| morganfainberg | jamielennox: in theory it should be safe (just looking at the change to HTTPClient init) the way it is changed. even with tests referencing it. | 00:01 |
|---|---|---|
| *** adjohn has quit IRC | 00:02 | |
| *** adjohn has joined #openstack-dev | 00:02 | |
| *** danwent has joined #openstack-dev | 00:02 | |
| jamielennox | morganfainberg: so my thought is more I think we should change it, looking at things like the if statement around like 286 - that sort of thing should be pushed up to the overriding cilent | 00:03 |
| bknudson | how do we tell users what's API and what's internal? | 00:03 |
| morganfainberg | jamielennox: beat me to my next question ;) | 00:03 |
| jamielennox | and that there is no reason for anyone (other than our few tests) to make use of the signature of the HTTPClient directly | 00:03 |
| jamielennox | bknudson: the crux of the question - and my main bugbear with python | 00:03 |
| jamielennox | maybe top 5 | 00:03 |
| bknudson | should something in the docstring say it? | 00:04 |
| jamielennox | afaik there is no standard way | 00:04 |
| morganfainberg | bknudson: thats a good idea actually. | 00:04 |
| roaet | Just out of curiosity, how long does it take for the jenkins auto-review to hit a change? | 00:05 |
| morganfainberg | this might be a good ML topic too. | 00:05 |
| jamielennox | but i know i'm doing a refactoring i'll push up later today that is going to remove a function or two that is public but there is no reason anyone *should* ever use it | 00:05 |
| jamielennox | i wsa going to use it as the base for discussion - but this works as well | 00:05 |
| morganfainberg | roaet: you mean after you upload your patchset how long does it take jenkins to +1/-1 it? you can see the status of everything in the queue at http://status.openstack.org/zuul/ it really depends on how many things are pending and where you are in the queue | 00:06 |
| bknudson | one option is to version and spec the keystoneclient like we version and spec the REST API | 00:06 |
| roaet | morganfainberg: thanks, I wasn't aware of that link | 00:06 |
| morganfainberg | roaet: happy to help | 00:06 |
| morganfainberg | bknudson: i would be in favor of that, actually | 00:06 |
| jamielennox | so in a way we do that because we have the V2 and V3 client - unfortuantely this is base | 00:06 |
| morganfainberg | bknudson: it would add a good deal of overhead both on the review/code submission side though | 00:07 |
| bknudson | then we'd leave the current keystoneclient alone and have a keystoneclient/v2 | 00:07 |
| jamielennox | but i do think we should aim for a keystoneclient 1.0 one of these days and this should be figured out before that | 00:07 |
| morganfainberg | jamielennox: the more i'm thinking about that, the more i think base should be just that, very very very basic structures | 00:07 |
| jamielennox | so one of the things that he is doing with this whole apiclient thing is moving us back to the more original idea that the base HTTPClient shouldn't be inheritted by v2 and v3 client instead it should just be the class you communicate through | 00:09 |
| morganfainberg | jamielennox: but by that token, a lot of what is there probably needs to be pulled upwards. | 00:09 |
| jamielennox | this is how nova, quantum etc work and i think how keystone was originally supposed to work | 00:09 |
| morganfainberg | its a big refactor, but i think it is a much cleaner approach. | 00:09 |
| jamielennox | it makes the use of HTTPClient in keystone tests make sense | 00:09 |
| morganfainberg | (i admit i haven't looked at the rest of the code in the changeset) | 00:10 |
| morganfainberg | yet. | 00:10 |
| jamielennox | and it makes it much easier to share an auth token around a number of clients | 00:10 |
| jamielennox | so it is the right thing i think | 00:10 |
| jamielennox | so in general we can't break compatibility for people that use v2_0.Client etc | 00:10 |
| *** zul has joined #openstack-dev | 00:10 | |
| morganfainberg | right | 00:10 |
| *** hailiang has joined #openstack-dev | 00:11 | |
| jamielennox | but i'd really like a way to know what is considered public - because none of our documentation ever suggests just using HTTPClient and i think it would break in most cases | 00:11 |
| *** zul has quit IRC | 00:11 | |
| *** lexinator has quit IRC | 00:11 | |
| jamielennox | but that doesn't mean someone isn't doing it | 00:11 |
| roaet | morganfainberg: Erm. I've noticed that my tox test, and jenkins, fails with pure upstream code. Is that normal? | 00:11 |
| *** lexinator has joined #openstack-dev | 00:12 | |
| jamielennox | roaet: it shouldn't but jenkins is temperamental at the moment | 00:12 |
| morganfainberg | jamielennox: i think this is a ML topic, but I do like the suggestion of the docstring (if there isn't another mechanism), bknudson - any other thoughts? | 00:12 |
| roaet | jamielennox: but my local env is failing too. 2.7 is ok, but 2.6 fails. | 00:12 |
| *** zul has joined #openstack-dev | 00:13 | |
| jamielennox | roaet: from upstream code? no it shouldn't fail - but you might be configured slightly differently | 00:13 |
| *** vipul is now known as vipul-away | 00:13 | |
| morganfainberg | jamielennox: wow, jenkins is being tempermental (watching gate queue atm) | 00:14 |
| jamielennox | roaet: my first step is generally to recreate the virtualenv | 00:14 |
| *** sushils has quit IRC | 00:14 | |
| jamielennox | roaet: then figure out if there is actually something that is unsupported by py2.6 | 00:14 |
| jamielennox | morganfainberg: i've hit this one on about 4 reviews now: https://bugs.launchpad.net/tempest/+bug/1212405 | 00:14 |
| uvirtbot | Launchpad bug 1212405 in horizon "test_basic_scenario fails in the gate with a 500 error" [Critical,Fix committed] | 00:14 |
| roaet | jamielennox: ok. is there a doc somewhere with recommended py26 settings? | 00:14 |
| bknudson | morganfainberg: seems like it's too late too do anything sane. | 00:15 |
| bknudson | the code is out there without docstrings... so we assume everything is public API? | 00:15 |
| jamielennox | roaet: not that i'm aware of, but figure out what test is failing - generally if it works in 2.7 then there is a function being used that isn't available in python2.6 and you have to work around it | 00:16 |
| *** lexinator has quit IRC | 00:16 | |
| morganfainberg | bknudson: and it would be a bear of a task to get it documented and validated… short of some crazy introspection code ot know how things got called | 00:16 |
| morganfainberg | that it is public/private. might be worth the effort, but ugh. it is a bit late. | 00:16 |
| roaet | jamielennox: I'm not sure if my change could possibly have caused 2.6 to begin failing, it was a single line. It appears it was failing on clone. | 00:17 |
| jamielennox | morganfainberg, bknudson: maybe we take these new patches as the start of a 1.0 effort where we docstring public methods | 00:17 |
| jamielennox | roaet: review? | 00:17 |
| roaet | https://review.openstack.org/#/c/42242/ | 00:17 |
| bknudson | there are some docs... which are totally out of date, of course. | 00:18 |
| jamielennox | roaet: ok, so definetly not a 2.6 issue | 00:18 |
| bknudson | https://github.com/openstack/python-keystoneclient/blob/master/doc/source/using-api.rst | 00:18 |
| roaet | jamielennox: Is it possible something upgraded from a dependency and broke the build? | 00:19 |
| bknudson | jamielennox: morganfainberg: maybe we can say that what's in the doc is the public API. | 00:20 |
| *** jasdeepH has quit IRC | 00:20 | |
| jamielennox | bknudson: so everything in there is suggesting using the v2 client | 00:20 |
| bknudson | it could use some updating. | 00:21 |
| jamielennox | roaet: always possible but for a minimal change like that it should work | 00:21 |
| morganfainberg | bknudson: it also means we need to force updates to the doc when changes are up for review. | 00:22 |
| *** jayg is now known as jayg|g0n3 | 00:22 | |
| *** colinmcnamara has quit IRC | 00:23 | |
| jamielennox | morganfainberg: not really, if we have a doc tag we can autogenerate a public-api list | 00:23 |
| morganfainberg | jamielennox, bknudson: i do like using docstrings so regardless of what people read in that doc, i think the public stuff should explicitly say "this is public", and if that is i 1.0 effort so be it, but i def. support that. | 00:23 |
| morganfainberg | jamielennox: i mean in the interim. | 00:24 |
| bknudson | another option is _prefix for private | 00:24 |
| bknudson | so like _HttpClient | 00:24 |
| *** vaml has quit IRC | 00:24 | |
| morganfainberg | that might be more pythonic, is there an accepted "convention" for classes in python? i know methods and functions are the _prefix | 00:25 |
| roaet | jamielennox: How sad. Jenkins failed it with the same error. I've rebuilt my test env. I guess there is something more hapening | 00:26 |
| jamielennox | roaet: I'm not familiar with running novaclient and i don't have a python2.6 handy to help you there unfortunately | 00:27 |
| bknudson | http://manpages.ubuntu.com/manpages/dapper/man1/epydoc2.4.1.html mentions public / private doc generation... wonder what it uses to know? | 00:27 |
| bknudson | jamielennox: morganfainberg: maybe should use pep8... http://www.python.org/dev/peps/pep-0008/#public-and-internal-interfaces | 00:28 |
| morganfainberg | oh use the __all__ mechanism? | 00:29 |
| *** SumitNaiksatam has joined #openstack-dev | 00:29 | |
| bknudson | and docstrings. | 00:29 |
| jamielennox | bknudson: without reading it pep8 wins all arguments | 00:29 |
| morganfainberg | yeah. | 00:29 |
| morganfainberg | ok, that makes sense. so _prefix, docstrings (and proper use of __all__ if warranted) looks like | 00:30 |
| roaet | jamielennox: thanks for the help. I'll take it to os-neutron | 00:30 |
| bknudson | I've never used __all__ so not really sure how it works. | 00:30 |
| morganfainberg | bknudson: it is for the from blah import * | 00:30 |
| morganfainberg | it specifies what you get. | 00:30 |
| *** dims has quit IRC | 00:30 | |
| morganfainberg | or so i understand. | 00:31 |
| morganfainberg | theoretically, from <blah> import * should only ever import publically consumable stuff. | 00:31 |
| bknudson | so for import keystonclient ... | 00:31 |
| jamielennox | right, it generally says when you import a library what people should have access to - so i guess httpclient.py should never have been importable | 00:31 |
| bknudson | it would exclude httpclient, base, shell, utils | 00:32 |
| bknudson | going off https://github.com/openstack/python-keystoneclient/tree/master/keystoneclient | 00:32 |
| jamielennox | actually thinking about that - if this is a problem then my change to move client.py -> httpclient.py is also a compatibility problem | 00:32 |
| *** vipul-away is now known as vipul | 00:32 | |
| morganfainberg | yep. | 00:32 |
| bknudson | jamielennox: someone commented on that after it merged. | 00:33 |
| jamielennox | i only thought about that in the context of the clients that depended on it | 00:33 |
| *** herndon has joined #openstack-dev | 00:33 | |
| *** Ryan_Lane has joined #openstack-dev | 00:33 | |
| morganfainberg | jamielennox: based upon that merge, i am guessing that we can assume httpclient is private… unless we offer client.py providing HTTPClient imported into its namespace | 00:34 |
| jamielennox | however as pointed out by lin-hua-cheng they apparently have code referencing client.py | 00:34 |
| jamielennox | in the event that i can get the discovery patch in before the next keystoneclient version we can patch that back up for them though | 00:35 |
| jamielennox | but yea, based on that it was assumed private | 00:35 |
| *** sarob_ has joined #openstack-dev | 00:35 | |
| *** lbragstad has joined #openstack-dev | 00:36 | |
| *** lexinator has joined #openstack-dev | 00:36 | |
| morganfainberg | jamielennox: if you provide a transparent HTTPClient object in client.py that logs and says it's deprecated/private and shouldn't be used… might be a good idea. | 00:37 |
| jamielennox | well i can pull the object into the other file and not have circular dependencies, i just can't have them in the same file | 00:37 |
| *** danwent has quit IRC | 00:37 | |
| morganfainberg | right. | 00:38 |
| jamielennox | but yes, for longer term we should mark deprecated | 00:39 |
| morganfainberg | but if we expect it to be private, we could provide an object that says as much. | 00:39 |
| morganfainberg | without breaking compat | 00:39 |
| *** sarob has quit IRC | 00:39 | |
| *** sarob_ has quit IRC | 00:40 | |
| *** alunch has joined #openstack-dev | 00:41 | |
| *** bswartz has quit IRC | 00:43 | |
| *** salv-orlando has quit IRC | 00:44 | |
| *** rwsu is now known as rwsu-away | 00:44 | |
| *** welldannit has quit IRC | 00:44 | |
| *** salv-orlando has joined #openstack-dev | 00:44 | |
| *** colinmcnamara has joined #openstack-dev | 00:44 | |
| *** bswartz has joined #openstack-dev | 00:45 | |
| *** dims has joined #openstack-dev | 00:45 | |
| *** SumitNaiksatam has quit IRC | 00:45 | |
| *** zaitcev has quit IRC | 00:48 | |
| *** gyee has quit IRC | 00:50 | |
| *** bknudson has quit IRC | 00:51 | |
| *** adjohn has quit IRC | 00:51 | |
| *** alunch has quit IRC | 00:52 | |
| *** sarob has joined #openstack-dev | 00:56 | |
| *** bdpayne has quit IRC | 00:57 | |
| *** alexb_ has quit IRC | 00:58 | |
| *** alop has quit IRC | 00:58 | |
| *** adjohn has joined #openstack-dev | 01:03 | |
| morganfainberg | jamielennox: https://review.openstack.org/#/c/39899/6/keystoneclient/httpclient.py similar concerns about the init on HTTPClient | 01:05 |
| morganfainberg | jamielennox: i think we're going to need to make an official call on this and enforce it in the not too distant future. | 01:05 |
| jamielennox | morganfainberg: that one's ok because it's adding an optional parameter | 01:05 |
| morganfainberg | right | 01:05 |
| *** dvarga has joined #openstack-dev | 01:06 | |
| morganfainberg | but generally speaking, it's a similar concern | 01:06 |
| jamielennox | but it's being used as a place to aggregate param handling between v2 and v3 and it shouldn't be | 01:06 |
| *** mjfork has quit IRC | 01:06 | |
| morganfainberg | yep | 01:06 |
| jamielennox | yes, i'm not sure if i have a personal limit on the number of parameters a function should have - but i'd suggest that one is pushing it | 01:07 |
| *** sarob has quit IRC | 01:09 | |
| *** sarob has joined #openstack-dev | 01:09 | |
| jamielennox | morganfainberg: so i just posted https://review.openstack.org/42254 which is an obvious function removal - but i don't think it should be considered an api break | 01:09 |
| *** ljjjustin has joined #openstack-dev | 01:09 | |
| jamielennox | because that function should never be considered for public consumption | 01:11 |
| *** adjohn has quit IRC | 01:11 | |
| *** alunch has joined #openstack-dev | 01:11 | |
| *** Ryan_Lane has quit IRC | 01:12 | |
| morganfainberg | jamielennox: you know… someone is using it. i'm sure </cynical> | 01:13 |
| *** sarob has quit IRC | 01:14 | |
| stevemar | jamielennox, ping | 01:14 |
| jamielennox | morganfainberg: god i hope not | 01:14 |
| jamielennox | yes stevemar? | 01:14 |
| stevemar | reviewing: https://review.openstack.org/#/c/38414/12/keystoneclient/client.py | 01:14 |
| stevemar | jamielennoxi have a silly question | 01:14 |
| stevemar | line 246, why the [1] at the end? | 01:14 |
| jamielennox | so versoins is a dictionary, so we convert it to a list of tuples when we do iteritems() and max will return eg ('2.0', v2_0.client.Client) | 01:16 |
| *** alunch has quit IRC | 01:16 | |
