a known issue. 14:33:47 <bmbouter> +1 as known issue 14:33:58 <mhrivnak> And yes, we should discuss priorities with them especially around this area. 14:34:01 <bmbouter> agreed 14:34:06 <jortel> agreed 14:34:19 <ipanova> i think what he tries to point out is that once you schedule a task then you are not able to list tasks in general 14:34:33 <ttereshc> !propose other accept + tterashc will put a comment 14:34:33 <ttereshc> #idea Proposed for #3042: accept + tterashc will put a comment 14:34:33 <pulpbot> ttereshc: Proposed for #3042: accept + tterashc will put a comment 14:34:36 <dalley> that seems problematic 14:35:21 <ttereshc> ipanova, any objections to accept? 14:35:27 <mhrivnak> That is a good point. It would be nice to make sure this isn't a regression. 14:35:27 <ipanova> nope 14:35:33 <mhrivnak> If it is, it could impact katello and others. 14:35:45 <ipanova> let me reproduce this and then close? 14:35:59 <ttereshc> !propose other accept + ipanova will try to reproduce 14:35:59 <ttereshc> #idea Proposed for #3042: accept + ipanova will try to reproduce 14:35:59 <pulpbot> ttereshc: Proposed for #3042: accept + ipanova will try to reproduce 14:36:03 <ttereshc> !accept 14:36:03 <ttereshc> #agreed accept + ipanova will try to reproduce 14:36:03 <pulpbot> ttereshc: Current proposal accepted: accept + ipanova will try to reproduce 14:36:04 <mhrivnak> Otherwise I wonder if it's a very specific circumstance causing something unexpected to be in the TaskStatus object that isn't serializable. 14:36:05 <ttereshc> #topic Sync Task hangs forever if wrong proxy setting provided - http://pulp.plan.io/issues/3049 14:36:05 <pulpbot> ttereshc: 4 issues left to triage: 3049, 3052, 3053, 3056 14:36:06 <pulpbot> Issue #3049 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:36:07 <pulpbot> Sync Task hangs forever if wrong proxy setting provided - http://pulp.plan.io/issues/3049 14:37:42 <daviddavis> accept as is? 14:37:57 <mhrivnak> That seems reasonable. 14:38:04 <jortel> +1 14:38:06 <dalley> +1 14:38:08 <mansari> +1 14:38:09 <ttereshc> !propose accept 14:38:09 <ttereshc> #idea Proposed for #3049: Leave the issue as-is, accepting its current state. 14:38:09 <pulpbot> ttereshc: Proposed for #3049: Leave the issue as-is, accepting its current state. 14:38:12 <mhrivnak> I'm interested to reproduce it to figure out what part of the connection process is hanging. 14:38:21 <mhrivnak> But not interested enough to make it a big priority. :) 14:38:29 <ttereshc> !accept 14:38:29 <ttereshc> #agreed Leave the issue as-is, accepting its current state. 14:38:29 <pulpbot> ttereshc: Current proposal accepted: Leave the issue as-is, accepting its current state. 14:38:30 <pulpbot> ttereshc: 3 issues left to triage: 3052, 3053, 3056 14:38:31 <ttereshc> #topic Default fields download_policy and sync_mode on importer create - http://pulp.plan.io/issues/3052 14:38:31 <pulpbot> Issue #3052 [NEW] (unassigned) - Priority: Normal | Severity: Low 14:38:32 <pulpbot> Default fields download_policy and sync_mode on importer create - http://pulp.plan.io/issues/3052 14:39:10 <mhrivnak> rewrite as a story? 14:39:14 <daviddavis> +1 14:39:29 <asmacdo> sure. i also suggested that sync_mode should not have a default 14:39:35 <ttereshc> !propose other convert to a story 14:39:35 <ttereshc> #idea Proposed for #3052: convert to a story 14:39:35 <pulpbot> ttereshc: Proposed for #3052: convert to a story 14:39:45 <ttereshc> I think we can discuss it before grooming 14:39:47 <bmbouter> +1 rewrite as story 14:39:50 <ttereshc> on the issue 14:39:54 <mhrivnak> Sounds good. 14:39:59 <ttereshc> !accept 14:39:59 <ttereshc> #agreed convert to a story 14:39:59 <pulpbot> ttereshc: Current proposal accepted: convert to a story 14:40:01 <ttereshc> #topic pclean does not remove migrations in the pulp_app or plugin projects - http://pulp.plan.io/issues/3053 14:40:01 <pulpbot> ttereshc: 2 issues left to triage: 3053, 3056 14:40:02 <pulpbot> Issue #3053 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:40:03 <pulpbot> pclean does not remove migrations in the pulp_app or plugin projects - http://pulp.plan.io/issues/3053 14:40:17 <bmbouter> oh man this happened to me 3 times! 14:40:50 <asmacdo> so, pclean just runs migrations, but doesnt make them 14:40:53 <bmbouter> doesn't pclean make migrations currently? 14:41:01 <daviddavis> I don't think so 14:41:09 <bmbouter> I'm not really sure it's what I remembered 14:41:14 <asmacdo> there isnt a nice way to make migrations since we have to specify each app 14:41:29 <asmacdo> and aliases cant know which plugins to do 14:41:45 <bmbouter> oh right 14:42:06 <dkliban> it can be smart and figure out which plugins are installed 14:42:11 <mhrivnak> Is a core problem here that if you change the data model and forget about migrations, your running code won't match the DB's schema? 