14:31:14 #startmeeting Pulp Triage 2016-05-20 14:31:14 #info mhrivnak has joined triage 14:31:15 Meeting started Fri May 20 14:31:14 2016 UTC and is due to finish in 60 minutes. The chair is mhrivnak. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:31:15 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:31:15 The meeting name has been set to 'pulp_triage_2016_05_20' 14:31:15 mhrivnak has joined triage 14:31:33 #info ttereshc has joined triage 14:31:33 !here 14:31:33 ttereshc has joined triage 14:31:42 #info pcreech has joined triage 14:31:42 !here 14:31:42 pcreech has joined triage 14:31:43 !here 14:31:43 #info ipanova has joined triage 14:31:43 ipanova has joined triage 14:32:00 !here 14:32:00 #info jcline has joined triage 14:32:01 jcline has joined triage 14:32:08 !next 14:32:10 18 issues left to triage: 1088, 1893, 1903, 1916, 1918, 1919, 1921, 1922, 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 14:32:10 #topic Incremental publish could use the wrong primary.xml during fast-forward publish - http://pulp.plan.io/issues/1088 14:32:11 RPM Support Issue #1088 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:32:12 Incremental publish could use the wrong primary.xml during fast-forward publish - http://pulp.plan.io/issues/1088 14:32:23 !here 14:32:23 #info asmacdo has joined triage 14:32:23 asmacdo has joined triage 14:33:06 #info preethi has joined triage 14:33:06 !here 14:33:06 preethi has joined triage 14:33:09 Questions? Comments? 14:33:21 This one looks ok to me.. Since it's already on the next sprint and looks to be up to doate 14:33:23 we discussed this one recently. 14:33:48 We even put it on the next sprint, it seems like it's behaving like high priority already 14:33:58 severity is probably low 14:34:37 #idea Proposed for #1088: Priority: Normal, Severity: Low 14:34:37 !propose triage medium low 14:34:37 Proposed for #1088: Priority: Normal, Severity: Low 14:34:42 Yeah, probably, although as we said yesterday it _might_ be responsible for mysterious bugs 14:34:51 I'm fine with medium low though 14:35:02 !here 14:35:02 #info bowlofeggs has joined triage 14:35:03 bowlofeggs has joined triage 14:35:27 Yeah, i think yesterday we did say it could be more importnat than it looks on the surface.. 14:35:29 Any last comments? 14:35:58 Being on the sprint means it'll get attention soon. 14:36:34 It *could* be, but we don't know if it is. So it's getting done, but severity is actual impact. As far as we know, this is hurting anything. 14:36:44 We still won't really know, even after fixing it, if it was hurting anything. 14:37:00 !here 14:37:00 #info jortel has joined triage 14:37:01 jortel has joined triage 14:37:08 any objections or new proposals? Otherwise I'll accept. 14:37:14 Sounds good to me 14:37:15 let's triage it 14:37:18 #agreed Priority: Normal, Severity: Low 14:37:18 !accept 14:37:18 Current proposal accepted: Priority: Normal, Severity: Low 14:37:20 #topic pulp_docker does not support Docker < 1.10 with manifests that were published from Docker >= 1.10 - http://pulp.plan.io/issues/1893 14:37:20 17 issues left to triage: 1893, 1903, 1916, 1918, 1919, 1921, 1922, 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 14:37:21 Docker Support Issue #1893 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:37:22 pulp_docker does not support Docker < 1.10 with manifests that were published from Docker >= 1.10 - http://pulp.plan.io/issues/1893 14:37:52 Did we get info on this from RCM yet? 14:38:08 Not that I know of. Who was following that? 14:38:22 bowlofeggs ? 