| jamielennox | so it's just getting the client class and dropping the version integer | 01:16 |
| jamielennox | /float | 01:16 |
| stevemar | jamielennox ah okay, so it's grabbing the class, cool | 01:16 |
| jamielennox | also i have been thinking about that since - the blueprint i was working off specified using floats for version number | 01:17 |
| *** sarob has joined #openstack-dev | 01:17 | |
| jamielennox | however in a float 3.15 < 3.2 | 01:17 |
| jamielennox | i'm not sure if we'll ever hit an api revision >= 10 but others must have | 01:18 |
| morganfainberg | jamielennox: that should be fine. 3.1.5 would be < 3.2 | 01:18 |
| morganfainberg | but 3.15 should be > 3.2 | 01:18 |
| stevemar | ah, 15 | 01:18 |
| morganfainberg | oh blah | 01:18 |
| morganfainberg | uhm. you might need to do some splitting | 01:18 |
| jamielennox | morganfainberg: :) | 01:18 |
| jamielennox | i don't think you can | 01:18 |
| stevemar | cross that bridge when it comes? | 01:18 |
| *** radez is now known as radez_g0n3 | 01:18 | |
| morganfainberg | str().split('.') | 01:19 |
| morganfainberg | :P | 01:19 |
| jamielennox | if you store it as a float then 2.1 == 2.10000 | 01:19 |
| *** alunch has joined #openstack-dev | 01:19 | |
| jamielennox | morganfainberg: the spec specifically says pass version as a float so that you can do integer comparison | 01:19 |
| morganfainberg | morganfainberg: oh. i see. | 01:19 |
| morganfainberg | wow that was the wrong target :P | 01:19 |
| morganfainberg | jamielennox: oh i see | 01:19 |
| jamielennox | lol | 01:19 |
| jamielennox | morganfainberg: do you get notified if you ping yourself? | 01:20 |
| morganfainberg | yes i do | 01:20 |
| morganfainberg | but i'm using adium atm, haven't moved to textual | 01:20 |
| jamielennox | so python versions are typically done as a tuple of major, minor | 01:20 |
| jamielennox | because you can do tuple < > | 01:21 |
| jamielennox | so (3, 0) < (3,5) < (3,25) | 01:21 |
| morganfainberg | the spec might warrant a change | 01:21 |
| jamielennox | < (4, 0) | 01:21 |
| stevemar | sounds like tuple is the way to go. | 01:21 |
| jamielennox | i don't remember how but the rules change slightly if you change the tuple length | 01:21 |
| *** jhesketh has quit IRC | 01:22 | |
| jamielennox | so (3, 0, 1) ~ (3, 0) | 01:22 |
| morganfainberg | >>> (3,0,1) < (3,0) | 01:22 |
| morganfainberg | False | 01:22 |
| morganfainberg | >>> (3,0,1) > (3,0) | 01:22 |
| morganfainberg | True | 01:22 |
| *** jhesketh has joined #openstack-dev | 01:23 | |
| jamielennox | seems to hold | 01:23 |
| morganfainberg | passing as a float doesn't preclude splitting it into a tuple for your gt/lt comparison | 01:23 |
| jamielennox | Sequence types also support comparisons. In particular, tuples and lists are compared lexicographically by comparing corresponding elements. This means that to compare equal, every element must compare equal and the two sequences must be of the same type and have the same length. | 01:24 |
| morganfainberg | which is really the only place that concern exists at that point | 01:24 |
| jamielennox | so (3, 0) != (3, 0, 0) | 01:24 |
| morganfainberg | yep. | 01:24 |
| jamielennox | but that user is just being a pain anyway | 01:24 |
| *** grepory has quit IRC | 01:25 | |
| jamielennox | morganfainberg: true | 01:25 |
| morganfainberg | jamielennox: after this convo, i've had to add that into that review :P sorry | 01:26 |
| morganfainberg | might as well make it behave as expected rather than need to patch it in the future. | 01:27 |
| stevemar | morganfainberg: i'm sure it's appreciated | 01:27 |
| jamielennox | morganfainberg: damnit, i should never give away known flaws in reviews | 01:27 |
| morganfainberg | jamielennox: hehe, stevemar already −1'd it :P | 01:27 |
| jamielennox | stevemar: :O how could you | 01:28 |
| stevemar | :D that was before the convo anyway | 01:28 |
| morganfainberg | jamielennox: but that is good food for thought, i'll be looking at version comparisons differently here on forward. | 01:28 |
| stevemar | jaimelennox, i've been too nice in my reviews, gotta be a bit meaner :O | 01:28 |
| stevemar | morganfainberg: good point | 01:29 |
| jamielennox | so i'll patch it up so that it works with floats but i should put that up to the main blueprint | 01:29 |
| jamielennox | so just accept both | 01:29 |
| *** gordc has joined #openstack-dev | 01:29 | |
| morganfainberg | jamielennox: that makes the most sense… i don't see it happening, but what if we wanted M.x.p revision? | 01:29 |
| morganfainberg | with a float… we can't do that | 01:30 |
| morganfainberg | with the tuple acceptance, we can do that in the future... | 01:30 |
| morganfainberg | again, this might be a ML worthy topic (across all projects) | 01:30 |
| morganfainberg | proper handling of version comparisons/version numbers | 01:30 |
| stevemar | morganfainberg: at ibm, some products had V.R.M.F for versioning, ran into a few version problems there :P | 01:33 |
| morganfainberg | god | 01:33 |
| jamielennox | morganfainberg: so i'll have to at least put it onto the previous discovery conversation in the ML | 01:33 |
| morganfainberg | jamielennox: yeah. | 01:33 |
| jamielennox | stevemar: ouch | 01:33 |
| stevemar | was never fun :P | 01:33 |
| jamielennox | stevemar: i bet it was java | 01:33 |
| morganfainberg | stevemar: AIX? :P | 01:33 |
| stevemar | morganfainberg: nah, WebSphere | 01:34 |
| morganfainberg | stevemar: oh. no no no… stop bringing up bad memories | 01:34 |
| morganfainberg | stevemar: (I had to support websphere in a previous life) | 01:34 |
| stevemar | morganfainberg: whats wrong with it?!? hehehe | 01:35 |
| stevemar | better quit talking about it, while i still have a job | 01:35 |
| jamielennox | i have a friend about to go into an ibm internship i think around websphere and such - i'm being encouraging | 01:35 |
| morganfainberg | stevemar: at least when i dealt with it… it was a loooooooooooong time ago | 01:35 |
| morganfainberg | so, lets just say, any reasons i have to dislike it might be/not be relevant anymore | 01:35 |
| stevemar | jamielennox, it'll be interesting at least | 01:36 |
| jamielennox | stevemar: i think it must teach you a lot | 01:36 |
| stevemar | jamielennox, yup, especially if you're a student, most interns/student don't have a clue about enterprise software, or how their servers are set up | 01:38 |
| stevemar | anywho | 01:38 |
| jamielennox | yea, work... | 01:39 |
| *** lexinator has quit IRC | 01:41 | |
| *** lexinator has joined #openstack-dev | 01:41 | |
| *** briancurtin has joined #openstack-dev | 01:42 | |
| *** erkules_ has joined #openstack-dev | 01:46 | |
| *** erkules has quit IRC | 01:46 | |
| *** Mandell has quit IRC | 01:48 | |
| morganfainberg | jamielennox: so.. if you needed to do a recheck with a bug id, will a reverify with a similar bug id likely be needed? | 01:49 |
| jamielennox | not necessarily | 01:49 |
| morganfainberg | ok | 01:49 |
| morganfainberg | i guess we shall see. | 01:49 |
| jamielennox | they are the same call but to different systems | 01:49 |
| morganfainberg | ahh | 01:49 |
| morganfainberg | right | 01:49 |
| jamielennox | you recheck to redo the jenkins check, you reverify if the merge failed | 01:49 |
| morganfainberg | well i was approving a changeset | 01:50 |
| *** Samos123 has quit IRC | 01:50 | |
| morganfainberg | which had a failure on jenkins check (solved with a recheck) | 01:50 |
| morganfainberg | so i was wondering if it would have the same issue in gate. | 01:50 |
| *** SumitNaiksatam has joined #openstack-dev | 01:51 | |
| jamielennox | so possibly, but it's not related to it having failed previously | 01:51 |
| morganfainberg | trying to get the stuff that has been lingering around either feedback or (if appropriate) pushed through before the h3 crunch on the horizon. | 01:51 |
| *** anteaya has quit IRC | 01:51 | |
| morganfainberg | because it's gonna get icky like all milestones | 01:51 |
| jamielennox | yea, i've been trying to look at reviews at the bottom of the list and there are a few that have been multi +1 that never got merged | 01:52 |
| jamielennox | worse i'd suggest | 01:52 |
| morganfainberg | there are ones with +2s that didn't go anywhere | 01:52 |
| morganfainberg | but at least now with keystoneclient the oldest update was 5 days ago | 01:53 |
| morganfainberg | thats a plus | 01:53 |
| *** dolphm has joined #openstack-dev | 01:54 | |
| jamielennox | morganfainberg: there is a lot happening in keystoneclient at the moment, it's good but it was nice and quiet a few weeks ago - didn't have to keep rebasing | 01:54 |
| morganfainberg | jamielennox: yeah. thats tough when you have a lot of dependant changes | 01:55 |
| *** lexinator has quit IRC | 01:55 | |
| morganfainberg | like your code. | 01:55 |
| *** Samos123 has joined #openstack-dev | 01:55 | |
| jamielennox | i stripped most back and i'll resubmit when they get through | 01:55 |
| jamielennox | i had a dependency chain of 8 there for a bit | 01:55 |
| morganfainberg | i need to rebase my 2 keystone changes now because of the test movement | 01:56 |
| jamielennox | that happened? | 01:56 |
| morganfainberg | yep, that change got in. | 01:56 |
| stevemar | morganfainberg: blame the guy who did it :P | 01:56 |
| morganfainberg | tests for keystone are now keystone/tests | 01:57 |
| dolphm | stevemar: o/ | 01:57 |
| morganfainberg | ^^ | 01:57 |
| uvirtbot | morganfainberg: Error: "^" is not a valid command. | 01:57 |
| jamielennox | morazi: nice, about time | 01:57 |
| dolphm | stevemar: "add steps to api spec that note the required body when sending oauth request to /auth/tokens" which steps | 01:57 |
| jamielennox | damn | 01:57 |
| morganfainberg | LOL @ the bot. | 01:57 |
| stevemar | lol good job bot | 01:57 |
| jamielennox | morganfainberg: nice, about time | 01:57 |
| stevemar | dolphm: ahoy | 01:57 |
| stevemar | dolphm: the bit about how body has to have methods [oauth] | 01:58 |
| dolphm | stevemar: ah | 01:58 |
| dolphm | stevemar: that's it? | 01:58 |
| *** tjones has quit IRC | 01:58 | |
| stevemar | for that, yes | 01:58 |
| dolphm | stevemar: does it otherwise implement the spec to the best of your knowledge? | 01:58 |
| stevemar | dolphm: yeppers | 01:58 |
| dolphm | stevemar: +2 | 01:58 |
| stevemar | dolphm: i kept the consumer_id in the issued tokens... that too :O | 01:59 |
| dolphm | stevemar: i saw | 01:59 |
| stevemar | i can remove it, but when I call delete, i'd have to keep calling get consumer_id for every access token | 01:59 |
| dolphm | stevemar: that's not broken by any means, it's just a bit redundant | 02:00 |
| stevemar | dolphm: yep | 02:00 |
| stevemar | dolphm: cool! | 02:00 |
| *** erkules_ is now known as erkules | 02:00 | |
| dolphm | stevemar: the methods: ["oauth"], "oauth": {} was just poor planning on my part | 02:00 |
| dolphm | or poor memory | 02:00 |
| stevemar | dolphm: meh, that's a 1 liner | 02:01 |
| dolphm | stevemar: want to propose an identity-api fix for that, or shall i? | 02:01 |
| stevemar | dolphm: i'll do it now quickly, was just trying to review some changes tonight | 02:02 |
| dolphm | stevemar: cool | 02:02 |
| *** gongysh has joined #openstack-dev | 02:02 | |
| stevemar | dolphm: tonight was the first night all week that I haven't stayed up late trying to get oauth done, so i'm taking it easy | 02:02 |
| stevemar | dolphm: beer in hand and doing reviews, what could go wrong?!? | 02:02 |
| morganfainberg | dolphm: i'm on a rebase spree, i'll rebase your drop diablo/essex migration support. | 02:02 |
| *** spzala has quit IRC | 02:03 | |
| stevemar | good guy morganfainberg: free rebases for everyone | 02:03 |
| dolphm | morganfainberg: thanks! | 02:03 |
| *** sarob has quit IRC | 02:03 | |
| *** spzala has joined #openstack-dev | 02:03 | |
| dolphm | morganfainberg: i started writing a tool to detect merge conflicts before they happen... i should finish that | 02:03 |
| morganfainberg | dolphm: i want to get that one in before one of the others that is mucking around with the legacy.py stuff. | 02:03 |
| *** sarob has joined #openstack-dev | 02:03 | |
| morganfainberg | dolphm: useful tool, will it support jgit's odd… sense of what can be merged? | 02:04 |
| morganfainberg | :P | 02:04 |
| dolphm | morganfainberg: do we use jgit somewhere? | 02:05 |
| morganfainberg | gerrit | 02:05 |
| dolphm | morganfainberg: i was just doing plain old cherry picks with git for now (i'm not sure what the actual merge process looked like) | 02:05 |
| morganfainberg | most of the merge conflicts i've been dealing with today/yesterday have been because gerrit gets confused (tests moving) | 02:05 |
| dolphm | actually i might be doing rebases | 02:06 |
| morganfainberg | dolphm: ah yeah i've been doing cherry-picks mostly | 02:06 |
| dolphm | e.g. rebase fetch head onto master | 02:06 |
| morganfainberg | right, i should start doing that, i think it's more reliable | 02:06 |
| *** yaguang has joined #openstack-dev | 02:06 | |
| *** sarob has quit IRC | 02:08 | |
| *** ljjjustin has quit IRC | 02:09 | |
| *** ljjjustin has joined #openstack-dev | 02:10 | |
| *** dvarga has quit IRC | 02:10 | |
| morganfainberg | hrm | 02:13 |
| morganfainberg | any idea how to derive jenkins failure in gate from a patchset? | 02:13 |
| clarkb | morganfainberg: yes, in the console log will be the zuul ref that was fetched | 02:15 |
| clarkb | morganfainberg: you should be able to fetch that yourself and run the tests | 02:15 |
| *** sarob has joined #openstack-dev | 02:16 | |
| *** ljjjustin has quit IRC | 02:16 | |
| morganfainberg | clarkb: i don't see the console log atm, i am guessing i need to hunt through jenkin's past runs to see what happened? | 02:16 |
| *** ljjjustin has joined #openstack-dev | 02:17 | |
| clarkb | morganfainberg: a link to the logs is left in the gerrit commetn | 02:17 |
| morganfainberg | https://review.openstack.org/#/c/40121/ | 02:17 |
| *** nati_ueno has quit IRC | 02:18 | |
| clarkb | the merge failure? that is actually done by zuul before jenkins runs any tests | 02:18 |
| morganfainberg | ah ok | 02:18 |
| morganfainberg | i see now | 02:18 |
| morganfainberg | i get it | 02:18 |
| clarkb | so change 39845 was ahead of 40121 and that caused a conflict | 02:18 |
| morganfainberg | hm. | 02:18 |
| morganfainberg | oh crud, i thought i grabbed that one | 02:19 |
| *** novas0x2a|laptop has quit IRC | 02:19 | |
| morganfainberg | clarkb: thanks for the help, i was very confused | 02:19 |
| morganfainberg | i missed pulling 39845 first before this one onto my local tree | 02:19 |
| morganfainberg | when checking | 02:19 |
| morganfainberg | clarkb: and i now see the conflict wheee. | 02:21 |
| morganfainberg | thanks again | 02:21 |
| clarkb | np | 02:21 |
| *** pmathews has joined #openstack-dev | 02:21 | |
| *** shang has joined #openstack-dev | 02:23 | |
| *** colinmcnamara has quit IRC | 02:24 | |
| *** colinmcnamara has joined #openstack-dev | 02:25 | |
| *** ljjjustin has quit IRC | 02:27 | |
| *** dims has quit IRC | 02:28 | |
| *** fabio has joined #openstack-dev | 02:28 | |
| *** ljjjustin has joined #openstack-dev | 02:32 | |
| morganfainberg | stevemar: i'll do the full pass on your latest patchset when i get home | 02:33 |
| stevemar | morganfainberg: cool - should be good (i hope) | 02:33 |
| morganfainberg | stevemar: i'm looking forward to seeing OAuth go in (though probably not as much as you are by now) | 02:34 |
| stevemar | morganfainberg: it'll be huge weight off my shoulders thats for sure | 02:34 |
| *** jasondotstar has quit IRC | 02:38 | |
| *** melwitt has quit IRC | 02:39 | |