14:42:33 <daviddavis> I believe so 14:42:35 <dkliban> mhrivnak: yeah 14:42:36 <daviddavis> we should consider just checking in migrations into source code to fix this. would be good to get in the habit of updating migrations when we update models. 14:42:40 <bmbouter> no it's that after you vagrant up it was broken for me 14:42:52 <bmbouter> I think it's from changing a git branch and the migrations are leftover 14:43:00 <mhrivnak> Or I guess since we don't have migrations checked in, if *anyone* changes the data layer then everyone has to worry about migrations? 14:43:08 <mhrivnak> Yeah ok, that makes sense. 14:43:13 <daviddavis> mhrivnak: yes exactly 14:43:19 <bmbouter> I know how to work around it so it's not a big one for me 14:43:21 <daviddavis> bmbouter: you sure you didn't do a git pull? 14:43:27 <bmbouter> I did git pull 14:43:37 <bmbouter> then changed the branch 14:43:37 <daviddavis> it's possible that the git pull changed your models then 14:43:48 <bmbouter> yeah that is def what happened 14:43:51 <asmacdo> bmbouter: how about a new dev alias to remove and recreate migrations 14:44:05 <asmacdo> s/function/alias 14:44:08 <bmbouter> yeah but it also can't know about the plugins 14:44:19 <bmbouter> I had to remove both the pulp_app and pulp_example migrations 14:44:42 <asmacdo> if it took args, you could list any plugins installed 14:44:58 <bmbouter> so even though its a bit annoying I can fix this in a few seconds now and since it's temporary I think we should close WONTFIX and identify the workaround 14:45:08 <asmacdo> pulp-manager remakemigrations pulp_app pulp_file pulp_example 14:45:18 <bmbouter> I would also be ok w/ that 14:45:18 <asmacdo> thats fine by me also 14:45:20 <bmbouter> haha 14:45:23 <bmbouter> what asmacdo said 14:45:33 <bmbouter> would that be helpful later on? 14:45:37 <bmbouter> if so then that would be best 14:45:44 <asmacdo> it would be helpful until we GA 14:45:57 <bmbouter> can we get some input on others on this? 14:46:18 <asmacdo> even after we commit migrations, we will probably want to blow them away and start over so we have 1 set of initial migrations 14:46:30 <ttereshc> it's not that annoying for me personally, I'm fine with workaround as well 14:46:48 <daviddavis> I don't think we'll be redoing migrations after we commit them or at least I hope not 14:47:08 <asmacdo> +1 WONT FIX +1/2 new dev functions 14:47:19 <daviddavis> +1 WONTFIX from me 14:47:29 <dalley> +1 14:47:33 <jortel> +1 14:47:34 <ttereshc> any volunteers to provide workaround in comments? 14:47:47 <asmacdo> I'll do it 14:47:51 <bmbouter> +1 to all this 14:48:00 <ttereshc> !propose other close as wontfix and asmacdo will provide workaround 14:48:00 <ttereshc> #idea Proposed for #3053: close as wontfix and asmacdo will provide workaround 14:48:00 <pulpbot> ttereshc: Proposed for #3053: close as wontfix and asmacdo will provide workaround 14:48:02 <bmbouter> yeah it's also not annoying to me after the first time it happened 14:48:08 <ttereshc> !accept 14:48:08 <ttereshc> #agreed close as wontfix and asmacdo will provide workaround 14:48:08 <pulpbot> ttereshc: Current proposal accepted: close as wontfix and asmacdo will provide workaround 14:48:09 <pulpbot> ttereshc: 1 issues left to triage: 3056 14:48:10 <ttereshc> #topic rapidly dispatching sync of the same repo causes resource manager to print traceback - http://pulp.plan.io/issues/3056 14:48:10 <pulpbot> Issue #3056 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:48:11 <ttereshc> thanks asmacdo 14:48:12 <pulpbot> rapidly dispatching sync of the same repo causes resource manager to print traceback - http://pulp.plan.io/issues/3056 14:48:29 <mhrivnak> It looks like we're missing a uniqueness constraint on the model. 14:49:38 <mhrivnak> Might be an easy fix, but might also require a quick audit to make sure all code that creates them is prepared for a dup key error. 14:50:01 <bmbouter> agreed 14:50:04 <mhrivnak> In any case, it definitely needs to be fixed. 14:50:06 <bmbouter> I think we should take this onto the sprint 14:50:09 <ttereshc> +1 14:50:12 <dkliban> +1 14:50:25 <daviddavis> +1 14:50:26 <ttereshc> !propose other accept + add to sprint 14:50:26 <ttereshc> #idea Proposed for #3056: accept + add to sprint 14:50:26 <pulpbot> ttereshc: Proposed for #3056: accept + add to sprint 14:50:26 <mhrivnak> !propose other high high and add to sprint 14:50:35 <jortel> +1 14:50:35 <mhrivnak> heh 14:50:48 <ttereshc> let's try again 14:50:51 <ttereshc> !propose other high high and add to sprint 14:50:51 <ttereshc> #idea Proposed for #3056: high high and add to sprint 14:50:51 <pulpbot> ttereshc: Proposed for #3056: high high and add to sprint 14:50:55 <ttereshc> !accept 14:50:55 <ttereshc> #agreed high high and add to sprint 14:50:55 <pulpbot> ttereshc: Current proposal accepted: high high and add to sprint 14:50:56 <pulpbot> ttereshc: No issues to triage. 14:50:59 <ttereshc> !end 14:50:59 <ttereshc> #endmeeting