14:38:33 bowlofeggs, am i correct in saying that this *only* affects RCM? 14:38:43 no this will affect everyone 14:38:52 agreed, it could affect any user. 14:38:53 the question is "how much" and "when" 14:39:02 isnt v1 basically gone? 14:39:08 it's not about v1/v2 14:39:16 it's about v2 schema 1 and v2 schema 2 14:39:29 i haven't heard from rcm about this though 14:39:36 other than their initial report 14:39:38 Who on their side was going to do the testing? 14:39:39 it's def important 14:39:47 not sure 14:40:03 +1 to deferring again till we can get more information on this 14:40:15 Who reported it to us? 14:40:21 (as that was the decision last week) 14:40:30 Or any suggestion on who we should follow up with? 14:40:37 i just wrote twaugh to ask 14:40:46 Great, thanks. 14:40:51 !skip 14:40:53 #topic RPM import traceback (non-utf-8 metadata slipping through) - http://pulp.plan.io/issues/1903 14:40:53 16 issues left to triage: 1903, 1916, 1918, 1919, 1921, 1922, 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 14:40:54 Pulp Issue #1903 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:40:55 RPM import traceback (non-utf-8 metadata slipping through) - http://pulp.plan.io/issues/1903 14:42:56 This only affects rpms that don't conform to the utf-8 standard requirement. 14:43:17 Doesn't that mean they are't rpms? 14:43:25 We've seen a handful of old SUSE packages in that category in the past. 14:43:29 smyers: lol 14:43:30 Similar to how malformed errata shouldn't be accepted as errata? 14:43:48 yeah this ticket should just be about better error reporting 14:44:04 "this RPM didn't use UTF-8, please work with your vendor to correct this." 14:44:16 I tend to agree. I also looked and the current behaviour for upload is to blindly encode to latin-1, which, as bowlofeggs noted in that issue, is dangerous and incorrect. 14:44:29 +1 14:44:31 !propose low low 14:44:32 Error: "propose" is not a valid command. 14:44:32 It is also what createrepo_c does. 14:44:34 FWIW 14:44:34 dammit 14:44:42 #idea Proposed for #1903: Priority: Low, Severity: Low 14:44:42 #info smyers has joined triage 14:44:42 !propose triage low low 14:44:42 smyers has joined triage 14:44:43 Proposed for #1903: Priority: Low, Severity: Low 14:44:51 Waht does createrepo_c do? 14:44:52 But we're drifting into how to fix it, rather than establishing priority. 14:45:06 createrepo and createrepo_c both will fall back to encoding as latin1 14:45:27 citing other bad behavior is still [citation-needed] 14:45:39 this is why python 3 was created without strings 14:45:57 or really just "this is why python 3 was created" 14:46:01 this being "this problem" 14:46:02 But, again, if we agree that non-utf-8 rpms are malformed, we need to fix the error message and not mangle strings 14:46:08 yep 14:46:46 any other thoughts on the priority? 14:46:48 yep 14:46:54 +1 to low low 14:47:10 low/low 14:47:13 how low can you go 14:47:14 #agreed Priority: Low, Severity: Low 14:47:14 !accept 14:47:14 Current proposal accepted: Priority: Low, Severity: Low 14:47:16 15 issues left to triage: 1916, 1918, 1919, 1921, 1922, 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 14:47:17 #topic Unable to install packages with `.` in the name on pip==8.1.2 - http://pulp.plan.io/issues/1916 14:47:17 Python Support Issue #1916 [NEW] (unassigned) - Priority: Normal | Severity: Urgent 14:47:18 Unable to install packages with `.` in the name on pip==8.1.2 - http://pulp.plan.io/issues/1916 14:47:29 he said this wasn't an issue and we should close 14:47:49 +1 14:48:06 / pulpbot 14:48:29 #idea Proposed for #1916: closed notabug 14:48:29 !propose other closed notabug 14:48:29 Proposed for #1916: closed notabug 14:48:46 smyers, will !accept cause pulpbot to close it? 14:49:34 pulpbot does not touch redmine except in a read-only way 14:49:34 Error: "does" is not a valid command. 14:49:36 heh 14:49:48 pulpbot echo I don't make changes to redmine. 14:49:48 I don't make changes to redmine. 14:50:08 Gotcha. 