| *** yaguang has quit IRC | 02:40 | |
| *** aeperezt has quit IRC | 02:44 | |
| *** dguitarbite has joined #openstack-dev | 02:53 | |
| *** macjack has joined #openstack-dev | 02:53 | |
| *** grepory has joined #openstack-dev | 02:53 | |
| *** xchu has joined #openstack-dev | 02:54 | |
| *** redbeard2 has joined #openstack-dev | 02:54 | |
| morganfainberg | stevemar: i have a nit... | 02:54 |
| morganfainberg | stevemar: i wont −1 for it… but it's a nit... | 02:54 |
| stevemar | morganfainberg :O oh no | 02:54 |
| morganfainberg | Copyrights are wrong. | 02:54 |
| stevemar | noooooooo | 02:54 |
| morganfainberg | Should be: Copyright 2013 Openstack Foundation | 02:54 |
| morganfainberg | LLC assets were moved over recently | 02:55 |
| *** jfriedly has quit IRC | 02:55 | |
| *** aeperezt has joined #openstack-dev | 02:55 | |
| morganfainberg | (or so i was told by a few people in my H102 keystoneclient patch) | 02:55 |
| stevemar | all the new files? | 02:55 |
| dolphm | morganfainberg: ++ | 02:55 |
| morganfainberg | yeah. in the new files. | 02:55 |
| dolphm | "All references to "OpenStack LLC" can be changed to "OpenStack Foundation" because the copyright was transferred when the new entity was created." | 02:55 |
| dolphm | https://wiki.openstack.org/wiki/Documentation/Copyright | 02:56 |
| morganfainberg | so, uhm. you should fix that. | 02:56 |
| stevemar | here I was using LLC like a sucker | 02:56 |
| stevemar | cool | 02:56 |
| stevemar | i'll do it quickly since both of you are online | 02:56 |
| dolphm | :) | 02:56 |
| morganfainberg | sounds good. | 02:56 |
| morganfainberg | i'll keep reading the current patchset. | 02:56 |
| *** danwent has joined #openstack-dev | 02:56 | |
| *** yaguang has joined #openstack-dev | 02:58 | |
| morganfainberg | stevemar: thank you for using the named string substitutions. seriously, so much easier to read. | 03:00 |
| stevemar | morganfainberg: what was that now? | 03:00 |
| morganfainberg | i was saying thanks for doing "%(name)s %(someotherthing)s" % { "name": "blah" .... | 03:01 |
| morganfainberg | for the string substitutions | 03:01 |
| stevemar | morganfainberg: also, i added the copyrights to the init.py files | 03:01 |
| stevemar | oh, that, sometimes I do, sometimes I don't | 03:01 |
| morganfainberg | stevemar: i think __init__.py doesn't need the copyrights. but it doesn't hurt :) | 03:02 |
| stevemar | i noticed that it changes from person to person, I noticed that ayoung likes to just use %s | 03:02 |
| morganfainberg | a single %s it doesn't matter | 03:02 |
| morganfainberg | it's when you have mulitples or like 4 or 5 | 03:02 |
| morganfainberg | its makes it much easier. | 03:02 |
| morganfainberg | '%s/%s/%s/%s/%s%s%s/%s/%s' | 03:03 |
| dolphm | stevemar: morganfainberg: fwiw, with more than one substitution in a localized string, you *must* you named parameters | 03:03 |
| morganfainberg | stevemar: then i have a couple of comments for you to fix. | 03:03 |
| stevemar | morganfainberg: go ahead | 03:04 |
| stevemar | dolphm: good to know | 03:04 |
| *** prekarat has joined #openstack-dev | 03:04 | |
| morganfainberg | actually just as dolph said that i saw one. | 03:04 |
| stevemar | ha, i knew I had a few somewhere... | 03:05 |
| *** prekarat has quit IRC | 03:05 | |
| *** prekarat has joined #openstack-dev | 03:05 | |
| *** lexinator has joined #openstack-dev | 03:05 | |
| *** jhesketh has quit IRC | 03:10 | |
| *** martine has joined #openstack-dev | 03:11 | |
| *** martine is now known as Guest44383 | 03:11 | |
| *** kushal has joined #openstack-dev | 03:12 | |
| *** bdpayne has joined #openstack-dev | 03:12 | |
| *** jhesketh has joined #openstack-dev | 03:12 | |
| *** lexinator has quit IRC | 03:13 | |
| stevemar | morganfainberg: it's been too long, that many nits? :O | 03:21 |
| morganfainberg | stevemar: uhm... | 03:21 |
| morganfainberg | not just nits. | 03:21 |
| *** herndon has quit IRC | 03:22 | |
| morganfainberg | stevemar: you missed some function signature updates, and a couple of methods | 03:22 |
| stevemar | morganfainberg: as i was writing that, i was thinking 'nits' might not be the right choice of word | 03:22 |
| morganfainberg | stevemar: i probably should have seen it last time around :( I feel bad for missing them when i pointed out the token revocations for the other drivers. | 03:22 |
| stevemar | morganfainberg: meh | 03:22 |
| morganfainberg | stevemar: so… i'm trying to make sure i don't miss anything this time around | 03:22 |
| morganfainberg | :) | 03:22 |
| morganfainberg | it is also a lot of code. i am so happy someone (i think it was dolphm) stumbled onto "f" in gerrit, and asked about it. super super useful. | 03:23 |
| morganfainberg | i never would have known that hotkey and how much awesome it is otherwise. | 03:23 |
| *** kenperkins has joined #openstack-dev | 03:23 | |
| *** jbresnah has quit IRC | 03:24 | |
| *** kenperkins has quit IRC | 03:24 | |
| *** jbresnah has joined #openstack-dev | 03:24 | |
| stevemar | ahh cool, it brings up the other files, not bad | 03:25 |
| morganfainberg | yep | 03:25 |
| *** fabio has quit IRC | 03:26 | |
| *** noslzzp has quit IRC | 03:29 | |
| morganfainberg | stevemar: comments posted | 03:29 |
| stevemar | morganfainberg: cool | 03:29 |
| morganfainberg | i didn't tag all the copyrights, just one of them. | 03:29 |
| morganfainberg | since i know you already got that | 03:29 |
| stevemar | cool cool | 03:30 |
| *** colinmcnamara has quit IRC | 03:31 | |
| stevemar | morganfainberg: i lol'ed at https://review.openstack.org/#/c/29130/60/keystone/common/sql/core.py | 03:31 |
| morganfainberg | i only caught that one because that was a nit i posted in a previous review ;) | 03:31 |
| *** SergeyLukjanov has joined #openstack-dev | 03:32 | |
| *** colinmcnamara has joined #openstack-dev | 03:32 | |
| stevemar | morganfainberg: can I store None in an sql db? | 03:36 |
| *** adalbas has quit IRC | 03:36 | |
| morganfainberg | well, it gets translated to a NULL iirc | 03:37 |
| *** adalbas has joined #openstack-dev | 03:37 | |
| morganfainberg | if those fields _must_ be populated, i.e. without the field being filled it'll cause strange behavior | 03:37 |
| morganfainberg | they shouldn't be allowed to be null | 03:37 |
| morganfainberg | my concern is that all the fields allow null, not sure if you intend that | 03:38 |
| *** kspear has quit IRC | 03:38 | |
| morganfainberg | stevemar: example on how that works | 03:39 |
| morganfainberg | https://github.com/openstack/keystone/blob/master/keystone/common/sql/migrate_repo/versions/001_add_initial_tables.py#L54 | 03:39 |
| morganfainberg | not sure how much is used elsewhere in keystone | 03:39 |
| morganfainberg | but it stood out that it might be worth a sanity check | 03:39 |
| stevemar | so setting nullable=False, means you can save a None to the backend | 03:40 |
| stevemar | that sounds counter-intuitive | 03:40 |
| morganfainberg | no it means you can't | 03:41 |
| *** pmathews has quit IRC | 03:41 | |
| morganfainberg | i wanted to make sure you can handle any of those fields being none | 03:42 |
| morganfainberg | nothing inherently wrong there, just a sanity check :) | 03:42 |
| morganfainberg | it's a lot of code, sometimes that is warranted | 03:42 |
| dolphm | expires_at looks like the only one that should be nullable=True | 03:43 |
| dolphm | and actually authorizing_user_id should be indexed, no? | 03:43 |
| dolphm | ah it is | 03:44 |
| dolphm | stevemar: i think you can just specify indexed=True (or index=True) on the column definition | 03:44 |
| *** xchu has quit IRC | 03:44 | |
| *** ljjjustin has quit IRC | 03:44 | |
| morganfainberg | dolphm: i think that is backend specific | 03:44 |
| morganfainberg | dolphm: the way he's doing it with the index object is how i've seen it everywhere else | 03:45 |
| morganfainberg | dolphm: but, there is a non-trival chance i am wrong | 03:45 |
| dolphm | morganfainberg: i swear there's an easy way to have sqlalchemy define the index column name automatically, etc | 03:46 |
| morganfainberg | dolphm: also, up late huh? | 03:46 |
| morganfainberg | dolphm: i shall go hunting now. i am curious! | 03:46 |
| dolphm | morganfainberg: http://docs.sqlalchemy.org/en/latest/core/schema.html#indexes | 03:47 |
| morganfainberg | dolphm: aha, so i am wrong! i was about 30 seconds behind you | 03:47 |
| morganfainberg | good to know | 03:48 |
| dolphm | morganfainberg: the way stevemar wrote it is definitely how we do it for belatedly added indexes | 03:48 |
| stevemar | i'm a few minutes behind | 03:48 |
| *** comay has quit IRC | 03:48 | |
| dolphm | stevemar: save a couple lines of code with Column(..., index=True), | 03:48 |
| morganfainberg | woohoo. my talk proposal is up for voting! | 03:49 |
| morganfainberg | i know what i am doing tomorrow.. voting for talks! :) | 03:49 |
| dolphm | morganfainberg: not for the design track, right? | 03:49 |
| morganfainberg | dolphm: no | 03:50 |
| morganfainberg | dolphm: a strategy one | 03:50 |
| morganfainberg | i haven't seen design track submissions page yet | 03:50 |
| dolphm | morganfainberg: whats the title? | 03:50 |
| dolphm | morganfainberg: yeah, i didn't think that was open yet | 03:50 |
| dolphm | morganfainberg: at least another month+ to go for design | 03:51 |
| morganfainberg | dolphm: yeah, i have a few i want to see happen | 03:51 |
| *** macjack has quit IRC | 03:52 | |
| morganfainberg | dolphm: i'm going to rebase my cleanup commits and see if i can figure out why dogpiles invalitates seem to fail when i cache the get_token (if i can get caching on some token stuff, id, and assignment, i'll be happy as a first pass) | 03:56 |
| morganfainberg | erm, meant to say rebase my cleanup ones tonight | 03:56 |
| *** xchu has joined #openstack-dev | 04:01 | |
| *** Guest44383 has quit IRC | 04:03 | |
| *** armax has quit IRC | 04:04 | |
| *** troytoman-away is now known as troytoman | 04:05 | |
| *** edmund has joined #openstack-dev | 04:06 | |
| *** martine has joined #openstack-dev | 04:08 | |
| *** martine is now known as Guest9599 | 04:09 | |
| jerryz | could anyone tell me why keystone on stable/grizzly does not have requirements.txt | 04:15 |
| jerryz | i have to install the deps manually | 04:15 |
| morganfainberg | jerryz: in grizzly the requirements are in tools/pip-requires | 04:16 |
| dolphm | jerryz: it's just in a different location ... ^ | 04:16 |
| *** spzala has quit IRC | 04:16 | |
| jerryz | but when i ran tox -epy27 | 04:16 |
| jerryz | i got module not found error | 04:16 |
| dolphm | jerryz: did you install tools/test-requires as well? | 04:16 |
| *** mdomsch has quit IRC | 04:17 | |
| stevemar | morganfainberg: dolphm: what was the decision about index/sql/nullable-ness? was busy adding named variabled substitutions everywhere | 04:17 |
| *** gordc has quit IRC | 04:17 | |
| jerryz | i just took a look at the two files, they dont have netaddr | 04:18 |
| morganfainberg | stevemar: you can do indexes inline, index=true, but don't have to (you could do it like you're doing now) | 04:18 |
| jerryz | before i did .tox/py27/bin/pip install netaddr, i got ImportError: No module named netaddr | 04:19 |
| morganfainberg | and i think dolphm said expires_at is likely the only column that should be nullable (i could go re look at it if you need some direction) | 04:19 |
| stevemar | morganfainberg: dolphm: yes, just add index=True, to replace the 'index' calls, but what about setting the expires field to nullabel=true? | 04:19 |
| stevemar | when we initially create the request token, the 'authorizing_user_id' is set to None also | 04:20 |
| dolphm | stevemar: then that should be nullable=True | 04:20 |
| morganfainberg | dolphm: isn't nullable=True the default? | 04:20 |
| dolphm | stevemar: personally, i can never remember what the default value of nullable is, so i always explicitly set it on every column | 04:20 |
| stevemar | and the verifier too, since it's not authorized yet | 04:20 |
| morganfainberg | stevemar: do what dolphm said, be explicit per column, that way no surprises | 04:21 |
| stevemar | hehe, alright | 04:21 |
| dolphm | morganfainberg: you're probably right | 04:21 |
| dolphm | morganfainberg: if it was nullable=False by default i'd probably be less paranoid about it | 04:21 |
| * stevemar nods | 04:22 | |
| jerryz | dolphm: i looked at my error log, pip installed the pip-requires and test-requires | 04:22 |
| jerryz | but those two files don't have netaddr | 04:23 |
| *** jhesketh_ has quit IRC | 04:23 | |
| morganfainberg | jerryz if there is a requirement for netaddr, and it's not there, probably should be filed as a bug. looking now. | 04:23 |
| *** afazekas_zz is now known as afazekas | 04:23 | |
| *** jhesketh has quit IRC | 04:23 | |
| *** jhesketh has joined #openstack-dev | 04:24 | |
| morganfainberg | jerryz: i'm not seeing any import of netaddr in stable/grizzly. | 04:24 |
| dolphm | jerryz: can you paste the backtrace you're seeing? | 04:25 |
| morganfainberg | jerryz: ^ | 04:25 |
| jerryz | Traceback (most recent call last): | 04:25 |
| jerryz | File "/home/jenkins/workspace/periodic-keystone-python27-stable-grizzly/tests/test_keystoneclient.py", line 360, in test_change_password_invalidates_token | 04:25 |
| jerryz | from keystoneclient import exceptions as client_exceptions | 04:25 |
| jerryz | File "/home/jenkins/workspace/periodic-keystone-python27-stable-grizzly/vendor/python-keystoneclient-master/keystoneclient/exceptions.py", line 7, in <module> | 04:25 |
| jerryz | from keystoneclient.openstack.common import jsonutils | 04:25 |
| jerryz | File "/home/jenkins/workspace/periodic-keystone-python27-stable-grizzly/vendor/python-keystoneclient-master/keystoneclient/openstack/common/jsonutils.py", line 44, in <module> | 04:25 |
| jerryz | import netaddr | 04:25 |
| jerryz | ImportError: No module named netaddr | 04:25 |
| morganfainberg | ooh keystoneclient. | 04:26 |
| dolphm | why is there a stable/grizzly keystoneclient?! | 04:26 |
| dolphm | what does that even MEAN?! | 04:26 |
| morganfainberg | dolphm: i don't see one. | 04:26 |
| *** neelashah has joined #openstack-dev | 04:26 | |
| morganfainberg | jerryz: i am guessing you have some kind of mixed dependency / bad dep resolution | 04:27 |
| dolphm | that has to be testing keystoneclient master against stable/grizzly? | 04:27 |
| *** rcrit has quit IRC | 04:27 | |
| morganfainberg | except master keystoneclient has netaddr in it's requirements.txt | 04:28 |
| *** jhesketh_ has joined #openstack-dev | 04:28 | |
| dolphm | morganfainberg: jerryz: netaddr is an explicit requirement https://github.com/openstack/python-keystoneclient/blob/master/requirements.txt | 04:28 |
| morganfainberg | unless it's a venv thing? | 04:28 |
| dolphm | jerryz: are you install keystone's requirements and trying to run keystoneclient? | 04:28 |
| dolphm | s/install/installing/ | 04:28 |
| *** alunduil has joined #openstack-dev | 04:29 | |
| *** jhesketh has quit IRC | 04:29 | |
| dolphm | nvm... test_keystoneclient.py so: yes | 04:29 |
| jerryz | i will take a look at what was going on | 04:29 |
| *** dguitarbite has quit IRC | 04:29 | |
| dolphm | hrm, test_keystoneclient.py is one of the integration tests in keystone | 04:30 |
| dolphm | i was thinking it was a base test class in keystoneclient | 04:30 |
| dolphm | s/class/module/ | 04:30 |
| stevemar | morganfainberg: 2 comments left to fix... 1) add test case for list_tokens with consumer (that is a good one, nice catch), and 2) update list token and add consumer | 04:33 |
| *** colinmcnamara has quit IRC | 04:33 | |
| morganfainberg | stevemar: nice. | 04:33 |
| *** Guest9599 has quit IRC | 04:34 | |
| *** briancurtin has quit IRC | 04:36 | |
| jerryz | dolphm: should we add that requirement in test-requires? | 04:37 |
| *** ljjjustin has joined #openstack-dev | 04:37 | |
| dolphm | jerryz: that keystone test is dependent on keystoneclient, which has it's own requirements that should be installed... however, i'm not sure how/when those requirements should be installed | 04:39 |
| dolphm | jerryz: it wouldn't make sense to track the requirements in two places though (especially in a stable branch), because they could change as keystoneclient evolves | 04:39 |
| stevemar | morganfainberg: ideas on how to impl list tokens for consumer_id? ... i can't abuse a query, theres no consumer_id in the token model | 04:40 |
| morganfainberg | stevemar: hm. how does the trust one work? | 04:40 |
| morganfainberg | stevemar: let me look at it. | 04:40 |
| stevemar | queries the model | 04:40 |
| *** nayward has joined #openstack-dev | 04:41 | |
| morganfainberg | oh, it's a top-level item in the token model | 04:41 |
| morganfainberg | stevemar: well, you could just do the same thing delete tokens is doing | 04:42 |
| stevemar | yeah, but even then at least in delete... i can query the user_id | 04:42 |
| *** bdpayne has quit IRC | 04:42 | |
| morganfainberg | list tokens requires a user_id | 04:42 |
| morganfainberg | you can still query it. | 04:43 |
| morganfainberg | def list_tokens(self, user_id, tenant_id=None, trust_id=None): | 04:43 |
| stevemar | gotta change up the comment in the driver! | 04:43 |
| morganfainberg | obviously not your branch because i don't have consumer_id. | 04:44 |
| jerryz | dolphm: i wonder how the jenkins slaves of openstack have all the deps, pre-installed? but my jenkins slave also has netaddr in the system, but tox does not regnoize it | 04:44 |
| morganfainberg | jerryz: i'm trying a stable grizzly tox right now on my local vm | 04:44 |
| morganfainberg | seeing if i can figure out what is going on. | 04:44 |
| dolphm | jerryz: the tox environment is probably ignoring your existing site packages? | 04:45 |
| stevemar | wait, user_id is required... ugh, it's late | 04:45 |
| clarkb | dolphm: jerryz: the deps are not preinstalled for the unittests in the jenkins slaves. And the only project htat should allow site packages is nova (for libvirt) | 04:46 |
| *** colinmcnamara has joined #openstack-dev | 04:46 | |
| morganfainberg | stevemar: it happens ;) | 04:47 |
| jerryz | clarkb: then how pip installed those deps required by different versions of keystoneclient | 04:47 |
| clarkb | jerryz: when you pip install keystoneclient they should be installed | 04:48 |
| clarkb | s/pip// | 04:48 |
| *** alunduil has quit IRC | 04:48 | |
| *** loteriety has left #openstack-dev | 04:50 | |
| morganfainberg | yep just duplicated the problem with stable grizzly | 04:50 |
| jerryz | clarkb: python-keystone is installed in my tox but netaddr wasn't | 04:51 |
| stevemar | morganfainberg: there's no endpoint to query to list tokens right? just have to use the token_api if i want to test out that it works? | 04:51 |
| *** bdpayne has joined #openstack-dev | 04:52 | |
| morganfainberg | stevemar: correct. | 04:52 |
| morganfainberg | stevemar: as far as i can tell that is | 04:53 |
| jerryz | clarkb: python-keystone should declare it has this deps , netaddr . is this how pip works? | 04:53 |
| *** sarob has quit IRC | 04:53 | |
| morganfainberg | stevemar: in fact, it looks like it's only ever used in delete tokens :P | 04:53 |
| *** prekarat has quit IRC | 04:53 | |
| clarkb | http://logs.openstack.org/periodic/periodic-keystone-python27-stable-grizzly/precise5/5/console.html.gz its failing on our daily periodic job as well | 04:53 |
| dolphm | morganfainberg: stevemar: correct | 04:53 |
| *** sarob has joined #openstack-dev | 04:54 | |
| stevemar | dolphm: i just figured out why | 04:54 |
| stevemar | dolphm: because doing so would be crazy-pants | 04:54 |
| morganfainberg | stevemar: and it would be a security issue to expose that data.. a list of _all_ active tokens for a user? or trust? or | 04:55 |
| stevemar | that's what i meant by crazy-pants | 04:55 |
| morganfainberg | yeah | 04:55 |
| *** prekarat has joined #openstack-dev | 04:56 | |
| stevemar | tokens have uuid's as their ids right? | 04:56 |
| dolphm | stevemar: or sha1 hashes for pki | 04:57 |
| stevemar | but it looks like the uuids are returned in list? | 04:57 |
| morganfainberg | stevemar: the unique_id is. | 04:57 |
| morganfainberg | which is the uuid or the sha1 | 04:57 |
| stevemar | is there anyway to translate that to pki to match what I got back from /auth/tokens earlier? | 04:57 |
| morganfainberg | (the primary key) | 04:57 |
| *** sdake_ has joined #openstack-dev | 04:57 | |
| *** afazekas_ has joined #openstack-dev | 04:58 | |
| morganfainberg | you could use the ._unique_id() function on the PKI id | 04:58 |
| stevemar | i will try that | 04:58 |
| morganfainberg | that is on.. uhm.. token.core.Token i think | 04:58 |
| *** sarob has quit IRC | 04:58 | |
| *** lexinator has joined #openstack-dev | 04:58 | |
| stevemar | i sheee | 04:58 |
| stevemar | looks like all that does is: cms.cms_hash_token(token_id) | 04:59 |
| morganfainberg | stevemar: correct | 04:59 |
| *** colinmcnamara has quit IRC | 05:00 | |
| jerryz | dolphm: morganfainberg: should i file a bug? | 05:00 |
| morganfainberg | clarkb: in my tox environment i'm seeing /vendor/python-keystoneclient-master, which is the only thing that could be referencing netaddr | 05:01 |
| morganfainberg | though the pip freeze says i have python-keystoneclient==0.2.5 (activated venv) | 05:01 |
| clarkb | morganfainberg: is this one of those silly tests that downloads a specific version of a thing and does functional testing? | 05:02 |
| morganfainberg | clarkb: i, i don't think so, i'm looking at stable/grizzly to see. | 05:03 |
| dolphm | clarkb: yes! | 05:03 |
| morganfainberg | give me a moment. | 05:03 |
| morganfainberg | trust dolphm, he's i'm sure right | 05:03 |
| stevemar | morganfainberg: just a quick flake8 test, and a change to run my tests again, and i'll submit new patch | 05:03 |
| morganfainberg | stevemar: cool. | 05:03 |
| morganfainberg | dolphm , clarkb: wait, we do grizzly stable against master for obvious reasons. it's just not doing the proper dep thing then? | 05:04 |
| morganfainberg | or am i taking crazy pills (lack of understanding) | 05:04 |
| *** alunduil has joined #openstack-dev | 05:04 | |
| stevemar | just a quick rebase .... | 05:05 |
| clarkb | morganfainberg: I think testing that particular combo is fine (mordred is workign on making it generalized) but the implementation is fragiel | 05:05 |
| *** SergeyLukjanov has quit IRC | 05:05 | |
| morganfainberg | clarkb: got it. | 05:05 |
| clarkb | morganfainberg: looks like it simply clones keystoneclient and does not install it | 05:06 |
| morganfainberg | clarkb: so, in this case should jerryz file a bug? and if so, is this infra? | 05:06 |
| clarkb | see keystone/test.py:checkout_vendor | 05:06 |
| morganfainberg | oh so it's us. | 05:06 |
| morganfainberg | ok. | 05:06 |
| clarkb | yes, no. This is a keystone stable/grizzly unittest bug | 05:06 |
| morganfainberg | yep yep got it | 05:06 |
| *** lexinator has quit IRC | 05:07 | |
| stevemar | morganfainberg: dolphm: new patch | 05:07 |
| morganfainberg | jerryz: yes you should file a bug: https://launchpad.net/keystone for stable/grizzly. if you need help, i can help file the bug. | 05:08 |
| *** jbresnah has quit IRC | 05:08 | |
| morganfainberg | stevemar: cool. i'll look at it in a few i need to get out of the office now that i've finished my rebase-a-polooza | 05:08 |
| stevemar | you are one dedicated employee | 05:09 |
| dolphm | stevemar: you set nullable=True (probably the default anyway?) but not nullable=False on the other fields | 05:09 |
| *** lexinator has joined #openstack-dev | 05:09 | |
| *** Mandell has joined #openstack-dev | 05:09 | |
| dolphm | stevemar: and drop the __table_args__ in favor of index=True | 05:10 |
| stevemar | dolphm: does that matter? they won't ever be null | 05:10 |
| morganfainberg | stevemar: we should enforce it with the SQL schema, it prevents someone from writing code and doing somethng dumb. | 05:10 |
| dolphm | stevemar: it matters because you're allowing them to be null, so inevitably you'll get null values in there and have to deal with it :P | 05:10 |
| stevemar | dolphm: you have trust issues with users :P | 05:11 |
| dolphm | stevemar: with developers, mostly | 05:11 |
| * stevemar grumble grumble | 05:11 | |
| morganfainberg | stevemar: and is it paranoia if they really will do it? | 05:11 |
| stevemar | i'll add them in | 05:11 |
| stevemar | so you want it added to ALL the columns? | 05:12 |
| dolphm | stevemar: "it says nullable=True, so I set consumer_id to NULL because it's obviously optional and now oauth done broke *files bug*" | 05:12 |
| stevemar | (except the ones with nullable=True) | 05:12 |
| dolphm | stevemar: yeah, specify nullable=[True|False] on every column... no harm | 05:13 |
| *** lexinator has quit IRC | 05:14 | |
| stevemar | dolphm: yesterday ayoung mentioned how much more extreme keystones coding standard has been recently | 05:14 |
| *** Shaan7 has quit IRC | 05:14 | |
| jerryz | morganfainberg: is there a way to specify branch ? | 05:14 |
| stevemar | i feel that | 05:15 |
| stevemar | dolphm: id too? | 05:15 |
| dolphm | stevemar: lol there's definitely more and more eyes on the code -- that means higher and higher standards for sure | 05:15 |
| morganfainberg | dolph: any issue with me re-approving https://review.openstack.org/#/c/40121/ since i did the rebase? it was because hte signatures changed and docstrings used that to do the patchset | 05:15 |
| morganfainberg | stevemar: it is safest to specify if unsure, i am not sure how all the sql backends handle a PK if you try and set it to null | 05:16 |
| morganfainberg | jerryz: you might need to just say it's stable/grizzly in the description | 05:17 |
| morganfainberg | jerryz: i'm 2x checking now. | 05:17 |
| *** ljjjustin has quit IRC | 05:18 | |
| dolphm | morganfainberg: +2'd. re-approval is cool as long as you're not making changes that should also be reviewed (anything more than a trivial rebase, for example) | 05:18 |
| *** nshaikh has joined #openstack-dev | 05:18 | |
| morganfainberg | dolphm: yeah no changes besides a bit more than the trival rebase script could determine (due to other changes inherited from merged commits i think) | 05:19 |
| *** neelashah has quit IRC | 05:19 | |
| *** xchu has quit IRC | 05:19 | |
| morganfainberg | jerryz: yeah i don't see anyway in the bug (besides in the description) that you can specify branch. | 05:20 |
| stevemar | dolphm: morganfainberg: https://etherpad.openstack.org/oauth | 05:22 |
| morganfainberg | stevemar: it isn't that i'm just a dedicated employee, its that i want to see the keystone review queue not get backed up. that means initially tossing some more time in :) esp. since i have a little spare time between releases at my dayjob (couple days) | 05:22 |
| morganfainberg | stevemar: consumer_id in RequestToken can be none? | 05:24 |
| *** edmund has quit IRC | 05:25 | |
| morganfainberg | oh wait | 05:25 |
| morganfainberg | misread i think | 05:25 |
| stevemar | you had me so confused | 05:25 |
| morganfainberg | sorry. totally misread | 05:25 |
| stevemar | so, I can drop the sql.Index(...) stuff? and just toss in index=True in the column that I want? | 05:25 |
| morganfainberg | stevemar: according to the SQLAlchemy docs, yes. | 05:26 |
| stevemar | funky | 05:26 |
| *** jerryz has quit IRC | 05:26 | |
| morganfainberg | though if you're adding indexes after the fact, the way you had it before is the "right" way since it isn't the 1st migration of the table. | 05:26 |
| stevemar | i'm going to start up keystone and make sure it actually loads... | 05:27 |
| morganfainberg | stevemar: those make sense to me, nullable vs non-nullable. | 05:27 |
| jamielennox | stevemar and others, my review for discovery you think don't make the VersionNotAvailable exception inherit DiscoveryFailure? https://review.openstack.org/#/c/38414/12/keystoneclient/exceptions.py | 05:28 |
| jamielennox | i had kind of looked at it as a version not available was a discovery failure, just one that you might want to handle seperately | 05:28 |
| *** xchu has joined #openstack-dev | 05:28 | |
| *** ljjjustin has joined #openstack-dev | 05:28 | |
| morganfainberg | jamielennox: i think that is a fair assumption. it is a type of discoveryfailure | 05:29 |
| stevemar | jamie, yes, you are right | 05:29 |
| stevemar | my mistake | 05:29 |
| dolphm | morganfainberg: concerning the review queue not getting backed up... https://pypi.python.org/pypi/next-review/ | 05:29 |
| jamielennox | no worries, just fixing things up and people are still around to ask questions of | 05:29 |
| morganfainberg | dolphm: oh hotness | 05:29 |
| morganfainberg | dolphm: thanks! :) | 05:29 |
| *** jprovazn has joined #openstack-dev | 05:30 | |
| *** ljjjustin has quit IRC | 05:31 | |
| *** ljjjustin has joined #openstack-dev | 05:32 | |
| *** bdpayne has quit IRC | 05:34 | |
| jamielennox | dolphm: i really need to give that another try, unfortunately it seems that panemero (or whatever the ssh library was) doesn't work with ssh keys with passwords - even though i have an ssh-agent | 05:34 |
| dolphm | jamielennox: i think that was a bug on my side ... someone contributed a fix | 05:35 |
| *** kspear has joined #openstack-dev | 05:35 | |
| jamielennox | dolphm: within the version that's on pip? | 05:35 |
| dolphm | jamielennox: yes | 05:35 |
| jamielennox | dolphm: cool, will try again | 05:36 |
| *** dolphm has quit IRC | 05:38 | |
| *** boris-42 has joined #openstack-dev | 05:40 | |
| *** adjohn has joined #openstack-dev | 05:42 | |
| *** jhesketh_ has quit IRC | 05:45 | |
| stevemar | dolphm: morganfainberg: new (and last?) patch!!! | 05:46 |
| *** jhesketh has joined #openstack-dev | 05:46 | |
| *** macjack has joined #openstack-dev | 05:46 | |
| *** ljjjustin has quit IRC | 05:46 | |
| *** bashok has joined #openstack-dev | 05:46 | |
| morganfainberg | stevemar: i'll probably look at it tomorrow morning at this point. | 05:47 |
| stevemar | thats cool | 05:47 |
| *** mohits has joined #openstack-dev | 05:47 | |
| *** matiu has quit IRC | 05:50 | |
| *** jbresnah has joined #openstack-dev | 05:50 | |
| *** eharney has quit IRC | 05:52 | |
| *** freedomhui has joined #openstack-dev | 05:53 | |
| *** hailiang has quit IRC | 05:55 | |
| *** alexb_ has joined #openstack-dev | 06:00 | |
| *** lexinator has joined #openstack-dev | 06:00 | |
| *** hashfail_ has joined #openstack-dev | 06:00 | |
| *** sballe has quit IRC | 06:05 | |
| *** xBsd has joined #openstack-dev | 06:06 | |
| *** fifieldt_ has joined #openstack-dev | 06:06 | |
| *** hailiang has joined #openstack-dev | 06:08 | |
| *** lexinator has quit IRC | 06:09 | |
| *** tmmt has quit IRC | 06:13 | |
| *** bashok has quit IRC | 06:14 | |
| *** bashok has joined #openstack-dev | 06:14 | |
| *** kirankv has joined #openstack-dev | 06:15 | |
| *** tmmt has joined #openstack-dev | 06:17 | |
| *** ljjjustin has joined #openstack-dev | 06:20 | |
| *** vartom117 has joined #openstack-dev | 06:23 | |
| *** grepory has quit IRC | 06:24 | |
| *** adjohn has quit IRC | 06:24 | |
| *** prekarat has quit IRC | 06:27 | |
| *** jbresnah has quit IRC | 06:28 | |
| *** jhesketh has quit IRC | 06:29 | |
| *** prekarat has joined #openstack-dev | 06:30 | |
| *** sballe has joined #openstack-dev | 06:31 | |