14:50:17 #agreed closed notabug 14:50:17 !accept 14:50:17 Current proposal accepted: closed notabug 14:50:19 14 issues left to triage: 1918, 1919, 1921, 1922, 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 14:50:19 #topic pulp depends on m2crypto, which will not be ported to Python 3 - http://pulp.plan.io/issues/1918 14:50:20 Pulp Issue #1918 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:50:21 pulp depends on m2crypto, which will not be ported to Python 3 - http://pulp.plan.io/issues/1918 14:50:31 Because giving IRC bots the ability to mess with our redmine issues is *terrifying* to me. :) 14:50:47 !propose high med 14:50:47 Error: "propose" is not a valid command. 14:50:55 daaaaaaammit I wrote the damn thing 14:50:59 !propose triage high med 14:50:59 #idea Proposed for #1918: Priority: High, Severity: Medium 14:51:00 Proposed for #1918: Priority: High, Severity: Medium 14:51:10 This might even be an easay fix 14:51:21 +1 we need to start prioritizing py3 14:51:25 smyers, I looked a bit and we use m2crypto in the consumer stuff I think 14:51:36 and m2 is no bueno anyway 14:51:38 !care 1928 14:51:48 hm 14:51:48 also in the bindings. We probably need to drop py 2.4 before we can do this. 14:51:49 Everywhere else is an easy fix, but python-cryptography is 2.6+ 14:52:00 Right 14:52:02 SCLs woo 14:52:11 seems this should be a story or task 14:52:17 yeah SCLs - we can still support EL5 that way 14:52:45 agreed on high/med 14:52:52 or even high/low 14:52:59 We also won't have a need to support el5 once we get to pulp 3.0. 14:53:03 Actually there might not be el5 scls :( 14:53:13 https://en.wikipedia.org/wiki/High/Low_%28Nada_Surf_album%29 14:53:14 Title: High/Low (Nada Surf album) - Wikipedia, the free encyclopedia (at en.wikipedia.org) 14:53:15 jcline: ☹ 14:53:19 Either way, High something. 14:53:40 it's not severe… yet 14:53:43 but it will be! 14:53:47 To jortel's point, I actually can't decide if it's an Issue, Story, or Task. 14:53:49 M2crypto is pretty much dead anyway. I would not be surprised if some were CVEs against it. 14:53:54 py 2's days are numbered 14:54:07 Seems like an enhancement until we move to py3 14:54:12 Oh good point. Refactor maybe? 14:54:12 But then it's definitely a bug? 14:54:13 right. bug=defect and this is not a defect. 14:54:14 i say not a story, but a refactor makes sense 14:54:19 it's not a feature 14:54:27 Refactor works 14:54:40 No behavior change, just the internals, seems perfect 14:54:41 I'm sure I could find defects in m2crypto, but I'm fine with whatever classification people want 14:54:44 it will be an issue if it's not done in time 14:54:49 so there's that aspect to it 14:55:05 seems like a task 14:55:07 and "in time" is sooner than you may think 14:55:18 !propose other High Priority still, but change to Refactor or Task tracker 14:55:18 #idea Proposed for #1918: High Priority still, but change to Refactor or Task tracker 14:55:19 Proposed for #1918: High Priority still, but change to Refactor or Task tracker 14:55:30 I wrote that weirdly. Go me. 14:55:40 jortel, why task instead of refactor? 14:55:52 let's move on rather than debate task vs. refactor 14:55:57 I'm not sure it makes a lot of difference, but am curious. :) 14:56:30 Ok, there are 3 suggestions for refactor, so I'll go with that for now and move on. 14:56:41 refactor seems more appropriate for changing how pulp does something but selecting a different lib seems like a task. I'm fine with refactor. 