| *** nshaikh has left #openstack-dev | 06:31 | |
| *** jhesketh has joined #openstack-dev | 06:32 | |
| *** sballe_ has joined #openstack-dev | 06:33 | |
| *** sballe has quit IRC | 06:36 | |
| *** nshaikh has joined #openstack-dev | 06:38 | |
| *** pschaef has joined #openstack-dev | 06:42 | |
| *** odyssey4me has joined #openstack-dev | 06:42 | |
| *** jerryz has joined #openstack-dev | 06:46 | |
| *** yaguang has quit IRC | 06:48 | |
| *** ljjjustin has quit IRC | 06:50 | |
| *** mmagr has joined #openstack-dev | 06:54 | |
| *** ljjjustin has joined #openstack-dev | 06:54 | |
| *** ifarkas has joined #openstack-dev | 06:56 | |
| *** jerryz has quit IRC | 06:57 | |
| *** jerryz has joined #openstack-dev | 06:58 | |
| *** yaguang has joined #openstack-dev | 07:00 | |
| *** reidrac has joined #openstack-dev | 07:04 | |
| *** mrunge has joined #openstack-dev | 07:04 | |
| *** afazekas has quit IRC | 07:06 | |
| *** afazekas_ is now known as afazekas | 07:06 | |
| *** yaguang has quit IRC | 07:07 | |
| *** odyssey4me has quit IRC | 07:07 | |
| *** odyssey4me has joined #openstack-dev | 07:08 | |
| *** jerryz has quit IRC | 07:08 | |
| *** pschaef has quit IRC | 07:09 | |
| *** jamielennox is now known as jamielennox|away | 07:12 | |
| *** odyssey4me2 has joined #openstack-dev | 07:12 | |
| *** odyssey4me has quit IRC | 07:12 | |
| *** odyssey4me has joined #openstack-dev | 07:14 | |
| *** prekarat has quit IRC | 07:15 | |
| *** prekarat has joined #openstack-dev | 07:15 | |
| *** prekarat has quit IRC | 07:16 | |
| *** prekarat has joined #openstack-dev | 07:16 | |
| *** odyssey4me2 has quit IRC | 07:16 | |
| *** hailiang has quit IRC | 07:17 | |
| *** aababilov has joined #openstack-dev | 07:19 | |
| aababilov | apiclient is introduced to keystoneclient: https://review.openstack.org/#/c/28043/ | 07:19 |
| aababilov | please review | 07:19 |
| *** aababilov has left #openstack-dev | 07:19 | |
| *** flaper87|afk is now known as flaper87 | 07:22 | |
| *** ljjjustin has quit IRC | 07:29 | |
| *** hailiang has joined #openstack-dev | 07:31 | |
| *** ljjjustin has joined #openstack-dev | 07:32 | |
| *** AnilV4 has joined #openstack-dev | 07:33 | |
| *** adjohn has joined #openstack-dev | 07:35 | |
| *** xqueralt-afk is now known as xqueralt | 07:36 | |
| *** MaxV has joined #openstack-dev | 07:37 | |
| *** Ryan_Lane has joined #openstack-dev | 07:38 | |
| *** pschaef has joined #openstack-dev | 07:38 | |
| *** jistr has joined #openstack-dev | 07:38 | |
| *** yaguang has joined #openstack-dev | 07:39 | |
| *** adjohn has quit IRC | 07:40 | |
| *** dina_belova has joined #openstack-dev | 07:42 | |
| bashok | salv-orlando: ping | 07:44 |
| *** stevemar has quit IRC | 07:44 | |
| salv-orlando | bashok: hi | 07:44 |
| *** psedlak has quit IRC | 07:44 | |
| bashok | salv-orlando: I saw your response on the network namespaces issue, I wanted to share my input, I faced similar issues in the past , that time I was told that a bug in ip-route2 packaged causing the issue while deleting the namespace, but till now I dont see any updates on that area. netns-cleanup also wont help (when you get "setting namespace failed" error), it defnitely needs some attention | 07:45 |
| *** lexinator has joined #openstack-dev | 07:45 | |
| salv-orlando | bashok: Did not mention that in the mail, but I am aware too of situations when clearing out the namespace becomes a nightmare | 07:45 |
| salv-orlando | you're right it's an iproute2 issue | 07:46 |
| salv-orlando | I have not seen any progress as well | 07:46 |
| salv-orlando | The only thing I've noticed is that this might happen when you try to remove a ns with active ifaces | 07:46 |
| *** dina_belova has quit IRC | 07:46 | |
| *** dina_belova has joined #openstack-dev | 07:47 | |
| bashok | salv-orlando: especially deleting the qrouter-* namespace is the most trickiest part , | 07:47 |
| salv-orlando | I guess because it has lots of interfaces | 07:47 |
| *** lexinator has quit IRC | 07:49 | |
| *** yolanda has joined #openstack-dev | 07:53 | |
| *** yolanda has quit IRC | 07:54 | |
| *** danwent has quit IRC | 07:54 | |
| *** yolanda has joined #openstack-dev | 07:54 | |
| *** SergeyLukjanov has joined #openstack-dev | 07:57 | |
| *** jerryz has joined #openstack-dev | 07:59 | |
| *** Ng_holiday is now known as Ng | 08:00 | |
| *** giulivo has joined #openstack-dev | 08:01 | |
| *** jaimegil has joined #openstack-dev | 08:03 | |
| *** jpich has joined #openstack-dev | 08:03 | |
| *** boris-42 has quit IRC | 08:05 | |
| *** boden has joined #openstack-dev | 08:05 | |
| *** psedlak has joined #openstack-dev | 08:07 | |
| *** lucasagomes has joined #openstack-dev | 08:08 | |
| *** jerryz has quit IRC | 08:11 | |
| *** cmark has joined #openstack-dev | 08:18 | |
| *** BobBall_Away is now known as BobBall | 08:18 | |
| *** ema has joined #openstack-dev | 08:20 | |
| *** Alexei_987 has joined #openstack-dev | 08:21 | |
| *** xBsd has quit IRC | 08:23 | |
| *** iartarisi has joined #openstack-dev | 08:26 | |
| *** llu has joined #openstack-dev | 08:29 | |
| *** jtomasek has quit IRC | 08:32 | |
| *** jtomasek has joined #openstack-dev | 08:33 | |
| *** sushils has joined #openstack-dev | 08:36 | |
| *** fbo_away is now known as fbo | 08:37 | |
| *** boris-42 has joined #openstack-dev | 08:37 | |
| *** ljjjustin has quit IRC | 08:39 | |
| *** troytoman is now known as troytoman-away | 08:40 | |
| *** vsergeyev has joined #openstack-dev | 08:46 | |
| *** henrynash has quit IRC | 08:46 | |
| sergmelikyan | What is happening with Zuul? | 08:47 |
| sergmelikyan | http://status.openstack.org/zuul/ | 08:47 |
| sergmelikyan | to many unstable builds! | 08:47 |
| *** shane-wang_ has quit IRC | 08:49 | |
| *** zbitter is now known as zaneb | 08:50 | |
| *** ruhe has joined #openstack-dev | 08:53 | |
| *** alexxu has joined #openstack-dev | 09:00 | |
| *** odyssey4me2 has joined #openstack-dev | 09:00 | |
| *** odyssey4me3 has joined #openstack-dev | 09:02 | |
| *** odyssey4me has quit IRC | 09:03 | |
| *** xBsd has joined #openstack-dev | 09:04 | |
| *** odyssey4me2 has quit IRC | 09:05 | |
| *** xqueralt is now known as xqueralt-afk | 09:06 | |
| *** martyntaylor has joined #openstack-dev | 09:16 | |
| *** athomas has joined #openstack-dev | 09:18 | |
| *** salv-orlando has quit IRC | 09:18 | |
| *** dina_belova has quit IRC | 09:26 | |
| *** bashok has quit IRC | 09:38 | |
| xBsd | ping | 09:39 |
| xBsd | what happened with zulu? why all jobs are freezes? | 09:40 |
| *** morganfainberg is now known as morganfainberg_a | 09:49 | |
| *** morganfainberg_a is now known as morganfainberg | 09:52 | |
| *** jhesketh has quit IRC | 09:53 | |
| *** morganfainberg is now known as morganfainberg|a | 09:54 | |
| *** morganfainberg|a is now known as morganfainberg | 09:55 | |
| *** morganfainberg is now known as morganfainberg|a | 09:56 | |
| *** rods has joined #openstack-dev | 09:59 | |
| gongysh | hi, is there a complete state transition diagram somewhere of the VM managed by openstack? | 10:04 |
| *** kushal has quit IRC | 10:05 | |
| *** hailiang has quit IRC | 10:07 | |
| *** kushal has joined #openstack-dev | 10:08 | |
| *** ruhe has quit IRC | 10:10 | |
| *** sumanthns has joined #openstack-dev | 10:13 | |
| *** kaushikc has joined #openstack-dev | 10:14 | |
| *** para__ has joined #openstack-dev | 10:15 | |
| *** mmagr has quit IRC | 10:17 | |
| *** salv-orlando has joined #openstack-dev | 10:19 | |
| *** para__ is now known as mmagr | 10:24 | |
| *** dina_belova has joined #openstack-dev | 10:26 | |
| *** xchu has quit IRC | 10:27 | |
| *** dina_belova has quit IRC | 10:31 | |
| *** pcm_ has joined #openstack-dev | 10:32 | |
| *** pcm_ has quit IRC | 10:32 | |
| *** pcm_ has joined #openstack-dev | 10:33 | |
| *** dina_belova has joined #openstack-dev | 10:34 | |
| *** dina_belova has quit IRC | 10:34 | |
| *** pschaef has quit IRC | 10:34 | |
| *** lexinator has joined #openstack-dev | 10:37 | |
| *** gongysh has quit IRC | 10:37 | |
| *** sumanthns has quit IRC | 10:37 | |
| *** mjfork has joined #openstack-dev | 10:39 | |
| *** dina_belova has joined #openstack-dev | 10:40 | |
| *** dina_bel_ has joined #openstack-dev | 10:40 | |
| *** dina_belova has quit IRC | 10:40 | |
| *** lexinator has quit IRC | 10:42 | |
| *** dina_bel_ has quit IRC | 10:42 | |
| *** yaguang has quit IRC | 10:43 | |
| *** ruhe has joined #openstack-dev | 10:45 | |
| *** bashok has joined #openstack-dev | 10:48 | |
| *** jruzicka has joined #openstack-dev | 10:49 | |
| *** AnilV4 has quit IRC | 10:52 | |
| *** derekh has joined #openstack-dev | 10:57 | |
| *** ndipanov is now known as nick_long | 10:58 | |
| *** openstackstatus has joined #openstack-dev | 11:06 | |
| openstackstatus | NOTICE: some sort of gating disruption has been identified--looking into it now | 11:07 |
| *** ChanServ changes topic to "some sort of gating disruption has been identified--looking into it now" | 11:07 | |
| *** odyssey4me3 has quit IRC | 11:08 | |
| *** dims has joined #openstack-dev | 11:08 | |
| mordred | morganfainberg|a: ping whne you're awake | 11:09 |
| *** giulivo has quit IRC | 11:11 | |
| *** martine has joined #openstack-dev | 11:15 | |
| *** martine is now known as Guest13805 | 11:15 | |
| *** odyssey4me3 has joined #openstack-dev | 11:16 | |
| *** lexinator has joined #openstack-dev | 11:17 | |
| *** giulivo has joined #openstack-dev | 11:17 | |
| *** lexinator has quit IRC | 11:21 | |
| openstackstatus | NOTICE: the log server has filled up, disrupting job completion--working on it now, ETA 12:30 UTC | 11:21 |
| *** ChanServ changes topic to "the log server has filled up, disrupting job completion--working on it now, ETA 12:30 UTC" | 11:21 | |
| *** nayward has quit IRC | 11:22 | |
| *** lucasagomes is now known as lucas-hungry | 11:24 | |
| *** kushal has quit IRC | 11:30 | |
| *** drewlander has joined #openstack-dev | 11:31 | |
| *** huats_ is now known as huats | 11:32 | |
| *** yamahata has quit IRC | 11:35 | |
| *** tmclaugh[work] has joined #openstack-dev | 11:36 | |
| *** e1mer has quit IRC | 11:41 | |
| *** FunnyLookinHat has joined #openstack-dev | 11:45 | |
| *** kirankv has quit IRC | 11:50 | |
| *** lexinator has joined #openstack-dev | 11:52 | |
| *** dina_belova has joined #openstack-dev | 11:53 | |
| *** flaper87 is now known as flaperboon | 11:53 | |
| *** dvarga has joined #openstack-dev | 11:56 | |
| *** lexinator has quit IRC | 11:56 | |
| *** dina_belova has quit IRC | 11:57 | |
| *** kushal has joined #openstack-dev | 12:00 | |
| openstackstatus | NOTICE: log server has a larger filesystem now--rechecking/reverifying jobs, ETA 12:30 UTC | 12:00 |
| *** ChanServ changes topic to "log server has a larger filesystem now--rechecking/reverifying jobs, ETA 12:30 UTC" | 12:00 | |
| *** markmc has joined #openstack-dev | 12:01 | |
| *** CaptTofu has quit IRC | 12:02 | |
| *** CaptTofu has joined #openstack-dev | 12:02 | |
| *** spzala has joined #openstack-dev | 12:02 | |
| *** CaptTofu has quit IRC | 12:04 | |
| *** CaptTofu has joined #openstack-dev | 12:04 | |
| *** CaptTofu has quit IRC | 12:05 | |
| *** CaptTofu has joined #openstack-dev | 12:05 | |
| *** jimjiang has joined #openstack-dev | 12:09 | |
| *** jimjiang has joined #openstack-dev | 12:10 | |
| *** jaimegil has quit IRC | 12:13 | |
| *** troytoman-away is now known as troytoman | 12:13 | |
| *** bashok has quit IRC | 12:16 | |
| *** wwallnrr__ has joined #openstack-dev | 12:16 | |
| *** mberwanger has joined #openstack-dev | 12:18 | |
| *** mrunge has quit IRC | 12:18 | |
| *** galstrom_zzz is now known as galstrom | 12:20 | |
| *** odyssey4me3 has quit IRC | 12:20 | |
| *** slagle has quit IRC | 12:21 | |
| *** ruhe has quit IRC | 12:22 | |
| *** ruhe has joined #openstack-dev | 12:24 | |
| *** gordc has joined #openstack-dev | 12:26 | |
| *** AnilV4 has joined #openstack-dev | 12:27 | |
| *** odyssey4me3 has joined #openstack-dev | 12:28 | |
| *** nshaikh has quit IRC | 12:29 | |
| *** rcrit has joined #openstack-dev | 12:29 | |
| *** nshaikh has joined #openstack-dev | 12:29 | |
| *** SergeyLukjanov has quit IRC | 12:30 | |
| *** woodspa has joined #openstack-dev | 12:31 | |
| *** alunch has quit IRC | 12:31 | |
| *** mberwanger has quit IRC | 12:33 | |
| *** mberwanger has joined #openstack-dev | 12:34 | |
| *** slagle has joined #openstack-dev | 12:34 | |
| *** mberwang_ has joined #openstack-dev | 12:35 | |
| *** kaushikc has quit IRC | 12:36 | |
| *** roz_ has joined #openstack-dev | 12:36 | |
| *** roz_ is now known as andrearoz | 12:36 | |
| *** kaushikc has joined #openstack-dev | 12:37 | |
| *** mberwanger has quit IRC | 12:38 | |
| *** radez_g0n3 is now known as radez | 12:39 | |
| *** redbeard2 has quit IRC | 12:39 | |
| *** dolphm has joined #openstack-dev | 12:40 | |
| openstackstatus | NOTICE: still rechecking/reverifying false negative results on changes, but the gate is moving again | 12:41 |
| *** ChanServ changes topic to "OpenStack development || Support is in #openstack" | 12:41 | |
| *** jasondotstar has joined #openstack-dev | 12:42 | |
| fungi | sergmelikyan: thanks for pitching in on rechecks--that's helping a ton! | 12:42 |
| *** lucas-hungry is now known as lucasagome | 12:43 | |
| *** lucasagome is now known as lucasagomes | 12:43 | |
| *** kaushikc has quit IRC | 12:43 | |
| xBsd | fungi: is jenkins01.openstack.org is up? | 12:44 |
| xBsd | fungi: it seems that a lot of stuff freeze on the node | 12:45 |
| *** dims has quit IRC | 12:48 | |
| *** dvarga has quit IRC | 12:50 | |
| fungi | xBsd: yes, we ran out of space for the jobs to upload their logs, so everything started going haywire. i added more space and it's working again now but stuff which failed earlier has to be rechecked or reverified so it re-runs | 12:52 |
| *** sandywalsh has quit IRC | 12:52 | |
| *** galstrom is now known as galstrom_zzz | 12:53 | |
| *** dina_belova has joined #openstack-dev | 12:53 | |
| xBsd | fungi: but I still couldn't open web interface on that node. is it ok? | 12:54 |
| *** bashok has joined #openstack-dev | 12:55 | |
| *** mberwang_ has quit IRC | 12:56 | |
| fungi | xBsd: the log links which got left for jobs which failed around those times point to log files which never got uploaded/directories which were never successfully created | 12:56 |
| fungi | so when we rerun jobs on those, working links will get posted once they complete | 12:57 |
| *** dstanek has joined #openstack-dev | 12:57 | |
| *** dina_belova has quit IRC | 12:57 | |
| *** ruhe has quit IRC | 12:58 | |
| *** mkollaro has joined #openstack-dev | 12:59 | |
| simo | russellb: what was the command to see the diff between 2 revisions in gerrit ? | 12:59 |
| xBsd | fungi: but I tried to open https://jenkins01.openstack.org/ without any additional links ) | 12:59 |
| *** SergeyLukjanov has joined #openstack-dev | 13:00 | |
| russellb | simo: $ git review -m <review#>,<rev1>-<rev2> | 13:00 |
| fungi | xBsd: it works for me. perhaps your browser is configured to try and protect you from servers which have self-signed certificates? | 13:00 |
| *** jjmb has quit IRC | 13:01 | |
| *** lexinator has joined #openstack-dev | 13:01 | |
| *** anteaya has joined #openstack-dev | 13:01 | |
| simo | russellb: uhmm my git review doesn't like -m | 13:01 |
| russellb | simo: the one in fedora is really old | 13:02 |
| russellb | simo: grab it from pip instead | 13:02 |
| fungi | simo: i think you need git-review 1.21 or later for -m to be available | 13:02 |
| russellb | -m is glorious | 13:02 |
| xBsd | fungi: don't think so, but thanks, I'll try to start reverify job :) | 13:03 |
| simo | russellb: do you know who is the fedora packager ? We should get it in fedora properly | 13:04 |
| fungi | xBsd: if you can load https://jenkins.openstack.org/ but not jenkins01 or jenkins02 it's probably because 01 and 02 have self-signed certificates | 13:04 |
| russellb | simo: i think pete zaitcev did it originally | 13:05 |
| fungi | simo: russellb: indeed, i released 1.23 yesterday and GheRivero_CH uploaded it to debian from a pub | 13:05 |
| *** ema_ has joined #openstack-dev | 13:05 | |
| russellb | simo: https://admin.fedoraproject.org/pkgdb/acls/name/git-review | 13:05 |
| *** dims has joined #openstack-dev | 13:05 | |
| simo | interestingly I do not find the component | 13:06 |
| *** herndon has joined #openstack-dev | 13:06 | |
| simo | going to open a bug to get it properly updated | 13:06 |
| *** lexinator has quit IRC | 13:06 | |
| *** sandywalsh has joined #openstack-dev | 13:07 | |
| *** bknudson has joined #openstack-dev | 13:07 | |
| *** dstanek has quit IRC | 13:07 | |
| *** ema has quit IRC | 13:08 | |
| fungi | simo: russellb: if rh/fed have an active maintainer for g-r i'll be sure to give them a heads up on future releases | 13:09 |
| xBsd | fungi: I'be made reverify but the jobs still freeze in zuul | 13:09 |
| simo | fungi: donno how active he is but we do have a maintainer | 13:09 |
| russellb | fungi: he should be active, he's still doing openstack stuff ... i think ... i don't see him online though | 13:09 |
| xBsd | fungi: http://status.openstack.org/zuul/ openstack/cinder42241,1 | 13:10 |
| flaperboon | russellb: fungi he's still active, he's mainly focused on swift | 13:10 |
| *** jayg|g0n3 is now known as jayg | 13:11 | |
| *** russellb is now known as rustlebee | 13:11 | |
| rustlebee | happy casual nick Friday | 13:11 |
| *** comstud is now known as bearhands | 13:11 | |
| *** ema_ is now known as ema | 13:12 | |
| *** mberwanger has joined #openstack-dev | 13:12 | |
| fungi | xBsd: yeah, looks like there could be some hung jobs on the jenkins servers i'll need to cancel and reverify | 13:12 |
| *** vartom117 has quit IRC | 13:13 | |
| fungi | simo: rustlebee: yeah, just didn't know if he was still active as a distro package maintainer. i know i've seen him working on openstack upstream stuff | 13:13 |
| xBsd | fungi: thanks ) | 13:14 |
| *** neelashah has joined #openstack-dev | 13:15 | |
| fungi | right, i eventually get "The proxy server could not handle the request GET /job/gate-tempest-devstack-vm-postgres-full/2296/." | 13:16 |
| *** dina_belova has joined #openstack-dev | 13:16 | |
| *** zul has quit IRC | 13:16 | |
| *** mberwanger has quit IRC | 13:21 | |
| *** redbeard2 has joined #openstack-dev | 13:21 | |
| *** dprince has joined #openstack-dev | 13:22 | |
| *** rfolco has joined #openstack-dev | 13:22 | |
| fungi | actually, i've started timing out getting to its web interface | 13:23 |
| xBsd | fungi: as I said :) | 13:23 |
| xBsd | something wrong ) | 13:23 |
| *** sergmelikyan has quit IRC | 13:23 | |
| *** dansmith is now known as Steely_Dan | 13:24 | |
| fungi | but it was working for me a little while ago, even after i started rechecking changes | 13:25 |
| *** jprovazn has quit IRC | 13:26 | |
| *** dina_belova has quit IRC | 13:26 | |
| *** vsergeyev has left #openstack-dev | 13:29 | |
| *** xchu has joined #openstack-dev | 13:30 | |
| xBsd | fungi | 13:30 |
| xBsd | fungi: btw what is official place for such bugs? | 13:31 |
| fungi | xBsd: here basically, or in #openstack-infra | 13:31 |
| *** jecarey has joined #openstack-dev | 13:31 | |
| fungi | i'm going to restart the jenkins process on jenkins01, but we may also need to restart zuul because i'm not sure if it will consider those jobs still running if it never hears back from them on the gearman queue | 13:32 |
| *** lexinator has joined #openstack-dev | 13:32 | |
| xBsd | let's burn! :) | 13:32 |
| openstackstatus | NOTICE: the earlier log server issues seem to have put one of the jenkins servers in a bad state, blocking the gate--working on that, ETA 14:00 UTC | 13:34 |
| *** ChanServ changes topic to "the earlier log server issues seem to have put one of the jenkins servers in a bad state, blocking the gate--working on that, ETA 14:00 UTC" | 13:34 | |
| *** dolphm_ has joined #openstack-dev | 13:34 | |
| *** Shaan7 has joined #openstack-dev | 13:35 | |
| fungi | xBsd: we do also take bug reports on launchpad as the openstack-ci project there, but when we're busy fighting fires we usually don't notice those since e-mail tends to take a back seat to irc | 13:36 |
| *** dolphm has quit IRC | 13:36 | |
| *** lexinator has quit IRC | 13:36 | |
| *** dolphm has joined #openstack-dev | 13:37 | |
| xBsd | fungi: I asked, cause I'd written here about this issue about 4 hours ago and didn't get any response | 13:40 |
| *** litong has joined #openstack-dev | 13:40 | |
| xBsd | fungi: but I got #os-infra is better place for that ;) | 13:40 |
| *** FunnyLookinHat has quit IRC | 13:40 | |
| *** alunch has joined #openstack-dev | 13:41 | |
| *** kbringard has joined #openstack-dev | 13:41 | |
| fungi | xBsd: the bigger issue is that right now all the coremudgeons for -infra are on the same piece of rock, so we tend to be asleep at similar times | 13:41 |
| *** jtomasek has quit IRC | 13:43 | |
| *** jtomasek has joined #openstack-dev | 13:43 | |
| *** jtomasek has quit IRC | 13:45 | |
| fungi | xBsd: i'm usually the first one awake an on line, so i didn't spot everyone's concerns in scrollback until about 1100 utc | 13:45 |
| *** SergeyLukjanov has quit IRC | 13:45 | |
| *** jtomasek has joined #openstack-dev | 13:45 | |
| fungi | and it was unfortunately a cascade failure, so i addressed the most immediate problem of the full log server filesystem but once it was back and working it took some time to be sure that there were further problems which seemed to have been triggered by that one | 13:46 |
| fungi | anyway, everything looks happy again now that the jenkins process has been forcibly killed and restarted on jenkins01 (jobs were continuing to run on jenkins.o.o and jenkins02, but the hung jobs on 01 were blocking results for older changes being returned and keeping the gate from advancing) | 13:48 |
| *** burt has joined #openstack-dev | 13:48 | |
| *** bswartz has quit IRC | 13:49 | |
| *** noslzzp has joined #openstack-dev | 13:50 | |
| *** sudorandom has joined #openstack-dev | 13:51 | |
| *** SergeyLukjanov has joined #openstack-dev | 13:51 | |
| *** otherwiseguy has joined #openstack-dev | 13:52 | |
| openstackstatus | NOTICE: the gate seems to be properly moving now, but some changes which were in limbo earlier are probably going to come back with negative votes now. rechecking/reverifying those too | 13:53 |
| *** ChanServ changes topic to "OpenStack development || Support is in #openstack" | 13:53 | |
| *** beagles is now known as seagulls | 13:54 | |
| *** bnemec is now known as oldben | 13:54 | |
| *** ruhe has joined #openstack-dev | 13:54 | |
| *** eharney has joined #openstack-dev | 13:55 | |
| sandywalsh | flaperboon, thanks for the input on the ack() issue ... I may need to pick your brain on some mongo questions next week if you're available? | 13:55 |
| *** adhately has joined #openstack-dev | 13:55 | |
| *** jimfehlig has joined #openstack-dev | 13:56 | |
| *** reidrac has quit IRC | 13:56 | |
| *** reidrac has joined #openstack-dev | 13:56 | |
| *** spzala has quit IRC | 13:58 | |
| *** spzala has joined #openstack-dev | 13:59 | |
| *** jtomasek has quit IRC | 13:59 | |
| *** jtomasek has joined #openstack-dev | 13:59 | |
| *** dolphm_ has quit IRC | 14:00 | |
| *** martyntaylor1 has joined #openstack-dev | 14:00 | |
| *** rwsu-away is now known as rwsu | 14:00 | |
| *** martyntaylor has quit IRC | 14:02 | |
| *** networkstatic has quit IRC | 14:02 | |
| *** datsun180b has joined #openstack-dev | 14:02 | |
| flaperboon | sandywalsh: np :), As Eric, I'm not 100% against that idea, I'd like to see if there's a better way to do it w/o actually exposing it through the API. (btw, When I mention that not all back-ends support ack, I'm also thinking about Marconi) | 14:02 |
| *** networkstatic has joined #openstack-dev | 14:03 | |
| *** kevinconway_ has joined #openstack-dev | 14:03 | |
| markmc | simo, oslo meeting now | 14:03 |
| markmc | sandywalsh, you also :) | 14:03 |
| *** datsun180b has quit IRC | 14:03 | |
| markmc | #openstack-meeting | 14:03 |
| flaperboon | sandywalsh: re mongo: Sure, whenever you want. I'll out from Wed-Friday, so we better chat on Mond and / or Tuesday | 14:03 |
| *** datsun180b has joined #openstack-dev | 14:03 | |
| sandywalsh | flaperboon, cool | 14:04 |
| sandywalsh | markmc, in there | 14:04 |
| markmc | sandywalsh, cool | 14:04 |
| *** kevinconway has quit IRC | 14:04 | |
| *** kevinconway_ is now known as kevinconway | 14:04 | |
| *** HenryG has joined #openstack-dev | 14:05 | |
| *** redbeard2 has quit IRC | 14:06 | |
| *** lexinator has joined #openstack-dev | 14:06 | |
| simo | markmc: joining | 14:07 |
| *** krtaylor has quit IRC | 14:08 | |
| *** briancurtin has joined #openstack-dev | 14:09 | |
| *** changbl has quit IRC | 14:10 | |
| *** slagle has quit IRC | 14:10 | |
| *** lexinator has quit IRC | 14:10 | |
| *** slagle has joined #openstack-dev | 14:12 | |
| *** FunnyLookinHat has joined #openstack-dev | 14:12 | |
| *** bswartz has joined #openstack-dev | 14:12 | |
| *** adhately has left #openstack-dev | 14:12 | |
| *** valgrind has joined #openstack-dev | 14:12 | |
| *** galstrom_zzz is now known as galstrom | 14:16 | |
| *** dprince has quit IRC | 14:18 | |
| *** jmontemayor has joined #openstack-dev | 14:18 | |
| *** tjones has joined #openstack-dev | 14:19 | |
| *** mohits has quit IRC | 14:19 | |
| *** freedomhui has quit IRC | 14:21 | |
| *** redbeard2 has joined #openstack-dev | 14:21 | |
| *** herndon has quit IRC | 14:23 | |
| *** redbeard2 has quit IRC | 14:24 | |
| *** dina_belova has joined #openstack-dev | 14:27 | |
| ayoung | dolphm, I was thinking about josecastro's recent post where they put up 3 copies of Keystone in HTTPD: one for uid/pw, one for Kerberos, and one for X509. THis patch, which makes sense under the present approach, won't work for that: https://review.openstack.org/#/c/41707/2 | 14:28 |
| ayoung | Something to think about for Icehouse | 14:28 |
| *** cthulhup_ has joined #openstack-dev | 14:28 | |
| *** aelkikhia has joined #openstack-dev | 14:28 | |
| *** aelkikhia has left #openstack-dev | 14:29 | |
| *** zul has joined #openstack-dev | 14:31 | |
| ayoung | simo, are you giving up on KDS for Havana? | 14:31 |
| ayoung | Please don't. | 14:31 |
| *** dina_belova has quit IRC | 14:32 | |
| simo | ayoung: already did | 14:32 |
| simo | seem everybody agrees is too risky | 14:32 |
| ayoung | I think it is very close. The API review is down to word choice, and the code is pretty simple. We were discussing if we could tag something as experimental | 14:32 |
| *** jimjiang has quit IRC | 14:32 | |
| simo | 2 PTLs do and Thierry also does | 14:32 |
| simo | ayoung: we can still commit it of course | 14:32 |
| *** dprince has joined #openstack-dev | 14:32 | |
| simo | as long as it is clearly marked experimental I guess | 14:33 |
| simo | and dolphm agrees | 14:33 |
| ayoung | yeah, I think that is the right approach | 14:33 |
| simo | but we will start using it only in Icehouse | 14:33 |
| ayoung | we tag that in the API doc and in the paste file | 14:33 |
| simo | so ... | 14:33 |
| *** esheffield has joined #openstack-dev | 14:33 | |
| simo | ayoung: nothing is using it right now, that's the point | 14:33 |
| simo | and nova won't at this point .. too late | 14:34 |
| ayoung | simo, but no one can build on it if it is not part of the project | 14:34 |
| simo | ayoung: which is why the idea is to push it in on the first few days IceHouse opens up | 14:34 |
| simo | ie keep review up to date and ready to go | 14:34 |
| ayoung | simo, it is the same thing as trusts...it is not just the people that are doing active development here that use our code. It has value to the community at large, and they can build on it in house if they want, but only if it is released | 14:35 |
| simo | just hold the actual push until Icehouse opens | 14:35 |
| simo | ayoung: we can make that case | 14:35 |
| simo | I am ok one way or the other | 14:35 |
| simo | I need to keep the code up and running anyway | 14:35 |
| *** rharwood has joined #openstack-dev | 14:37 | |
| ayoung | ttx, dolphm markmc would you be OK if we use the KDS code as the first example of tagging a feature as experimental? I thin the trules would be: annotated in the API doc, annotated in the paste file, disabled by default. | 14:37 |
| *** lexinator has joined #openstack-dev | 14:37 | |
| *** nayward has joined #openstack-dev | 14:37 | |
| *** dina_belova has joined #openstack-dev | 14:37 | |
| dolphm | ayoung: does "experimental" mean "will not receive security patches" or "security issues will be public by default"? | 14:38 |
| ayoung | dolphm, yes, I think so. | 14:38 |
| dolphm | ayoung: otherwise, you're not buying yourself anything | 14:39 |
| *** boris-42 has quit IRC | 14:39 | |
| ayoung | "This code is here to aid in integration. It will benefit from increased scrutiny by the largest audience..." | 14:39 |
| dolphm | ayoung: i wouldn't call it "experimental" then, i'd call it what it is: "completely unsupported" | 14:41 |
| *** dina_belova has quit IRC | 14:41 | |
| *** lexinator has quit IRC | 14:41 | |
| *** _TheDodd_ has joined #openstack-dev | 14:42 | |
| ayoung | dolphm, well, "Experimental" is how features like this are tagged in the Linux Kernel development process, so there is some precedent. "Experimental and Unsupported" works for me | 14:42 |
| dolphm | (it's actually not a bad idea to consider *all* new features as completely unsupported the *first* time they see a release...) | 14:44 |
| *** stevemar has joined #openstack-dev | 14:44 | |
| dolphm | i bet ttx would love that :P | 14:44 |
| dolphm | "you're on your own attempting to use this feature in a production capacity. if the feature still exists in the next release, it will begin to receive stable backports." | 14:46 |
| *** ruhe has quit IRC | 14:47 | |
| ayoung | dolphm, http://paste.fedoraproject.org/32581/76664403/ | 14:47 |
| ayoung | that would be in keystone-paste.ini | 14:47 |
| *** henrynash has joined #openstack-dev | 14:47 | |
| *** reidrac has left #openstack-dev | 14:47 | |
| *** rnirmal has joined #openstack-dev | 14:48 | |
| ayoung | dolphm, also, we move the api review from https://github.com/openstack/identity-api/tree/master/openstack-identity-api/v3 to https://github.com/openstack/identity-api/tree/master/openstack-identity-api/experimental | 14:48 |
| dolphm | ayoung: wouldn't it be v3/experimental/ ? | 14:49 |
| dolphm | ayoung: or if it's really a top level api, it either needs it's own repo or you're talking about drafting v4 | 14:50 |
| *** henrynash has quit IRC | 14:50 | |
| *** krtaylor has joined #openstack-dev | 14:51 | |
| ayoung | dolphm, I think we keep something in /experimental until we decide where it is going to live for real. But are extensions supposed to be under /v2 /v3 /v4 anyway, or should they be under /extensions? | 14:51 |
| dolphm | ayoung: i'm pretty sure glance shipped with an unsupported api for at least a release ? ... i'd inquire with ttx on the precedent an unsupported release process | 14:51 |
| ayoung | dolphm, I though only the extension discovery was going to be versioned. | 14:52 |
| *** rcleere has joined #openstack-dev | 14:52 | |
| dolphm | ayoung: an extension would need to be documented for any top level api that it extends; the versioning presented by /extensions should imply on which top level api the extension can be expected to exist | 14:53 |