14:56:50 #agreed High Priority still, but change to Refactor or Task tracker 14:56:50 !accept 14:56:50 Current proposal accepted: High Priority still, but change to Refactor or Task tracker 14:56:52 13 issues left to triage: 1919, 1921, 1922, 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 14:56:53 #topic pulp depends on python-ldap, which will not be ported to Python 3 - http://pulp.plan.io/issues/1919 14:56:53 Pulp Issue #1919 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:56:54 pulp depends on python-ldap, which will not be ported to Python 3 - http://pulp.plan.io/issues/1919 14:57:08 same 14:57:24 #idea Proposed for #1919: High Priority still, but change to Refactor or Task tracker 14:57:24 !propose other High Priority still, but change to Refactor or Task tracker 14:57:24 Proposed for #1919: High Priority still, but change to Refactor or Task tracker 14:57:27 !suggest Consensus for #1918 was to use the refactor tracker 14:57:27 #idea Consensus for #1918 was to use the refactor tracker 14:57:36 !accept 14:57:37 #agreed High Priority still, but change to Refactor or Task tracker 14:57:37 Current proposal accepted: High Priority still, but change to Refactor or Task tracker 14:57:38 #topic pulp-rpm depends on pyliblzma, which will not be ported to Python 3 - http://pulp.plan.io/issues/1921 14:57:39 12 issues left to triage: 1921, 1922, 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 14:57:40 RPM Support Issue #1921 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:57:41 pulp-rpm depends on pyliblzma, which will not be ported to Python 3 - http://pulp.plan.io/issues/1921 14:57:44 !propose other High Priority still, but change to Refactor or Task tracker 14:57:44 #idea Proposed for #1921: High Priority still, but change to Refactor or Task tracker 14:57:45 Proposed for #1921: High Priority still, but change to Refactor or Task tracker 14:57:46 !suggest Consensus for #1919 was also to use the refactor tracker 14:57:46 #idea Consensus for #1919 was also to use the refactor tracker 14:57:55 :) 14:57:56 #agreed High Priority still, but change to Refactor or Task tracker 14:57:56 !accept 14:57:56 Current proposal accepted: High Priority still, but change to Refactor or Task tracker 14:57:56 haha 14:57:58 11 issues left to triage: 1922, 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 14:57:58 #topic unhandled exception when changing feed to equivalent repo - http://pulp.plan.io/issues/1922 14:57:59 RPM Support Issue #1922 [NEW] (unassigned) - Priority: Normal | Severity: Medium 14:58:00 unhandled exception when changing feed to equivalent repo - http://pulp.plan.io/issues/1922 14:58:11 !suggest Consensus for #1921 was also to use the refactor tracker 14:58:11 #idea Consensus for #1921 was also to use the refactor tracker 14:58:15 hahaha 14:58:16 sorry smyers 14:58:20 gotta get those minutes lookin' good 14:58:58 This issue hurts my brain. 14:59:03 For the record. 14:59:12 !suggest This issue hurts my brain. 14:59:12 #idea This issue hurts my brain. 14:59:22 (Now it' son the record) 14:59:37 Crap, I forgot to note who reported this to me. I'm pretty sure it was omaciel . 15:00:05 So is the solution to just overwrite the link if it already exists? 15:00:11 "cloud-qe" seems to back up your guess mhrivnak 15:00:22 I don't get what's going enough well enough to triage it. 15:00:42 does seem like a use case that should work, though not sure how important it is 15:00:56 !propose triage med med 15:00:56 #idea Proposed for #1922: Priority: Normal, Severity: Medium 15:00:57 Proposed for #1922: Priority: Normal, Severity: Medium 15:01:03 I think it's basically saying if there are 2 mirrors of the same content, and you switch the feed from one to the other, it breaks. 15:01:20 And in the case where that was seen, both mirrors happened to also be pulps. 15:01:42 could be that pulp doesn't write the data consistently in publish 15:01:47 seems like higher than med/med 15:01:52 because the CDN is mirrors all the way down 15:01:55 and that works 15:02:14 jortel: only if there's a real use case for this 15:02:15 Indeed, we know for sure the two pulps will have slightly different data in repodata/ 15:02:18 which i don't know of 15:02:46 without a use case, i might even argue low/med 15:02:52 Yeah, that's my feeling as well. 15:02:53 There are people using multiple pulps behind a load balancer, and active work to improve that use case. 15:03:08 yeah that's a use case 15:03:43 !propose med high 15:03:43 Error: "propose" is not a valid command. 15:03:46 So Pulp A and Pulp B are behind a load balancer, something else is keeping them synced, Pulp C syncs from the load balancer 15:03:50 !propose triage med high 15:03:50 #idea Proposed for #1922: Priority: Normal, Severity: High 15:03:50 Alright 15:03:50 Proposed for #1922: Priority: Normal, Severity: High 15:03:52 I like it 15:04:01 Makes sense now :) 15:04:03 works for me 15:04:04 !accept 15:04:04 Current proposal accepted: Priority: Normal, Severity: High 15:04:04 #agreed Priority: Normal, Severity: High 15:04:06 #topic POST /pulp/api/v2/content/actions/delete_orphans/ is broken - http://pulp.plan.io/issues/1923 15:04:06 10 issues left to triage: 1923, 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 15:04:07 Pulp Issue #1923 [NEW] (unassigned) - Priority: Normal | Severity: Medium 15:04:08 POST /pulp/api/v2/content/actions/delete_orphans/ is broken - http://pulp.plan.io/issues/1923 15:04:21 actually, putting two pulps behind a load balancer that aren't using NFS is not supported 15:04:26 so i don't think that's a real use case 15:04:31 i suggest going back to that last one 15:04:34 if we can 15:05:04 they would be using NFS and would be identical 15:05:15 I think the best way is to bring it up after we get through the list? 15:05:29 sure 15:05:30 #1923 sounds like a regression and bad 15:05:35 sounds good. 15:05:42 agreed on #1923 15:06:01 +1 15:06:09 smyers, i wouldnt say its that bad 15:06:18 It's a regression. 15:06:20 There is a clear workaround it seems. 15:06:21 there is another API endpoint that does the same thing 15:06:24 Regressions are bad. 15:06:29 it doesn't say how it's broken though right? 15:06:56 yeah regressions are bad 15:06:57 +1 15:06:59 Having two API endpoints that do the same thing is also bad, but that's not justification to break one of them. 15:07:19 I don't think anyone is arguing it's justified to break one. 15:07:33 So...we should fix this one. 15:07:35 Immediately. 15:07:36 yeah 15:07:47 But knowing that there's a workaround for an issue often affects the severity decision. 15:07:48 !propose triage med high 15:07:48 #idea Proposed for #1923: Priority: Normal, Severity: High 15:07:49 Proposed for #1923: Priority: Normal, Severity: High 15:07:52 !propose triage Urgent High 2.8.4 15:07:52 Error: Unknown Severity 15:08:05 Iiiinteresting. 15:08:16 i could do urgent i supose ☺ 15:08:21 with two p's even 15:08:21 Not having a workaround would make it Urgent. Having a workaround means it's high severity. 15:08:23 on this is interesting 15:08:39 it may have been broken for a very very long time 15:08:48 That's believable. 15:08:49 i should have filed a bug on this, but i saw it jan 2015 15:08:51 https://github.com/pulp/pulp/pull/1579#issuecomment-71904003 15:08:53 Title: django conversion /v2/content/ by asmacdo · Pull Request #1579 · pulp/pulp · GitHub (at github.com) 15:09:42 i'm ok with any of the proposed priorities 15:09:55 Ya, I also don't feel strongly about any of them. 15:09:55 point is we should fix it 15:10:17 smyers seems to feel strongly, so i say we go with his proposal 15:10:18 I'm proposing it as a blocker for the next bugfix release (Urgent Prio) because it's a backward-incompabile API change. 