| *** pasquier-s has quit IRC | 14:54 | |
| ayoung | dolphm, So something like kds, which does not extend anything that is currently core is a top level api in and of itself. I just extends Keystone. So if we are targetting, say v3.5 we could put it in /v3/extensions and for now it would go in /v3/experimental. | 14:55 |
| dolphm | ayoung: are you really proposing that kds is it's own application?? | 14:56 |
| *** freedomhui has joined #openstack-dev | 14:56 | |
| dolphm | ayoung: if that's how it's implemented, then it doesn't belong anywhere near our repo | 14:56 |
| *** fifieldt_ has quit IRC | 14:58 | |
| *** danwent has joined #openstack-dev | 14:59 | |
| *** herndon has joined #openstack-dev | 14:59 | |
| ayoung | dolphm, so, there is an implied relationship that we don't explicitly have right now, between Endpoints and the "targets" of Kds. Keystone doesn't currently have a way to distinguish between, say, two differen Nova Compute Nodes. However, from a security standpoint, you would want each to be its own principal, and to have separate security credentials. KDS is a move in that direction. When you create a token, and that token is supposed to be scop | 14:59 |
| ayoung | ed to the service that is going to consume it, you would ideally say "and this can only be done on node123456" but we don';t have that today | 14:59 |
| *** ffio has joined #openstack-dev | 14:59 | |
| ayoung | Noe, from an end user perspective, we don't want to show that to them anyway...too much detail. | 14:59 |
| ayoung | But there is no other service in OpenStack that plays this "Internal Identity Management" role either | 15:00 |
| ayoung | Policy kindof falls into the same realm | 15:00 |
| ayoung | End users should not really be messing with policy. | 15:00 |
| *** sthaha has quit IRC | 15:00 | |
| ayoung | Although there you could make the argument "I need to be able to see the policy in order to give appropriate roles to users." | 15:01 |
| *** woodspa_ has joined #openstack-dev | 15:01 | |
| *** mikemar10 has joined #openstack-dev | 15:01 | |
| dolphm | ayoung: you're avoiding the question | 15:01 |
| ayoung | dolphm, no, I am thinking through it...you have an interesting point.... | 15:01 |
| *** yaguang has joined #openstack-dev | 15:03 | |
| *** athomas has quit IRC | 15:03 | |
| ayoung | dolphm, I could easily say it is its own application. But then, I also think that the identity backends and the assignment backends are separate applications as well, they are just consumed by the token issuing application. | 15:03 |
| ayoung | But KDS does not tie in with Tokens. | 15:03 |
| ayoung | Then again, neither does policy. | 15:03 |
| *** jvrbanac has joined #openstack-dev | 15:03 | |
| simo | ayoung: it consumes them only for administrative duties like any other component | 15:03 |
| ayoung | Catalog...is consumed by tokens. | 15:03 |
| dolphm | ayoung: "kds, which does not extend anything that is currently core is a top level api in and of itself" <-- you're describing a stand alone service | 15:03 |
| *** ruhe has joined #openstack-dev | 15:03 | |
| ayoung | dolphm, true | 15:04 |
| *** woodspa has quit IRC | 15:04 | |
| *** AnilV4 has quit IRC | 15:05 | |
| ayoung | dolphm, we had this discussion back at the start of the efforts. There really is nowhere that KDS could run, but starting up a whole project just for Keys seemed too heavyweight | 15:05 |
| ayoung | I thought you were in on that discussion | 15:05 |
| dolphm | ayoung: exactly which application pipeline should "kds_extension" be deployed to? | 15:05 |
| ayoung | dolphm, I would say admin. | 15:05 |
| *** colinmcnamara has joined #openstack-dev | 15:06 | |
| ayoung | dolphm, the only reason it would be public is if we were making it avaialable as a service to end users. I could see that being valuable, but it is not what we are proposing for now | 15:06 |
| *** athomas has joined #openstack-dev | 15:06 | |
| ayoung | it would make sense as part of the Queuing-as-a-service effort | 15:06 |
| dolphm | ayoung: let me try again... we have exactly 5 application pipelines in keystone.conf.sample. which one should "kds_extension" be included in? https://github.com/openstack/keystone/blob/master/etc/keystone-paste.ini | 15:07 |
| ayoung | [pipeline:admin_api] | 15:07 |
| dolphm | ayoung: hrm, so KDS is a /v2.0/ extension then? | 15:08 |
| simo | currently it sits in api_v3 | 15:08 |
| *** dolphm_ has joined #openstack-dev | 15:09 | |
| *** MaxV has quit IRC | 15:09 | |
| ayoung | dolphm, well, I was not thinking in terms of versions. As I said before, I wasn't thinking that extensionwere falling under vbersions. So, yeah, it would haveto be v3. | 15:09 |
| *** dolphm has quit IRC | 15:09 | |
| ayoung | But then the role that provides access to it should only be given (at least for now) to service_users | 15:09 |
| *** pabelanger is now known as nubbie | 15:10 | |
| dolphm_ | so, if it's in api_v3, then it's deployed indiscriminately on both ports :5000 and :35357 and relies on RBAC | 15:10 |
| *** dina_belova has joined #openstack-dev | 15:10 | |
| *** dina_belova has quit IRC | 15:12 | |
| *** dina_belova has joined #openstack-dev | 15:12 | |
| *** changbl has joined #openstack-dev | 15:12 | |
| dolphm_ | ayoung: however, "service_users" is an arbitrary deployment-specific concept, in which case relying on RBAC is perfect... correct? | 15:13 |
| *** yaguang has quit IRC | 15:13 | |
| *** matiu has joined #openstack-dev | 15:13 | |
| *** matiu has quit IRC | 15:13 | |
| *** matiu has joined #openstack-dev | 15:13 | |
| *** cthulhup_ has quit IRC | 15:15 | |
| *** martyntaylor1 has quit IRC | 15:17 | |
| *** nayward has quit IRC | 15:17 | |
| *** noslzzp has quit IRC | 15:17 | |
| *** mrodden has quit IRC | 15:17 | |
| *** markmcclain has joined #openstack-dev | 15:17 | |
| *** martyntaylor has joined #openstack-dev | 15:18 | |
| *** bashok has quit IRC | 15:19 | |
| *** dolphm has joined #openstack-dev | 15:19 | |
| ayoung | dolphm_, yes, RBAC is made for this. There really is no reason that end users cannot make use of KDS, and I could see it being useful for securing Queue traffic beyond the scoped of the Queues supporting Nova etc | 15:19 |
| *** bashok has joined #openstack-dev | 15:19 | |
| *** bashok_ has joined #openstack-dev | 15:19 | |
| ayoung | It really is a generic mechanism for two entities to share a symmetric key. We had discussed using Kerberos to do this, but simo and I both felt that requiring Kerberos for Securing OpenStack was too high a bar to pass. | 15:20 |
| *** seagulls has quit IRC | 15:20 | |
| *** beagles has joined #openstack-dev | 15:21 | |
| simo | ayoung: technically we can bolt kerberos on/around KDS | 15:21 |
| ayoung | simo, Shh don't give away our secret sauce... :) | 15:21 |
| ayoung | Seriously, though, it does make sense to have it as a web service | 15:21 |
| *** nayward has joined #openstack-dev | 15:22 | |
| *** dolphm_ has quit IRC | 15:22 | |
| ayoung | and as such, it nicely augments any crypto implementation, to include NSS. OpenSSL, or Kerberos | 15:22 |
| *** bashok__ has joined #openstack-dev | 15:22 | |
| *** valgrind has quit IRC | 15:23 | |
| ayoung | So, while Nova devs won't be able to make use of it until Icehouse, it will be avaialble for the Marconi project to make use of as well. | 15:23 |
| *** Mandell has quit IRC | 15:23 | |
| *** bashok has quit IRC | 15:23 | |
| *** bashok_ has quit IRC | 15:24 | |
| *** dprince has quit IRC | 15:24 | |
| *** Guest13805 has quit IRC | 15:25 | |
| *** yaguang has joined #openstack-dev | 15:26 | |
| *** henrynash has joined #openstack-dev | 15:27 | |
| henrynash | ayoung, bknudson: ok, let's give https://review.openstack.org/#/c/39530/ another go after yet another rebase… | 15:28 |
| *** mrodden has joined #openstack-dev | 15:28 | |
| *** colinmcnamara has quit IRC | 15:28 | |
| *** colinmcnamara has joined #openstack-dev | 15:29 | |
| *** tjones has quit IRC | 15:30 | |
| *** bashok_ has joined #openstack-dev | 15:30 | |
| ayoung | henrynash, all logging changes? | 15:30 |
| henrynash | ayoung: yes | 15:31 |
| *** freedomhui has quit IRC | 15:31 | |
| *** yaguang has quit IRC | 15:31 | |
| *** rharwood has quit IRC | 15:32 | |
| *** yolanda has quit IRC | 15:32 | |
| *** krtaylor has quit IRC | 15:32 | |
| *** mrodden has quit IRC | 15:32 | |
| henrynash | ayoung: …and I updated the wrappers in https://review.openstack.org/#/c/38308/ | 15:33 |
| *** spawnofbelliott has quit IRC | 15:33 | |
| *** bashok__ has quit IRC | 15:33 | |
| ayoung | henrynash, I just did the diff on the domain backend patch and it looks good, so reapproved | 15:34 |
| henrynash | ayoung: thx | 15:34 |
| *** belliott has joined #openstack-dev | 15:35 | |
| henrynash | ayoung: what's urgent that needs reviewing from your side...? | 15:35 |
| *** cthulhup_ has joined #openstack-dev | 15:35 | |
| *** pmathews has joined #openstack-dev | 15:36 | |
| ayoung | henrynash, KDS stuff, mostly. I am trying to stay in review mode, and not writ new code right now...but I have some bug fixes on the way. | 15:36 |
| *** grepory has joined #openstack-dev | 15:36 | |
| ayoung | henrynash, did you deal with those split id crashes? | 15:36 |
| *** mrodden has joined #openstack-dev | 15:37 | |
| ayoung | https://bugs.launchpad.net/keystone/+bug/1210590 and https://bugs.launchpad.net/keystone/+bug/1211445 are both assigned to you | 15:37 |
| uvirtbot | Launchpad bug 1210590 in keystone "Split backend crashes with AttributeError" [Critical,Confirmed] | 15:37 |
| ayoung | but I can take the second one | 15:37 |
| dolphm | ayoung: i owe another attempt at reproducing | 15:37 |
| henrynash | ayoung: no..so far I haven't reproduced…Dolphm was seeing if he could get more data... | 15:37 |
| ayoung | OK. Suspect it is related the UNicode and LDAP. | 15:38 |
| henrynash | dolphm: I tried hitting it…and I checked the unit tests - all which seem to cover this case…but there must be something I'm missing.... | 15:38 |
| *** nshaikh has quit IRC | 15:38 | |
| ayoung | dolphm, for the gate, once henrynash 's commit lands, suggest we make the split backend the default for tempest testing. THat way we can run both SQL and LDAP through their paces in one test suite. | 15:39 |
| dolphm | henrynash: i'm trying now | 15:39 |
| henrynash | dolphm: ok…if can get more data, I'll drive it to fix tonight | 15:39 |
| ayoung | Maybe run a few of the tests twice, once against "default domain" which would be LDAP and once against a dynamically created domain which would be in SQL | 15:39 |
| ayoung | henrynash, so, how is the decorator on your new patch applied? | 15:41 |
| ayoung | Is it using a method on the controller? | 15:42 |
| henrynash | ayoung: so @targetprotected will do a get on the target entity and then call the policy engine using it and any api parameters | 15:42 |
| henrynash | ayoung: yes, @targetprotected is in the controller | 15:42 |
| ayoung | henrynash, hmm, that will make it hard to migrate it to oslo common, where it belongs long term | 15:43 |
| ayoung | why'd you use that approach? | 15:43 |
| dolphm | henrynash: ayoung: UTR! :) downgrading priority | 15:43 |
| henrynash | ayoung: err, I thought that's what you were advocating…and that's how all the other protection is being done | 15:44 |
| *** tjones has joined #openstack-dev | 15:44 | |
| *** markmc has quit IRC | 15:44 | |
| *** yaguang has joined #openstack-dev | 15:44 | |
| ayoung | henrynash, nah, there was an article I linked which showed how the whole decorator could be done as a stand alone class. __init__ is called before the function, and __call__ is called on the way out. THen, we can extract the logic from the controller class and make the policy enforcement reusable | 15:45 |
| ayoung | henrynash, let me give it a hack. I want to see if this actually works | 15:45 |
| henrynash | ayoung: …so are you advocating changing ALL the existing protection decorators? | 15:46 |
| henrynash | dolphm: (what's UTR ?_ | 15:46 |
| *** alexb_ has quit IRC | 15:47 | |
| dolphm | henrynash: unable to reproduce | 15:47 |
| henrynash | dolphm: ahh, right! | 15:47 |
| stevemar | shardy: i just replied to your note, with stuff that was completely useless | 15:48 |
| *** ifarkas has quit IRC | 15:48 | |
| stevemar | shardy: i mis-read the question that was asked, sorry :( | 15:48 |
| *** tiamar has quit IRC | 15:48 | |
| *** bdpayne has joined #openstack-dev | 15:48 | |
| *** matiu has quit IRC | 15:48 | |
| shardy | stevemar: no worries, maybe ayoung can help ;) | 15:48 |
| stevemar | shardy: do you have any more info on why it's not working? | 15:49 |
| *** xchu has quit IRC | 15:49 | |
| *** edmund has joined #openstack-dev | 15:50 | |
| shardy | stevemar: well that's the problem, every attempt to authenticate with a trust_id specified in the scope just gives 401 | 15:50 |
| shardy | so without some sample known-good requests, it's hard to see where it's going wrong | 15:50 |
| shardy | stevemar: I can clean up my hacked up test code and post it if needed | 15:50 |
| ayoung | for BRANCH in `git branch | grep review ` ; do git branch -D $BRANCH ; done | 15:50 |
| dolphm | shardy: is your question on list directly related to your bug? if so, you should cross link them | 15:50 |
| shardy | dolphm: well I was going to, but it sounds like the bug may just be a non-obvious error due to the v2 token? | 15:51 |
| shardy | Now I'm just trying to get it working with username/password | 15:51 |
| shardy | all via the v3 api | 15:51 |
| stevemar | shardy: do you never get back a token id at all? | 15:51 |
| ayoung | henrynash, let me put it this way: we are the people most closely looking at policy enforcement, but others will need to consume it. We should be looking to take our extensions and push them to Oslo in the near future. | 15:51 |
| ayoung | THat will imply a refactoring effort to extract them from the rest of the controller code, or to set up a reasonable interface between controller and policy | 15:52 |
| shardy | stevemar: If I specify OS-TRUST:trust in the scope of the request I always get 401 back | 15:52 |
| shardy | stevemar: I may very well be doing something stupid ;) | 15:52 |
| dolphm | shardy: if you're tackling the same problem in both places, providing the extra context will help people trying to help you | 15:52 |
| shardy | but I've burned a lot of time trying to figure out what ;) | 15:52 |
| jaypipes | dhellmann: ok, addressed the one issue you found with the downgrade migration on https://review.openstack.org/#/c/41316/. Should be good to go. thx. | 15:52 |
| dolphm | shardy: no reason to maintain discrete conversations | 15:52 |
| ayoung | shardy, what is happening? | 15:53 |
| dolphm | ayoung: should a v2 token be able to consume a trust on v3? | 15:53 |
| shardy | dolphm: So you'd like me to update the thread with bug info or vice versa? | 15:53 |
| dolphm | shardy: both | 15:53 |
| *** aeperezt_ has joined #openstack-dev | 15:53 | |
| shardy | dolphm: Ok, will do, thanks | 15:53 |
| *** aeperezt_ has quit IRC | 15:53 | |
| ayoung | dolphm, not quite sure what your are asking. By consume, do you mean create a new token? IE I have a v2 token, and I pass that and a trustid to the v3 api? | 15:54 |
| dolphm | ayoung: yes | 15:54 |
| shardy | ayoung: I'm struggling with consuming trusts, getting a trust scoped token, ref bug #1212778 and openstack-dev thread | 15:54 |
| uvirtbot | Launchpad bug 1212778 in keystone "Authenticating with trust ID fails with 500 error" [Undecided,Incomplete] https://launchpad.net/bugs/1212778 | 15:54 |
| *** otherwiseguy has quit IRC | 15:54 | |