15:10:20 ill add a comment that it should probably point to the same view as the working endpoint 15:10:39 So we should either release pulp 3 or fix it. 15:10:46 +1 15:11:02 we do have 2.2 users about to upgrade too ☺ 15:11:33 #idea Proposed for #1923: urgent medium 2.8.4 15:11:33 !propose other urgent medium 2.8.4 15:11:33 Proposed for #1923: urgent medium 2.8.4 15:11:41 Is that what we're thinking? 15:11:47 medium sev is good, yeah 15:11:55 "Partial but non-critical functionality loss, or other loss of functionality where users are still able to perform their critical tasks." 15:12:04 i.e. "Use the other endpoint" 15:12:34 Would you agree that if it was broken in 2.7 also, the priority might be lower? 15:12:39 Nope. 15:12:46 Nope++ 15:12:46 Nope's karma is now 1 15:12:50 heh 15:12:58 The moment the regression is discovered, it should immediately be recitifed. 15:13:11 !accept 15:13:11 #agreed urgent medium 2.8.4 15:13:11 Current proposal accepted: urgent medium 2.8.4 15:13:13 9 issues left to triage: 1924, 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 15:13:14 #topic pulp-server and python-pulp-streamer rpms both provide the pulp_streamer.service systemd unit - http://pulp.plan.io/issues/1924 15:13:14 Packaging Issue #1924 [NEW] (unassigned) - Priority: Normal | Severity: Medium 15:13:15 pulp-server and python-pulp-streamer rpms both provide the pulp_streamer.service systemd unit - http://pulp.plan.io/issues/1924 15:14:12 Seems bad. What's the impact on a user? 15:15:12 ichimonji10: ++ 15:15:12 ichimonji10: 's karma is now 3 15:15:19 ipanova: ++ 15:15:19 ipanova: 's karma is now 3 15:15:21 teamwork++ 15:15:21 teamwork's karma is now 4 15:16:01 !propose skip 15:16:01 #idea Proposed for #1924: Skip this issue for this triage session. 15:16:02 Proposed for #1924: Skip this issue for this triage session. 15:16:15 Why are we skipping? 15:16:17 Unless someone has more info, perhaps let's ask jcline to add some. 15:16:20 Ah, there you are. 15:16:21 :) 15:16:26 What's the impact on a user? 15:16:45 Well, anyone wanting to use the yum file utilities is going to be confused. 15:16:56 It's an easy fix and it's just a packaging problem, it seems 15:17:08 this issue also exists in Fedora ☹ 15:17:09 I'm surprised yum doesn't get upset about installing both, but it doesn't. 15:17:19 Right, I'm also surprised by that. 15:17:21 that is surprising 15:17:28 So it doesn't actually break anything? 15:17:38 I really would've thought it would break everything. 15:17:39 yeah, I'd expect yum/dnf to fail when trying to install both packages 15:17:52 $dnf provides \*pulp_streamer.ser\* 15:17:55 smyers, I would also expect it to break everything. Hasn't so far, but who knows 15:18:00 pulp-server-2.8.2-2.fc25.noarch : The pulp platform server 15:18:01 Repo : rawhide 15:18:03 python2-pulp-streamer-2.8.2-2.fc25.noarch : The pulp lazy streamer 15:18:04 Repo : rawhide 15:18:06 lol wut 15:18:20 I know, right? Seems like maybe there's a bug in dnf/yum that lets this work 15:18:22 But...if it *doesn't*, then it might not be a high severity 15:18:26 #idea Proposed for #1924: Priority: High, Severity: Low 15:18:26 !propose triage high low 15:18:26 Proposed for #1924: Priority: High, Severity: Low 15:18:31 yeah probably just low sev 15:18:34 So yeah, high low's great 15:18:40 Let's fix it and talk with upstream 15:18:43 https://en.wikipedia.org/wiki/High/Low_%28Nada_Surf_album%29!propose 15:18:47 hahaha 15:18:50 haha 15:18:57 btw, that album is really good 15:19:02 !accept 15:19:02 #agreed Priority: High, Severity: Low 15:19:02 Current proposal accepted: Priority: High, Severity: Low 15:19:04 #topic v1 sync errors - http://pulp.plan.io/issues/1925 15:19:04 8 issues left to triage: 1925, 1926, 1928, 1929, 1930, 1931, 1932, 1933 15:19:05 Docker Support Issue #1925 [NEW] (unassigned) - Priority: Normal | Severity: Medium | Target Release: 2.8.4 15:19:06