| shardy | ayoung: Yeah I'm trying to do the bit marked "Consuming a trust with: POST /auth/tokens" in the API docs | 15:55 |
| *** zul has quit IRC | 15:55 | |
| stevemar | shardy: i was just going to link that: https://github.com/openstack/identity-api/blob/master/openstack-identity-api/v3/src/markdown/identity-api-v3-os-trust-ext.md#consuming-a-trust-with-post-authtokens | 15:55 |
| ayoung | well, 500 is never the right response. But that aside, I think the answer we settled on was 'yes' V2 and V3 tokens can both be passed to V3 API, so trusts should work with a v2 token | 15:55 |
| shardy | so I get a trust-scoped token back, ie authenticate a v3.Client via a trust | 15:55 |
| ayoung | keystone/keystone/auth/plugins/token.py line 42.... | 15:56 |
| *** henrynash has quit IRC | 15:56 | |
| *** freedomhui has joined #openstack-dev | 15:56 | |
| dolphm | ayoung: "should" ... is that tested? | 15:56 |
| ayoung | https://github.com/openstack/keystone/blob/master/keystone/auth/plugins/token.py#L42 | 15:56 |
| dolphm | ayoung: if it's not tested, then it doesn't work.. | 15:57 |
| ayoung | dolphm, It was, but there has been a lot of rework of the code since then. But that doesn't seem to be the problem | 15:57 |
| ayoung | Seems like the token plugin is not handling something cleanly, and that is irrelevant of trusts | 15:57 |
| *** matiu has joined #openstack-dev | 15:57 | |
| *** matiu has joined #openstack-dev | 15:57 | |
| ayoung | KeyError: 'token' | 15:58 |
| *** tjones has quit IRC | 15:58 | |
| ayoung | token_ref['token_data']['token'] seems to me that is the v3 code called against a v2 token, if I had to guess | 15:58 |
| *** jpeeler has quit IRC | 15:58 | |
| *** jpeeler has joined #openstack-dev | 15:59 | |
| ayoung | dolphm, in this case, there is no difference between creating a token from a trust and any other operations performed during authenticate. It looks like passing a v2 token to the v3 api is broken, but I thought we had tests for that. | 15:59 |
| *** rharwood has joined #openstack-dev | 16:00 | |
| *** kevinconway has quit IRC | 16:00 | |
| ayoung | shardy, see the logic executed here, and note that the same logic is missing for the code plugins/token.py https://github.com/openstack/keystone/blob/master/keystone/common/controller.py#L32 | 16:01 |
| *** iartarisi has quit IRC | 16:01 | |
| ayoung | dolphm, should the token auth plugin delegate the validation of the token data to the token provider? I think that both our controller code and our auth/plugins/token.py code are doing differnent thing, and they should hit a unified code path | 16:02 |
| *** nayward has quit IRC | 16:04 | |
| dolphm | ayoung: i would expect the token auth plugin to call the token driver directly... it shouldn't have to go through the provider? | 16:04 |
| ayoung | dolphm, driver is just for storage. THis is a format issue. I thought token format was considered the provice of the provider. | 16:05 |
| ayoung | province | 16:05 |
| dolphm | ayoung: regardless of format, i would expect you to be able to validate a token against the token driver | 16:05 |
| *** ruhe has quit IRC | 16:06 | |
| *** alop has joined #openstack-dev | 16:07 | |
| ayoung | dolphm, driver should say "yes, that is atoken" but when it comes to reading the data out of the token. THe controller code does that for policy enforcement. IN the case of the token driver..it might be a mistake to have that in the plugin. | 16:07 |
| ayoung | user_context.setdefault( | 16:07 |
| ayoung | 'user_id', token_ref['token_data']['token']['user']['id']) | 16:07 |
| dolphm | ayoung: the mistake is that the data persisted as a "token" varies by v2 / v3 / uuid / pki | 16:08 |
| ayoung | dolphm, so it is failing trying to pull the userid out of the token | 16:08 |
| ayoung | that whole block is v3 specific | 16:08 |
| *** woodspa__ has joined #openstack-dev | 16:08 | |
| *** cthulhup_ has quit IRC | 16:08 | |
| *** zbitter has joined #openstack-dev | 16:09 | |
| ayoung | dolphm, So, the provider should have a "normalize_data" call or something. Pass it the token you have, and get the data back in V3 format | 16:09 |
| dolphm | ayoung: because tokens aren't persisted consistently, all the code that depends on the token driver is confusing (to say the least) probably broken (in many more ways than just that one) | 16:09 |
| *** cococafe has joined #openstack-dev | 16:10 | |
| ayoung | dolphm, so we can either change it in the persistance, or we can do the George Lucas thing and fix it in postproduction. | 16:10 |
| ayoung | Until we fix persisitance, I would argue that we need a fix common to the auth plugin and the policy code | 16:10 |
| ayoung | and...that fix really should live in the client, because the same deal for people consuming tokens.... | 16:11 |
| *** sarob has joined #openstack-dev | 16:11 | |
| *** nachi has joined #openstack-dev | 16:11 | |
| *** fbo is now known as fbo_away | 16:11 | |
| ayoung | dolphm, I'll take the above bug for now, and I think we need to have at least on session in Icehouse on "Keystone client: how do we cleanup the mess" | 16:12 |
| *** woodspa_ has quit IRC | 16:12 | |
| *** yaguang has quit IRC | 16:12 | |
| *** zaneb has quit IRC | 16:12 | |
| dolphm | ayoung: token-related code in the service is a much bigger mess than anything on the client side | 16:12 |
| sdake_ | mic is on | 16:13 |
| sdake_ | not sure how to fix - guess should test it out next time :) | 16:13 |
| ayoung | dolphm, agreed. But token format processing should be common to server and client. Keystone the server needs to consume the Keystone client, with common behaviour being fixed there. Or at least keystone/common really does need to be common. | 16:14 |
| *** ftcjeff has joined #openstack-dev | 16:14 | |
| *** armax has joined #openstack-dev | 16:14 | |
| *** zbitter is now known as zaneb | 16:16 | |
| *** yolanda has joined #openstack-dev | 16:16 | |
| shardy | ayoung, dolphm: thanks for working through this | 16:16 |
| shardy | does the Icehouse comment mean we're unlikely to fix this in time for h3? | 16:17 |
| *** giulivo has quit IRC | 16:17 | |
| *** dprince has joined #openstack-dev | 16:17 | |
| dolphm | shardy: your bug is fixable | 16:18 |
| dolphm | shardy: keystone is not :P | 16:18 |
| stevemar | shardy: we try our best to keep it from falling apart everyday | 16:18 |
| *** krtaylor has joined #openstack-dev | 16:18 | |
| shardy | stevemar: haha | 16:19 |
| *** tjones has joined #openstack-dev | 16:19 | |
| *** jistr has quit IRC | 16:19 | |
| shardy | dolphm, ayoung: I'm starting to thing it may be safer to push the heat-trusts work to Icehouse, until this is all a bit better proven | 16:20 |
| ayoung | dolphm, where would I expect to see a test for validating a v2 token in the V3 api? test_v3_auth, right? | 16:20 |
| shardy | ayoung: I was looking at test_v3_v2_intermix as an example request | 16:21 |
| ayoung | shardy, well, you are using a v2 token. Mixing v2 and v3 is a bit wonky, and this is new stuff | 16:21 |
| *** cthulhup_ has joined #openstack-dev | 16:21 | |
| *** kevinconway has joined #openstack-dev | 16:21 | |
| ayoung | shardy, yes, me to, but that is the reverse of what you are doing | 16:21 |
| dolphm | ayoung: yep | 16:21 |
| ayoung | shardy, this is development, this is where we find bugs. You found a bug. | 16:21 |
| shardy | ayoung: I've still not managed to request a trust scoped token using just the v3 API either.. | 16:21 |
| ayoung | We fix bugs | 16:21 |
| shardy | ayoung: sure, I'm more than happy to help with the process, just truying to figure out how lengthy it's likely to be ;) | 16:22 |
| shardy | I've spent too long getting the keystoneclient stuff done and left myself short on time ;) | 16:22 |
| ayoung | shardy, this one will be relatively fast | 16:22 |
| dolphm | s/fix/create/ | 16:22 |
| *** jistr has joined #openstack-dev | 16:23 | |
| shardy | create/fix/refactor/fix-again, the cycle of software ;) | 16:23 |
| *** ema has quit IRC | 16:24 | |
| *** dstufft_ has joined #openstack-dev | 16:26 | |
| *** dstufft has quit IRC | 16:26 | |
| *** dstufft_ is now known as dstufft | 16:26 | |
| dolphm | anyone want to review oauth? :D https://review.openstack.org/#/c/29130/ | 16:27 |
| *** jbresnah has joined #openstack-dev | 16:28 | |
| simo | dolphm: I started a few days ago, but it is big and requires a lot of back and forth to see all that is going on, will take a while | 16:29 |
| *** martyntaylor has quit IRC | 16:30 | |
| *** noslzzp has joined #openstack-dev | 16:30 | |
| *** zul has joined #openstack-dev | 16:30 | |
| *** odyssey4me3 has quit IRC | 16:30 | |
| *** dolphm has quit IRC | 16:31 | |
| stevemar | dolphm: i'm sure morganfainberg will want another look at it | 16:34 |
| *** datsun180b_ has joined #openstack-dev | 16:34 | |
| ayoung | stevemar, I'll take another look, too. | 16:35 |
| ayoung | I have to deal with shardy 's issue first, though, as I think it is a show stopper | 16:35 |
| *** jistr has quit IRC | 16:37 | |
| *** afazekas has quit IRC | 16:37 | |
| *** vkmc has joined #openstack-dev | 16:37 | |
| *** vkmc has quit IRC | 16:37 | |
| *** vkmc has joined #openstack-dev | 16:37 | |
| *** datsun180b has quit IRC | 16:37 | |
| *** datsun180b_ is now known as datsun180b | 16:37 | |
| *** yolanda has quit IRC | 16:38 | |
| *** ruhe has joined #openstack-dev | 16:40 | |
| *** dolphm has joined #openstack-dev | 16:41 | |
| *** henrynash has joined #openstack-dev | 16:42 | |
| *** jkyle has quit IRC | 16:44 | |
| *** jduhamel_ has quit IRC | 16:45 | |
| *** jduhamel has joined #openstack-dev | 16:49 | |
| *** adjohn has joined #openstack-dev | 16:49 | |
| *** hartsocks has joined #openstack-dev | 16:49 | |
| *** ms_brown has joined #openstack-dev | 16:49 | |
| *** fabio__ has joined #openstack-dev | 16:50 | |
| *** alexxu has quit IRC | 16:51 | |
| *** kirankv has joined #openstack-dev | 16:51 | |
| *** hartsocks has quit IRC | 16:53 | |
| *** jecarey has quit IRC | 16:54 | |
| *** jpich has quit IRC | 16:54 | |
| *** alunduil has quit IRC | 16:55 | |
| *** spzala has quit IRC | 16:56 | |
| *** mkerrin has quit IRC | 16:56 | |
| *** sushils has quit IRC | 16:56 | |
| *** lucasagomes is now known as lucas-away | 16:56 | |
| *** sdake_ has quit IRC | 16:58 | |
| *** sdake_ has joined #openstack-dev | 16:58 | |
| *** sdake_ has quit IRC | 16:59 | |
| *** sdake_ has joined #openstack-dev | 16:59 | |
| *** dvarga has joined #openstack-dev | 16:59 | |
| *** nati_ueno has joined #openstack-dev | 16:59 | |
| *** aelkikhia has joined #openstack-dev | 16:59 | |
| *** Ryan_Lane has quit IRC | 17:02 | |
| *** mmagr has quit IRC | 17:02 | |
| *** lucas-away is now known as lucasagomes | 17:02 | |
| *** yolanda has joined #openstack-dev | 17:03 | |
| *** lucasagomes has quit IRC | 17:04 | |
| *** derekh has quit IRC | 17:04 | |
| *** aelkikhia has quit IRC | 17:06 | |
| *** krtaylor has quit IRC | 17:06 | |
| *** sandeepr_ltp has joined #openstack-dev | 17:07 | |
| *** alexb_ has joined #openstack-dev | 17:09 | |
| *** rharwood has quit IRC | 17:09 | |
| *** bashok_ has quit IRC | 17:09 | |
| *** cthulhup_ has quit IRC | 17:10 | |
| *** rharwood has joined #openstack-dev | 17:11 | |
| *** kushal has quit IRC | 17:13 | |
| *** lexinator has joined #openstack-dev | 17:14 | |
| *** spzala has joined #openstack-dev | 17:16 | |
| *** jayg is now known as jayg|g0n3 | 17:17 | |
| *** stevemar is now known as stevemar-afk | 17:18 | |
| *** aelkikhia has joined #openstack-dev | 17:19 | |
| *** psedlak has quit IRC | 17:20 | |
| *** dina_belova has quit IRC | 17:21 | |
| ayoung | dolphm, should the V3 api accept v2 tokens? | 17:21 |
| *** SergeyLukjanov has quit IRC | 17:22 | |
| *** gordc has quit IRC | 17:22 | |
| *** bdpayne has quit IRC | 17:24 | |
| *** kushal has joined #openstack-dev | 17:25 | |
| *** bdpayne has joined #openstack-dev | 17:25 | |
| *** mdomsch has joined #openstack-dev | 17:27 | |
| *** Mandell has joined #openstack-dev | 17:28 | |
| *** adjohn has quit IRC | 17:28 | |
| *** kirankv has quit IRC | 17:31 | |
| *** jfriedly has joined #openstack-dev | 17:31 | |
| *** sarob has quit IRC | 17:31 | |
| *** sarob has joined #openstack-dev | 17:32 | |
| *** noslzzp has quit IRC | 17:32 | |
| *** nayward has joined #openstack-dev | 17:33 | |
| *** kushal has quit IRC | 17:33 | |
| *** cthulhup has joined #openstack-dev | 17:36 | |
| *** sarob has quit IRC | 17:36 | |
| *** sarob has joined #openstack-dev | 17:37 | |
| *** melwitt has joined #openstack-dev | 17:37 | |
| *** holms has joined #openstack-dev | 17:37 | |
| *** yolanda has quit IRC | 17:37 | |
| *** sarob has quit IRC | 17:38 | |
| *** sarob has joined #openstack-dev | 17:39 | |
| *** rods` has quit IRC | 17:39 | |
| *** morganfainberg|a is now known as morganfainberg | 17:41 | |
| *** rods` has joined #openstack-dev | 17:41 | |
| *** yolanda has joined #openstack-dev | 17:41 | |
| *** burt has quit IRC | 17:41 | |
| morganfainberg | dolphm: in the latest patch on oAuth did you mean to remove the nullable=false args? | 17:41 |
| *** afazekas has joined #openstack-dev | 17:41 | |
| morganfainberg | mordred: pong (but only here for a few minutes before heading into the office) | 17:42 |
| *** herndon has quit IRC | 17:42 | |
| mordred | morganfainberg: hey | 17:43 |
| morganfainberg | mordred: what can i do for ya? | 17:43 |
| *** sarob has quit IRC | 17:43 | |
| *** noslzzp has joined #openstack-dev | 17:45 | |
| mordred | morganfainberg: I do not recall why I was pinging you | 17:45 |
| morganfainberg | mordred: hehe, ok then. | 17:45 |
| *** dkehn_ has joined #openstack-dev | 17:45 | |
| *** dkehn has quit IRC | 17:45 | |
| *** kushal has joined #openstack-dev | 17:45 | |
| *** dvarga has quit IRC | 17:47 | |
| holms | y | 17:47 |
| *** otherwiseguy has joined #openstack-dev | 17:49 | |
| *** sarob has joined #openstack-dev | 17:49 | |
| ayoung | shardy, so the issue is that you are passing a v2 token to a v3 api. Is there any reason you can't use a V3 token instead of a v2 token? | 17:49 |
| holms | i wonder if it even possible to get a remote job, for openstack contribution.. | 17:49 |
| *** carl_baldwin has joined #openstack-dev | 17:50 | |
| holms | any lucky people in here who has this job :P? | 17:51 |
| *** kushal has quit IRC | 17:51 | |
| *** kushal has joined #openstack-dev | 17:51 | |
| *** SumitNaiksatam has quit IRC | 17:51 | |
| *** ffio has quit IRC | 17:51 | |
| *** ffio_ has joined #openstack-dev | 17:51 | |
| *** ruhe has quit IRC | 17:52 | |
| *** SumitNaiksatam has joined #openstack-dev | 17:53 | |
| *** cthulhup has quit IRC | 17:56 | |
| dolphm | morganfainberg: no? :( | 17:56 |
| dolphm | ayoung: and yes | 17:56 |
| *** athomas has quit IRC | 17:57 | |
| dolphm | morganfainberg: fixing | 17:57 |
| *** eharney has quit IRC | 17:57 | |
| *** athomas has joined #openstack-dev | 17:57 | |
| *** eharney has joined #openstack-dev | 17:59 | |
| *** eharney has quit IRC | 17:59 | |
| *** eharney has joined #openstack-dev | 17:59 | |
| *** morganfainberg is now known as morganfainberg|a | 17:59 | |
| *** athomas has quit IRC | 17:59 | |
| *** dkehn_ has quit IRC | 18:02 | |
| dolphm | morganfainberg|a: updated, and also fixed a merge conflict with the config changes | 18:02 |
| *** briancurtin has quit IRC | 18:03 | |
| *** dkehn has joined #openstack-dev | 18:03 | |
| *** jasdeepH has joined #openstack-dev | 18:05 | |
| *** nick_long has quit IRC | 18:05 | |
| *** jbresnah has quit IRC | 18:07 | |
| *** dolphm has quit IRC | 18:08 | |
| *** _TheDodd_ has quit IRC | 18:12 | |
| *** anniec has joined #openstack-dev | 18:13 | |
| *** jergerber has quit IRC | 18:13 | |
| *** morganfainberg|a is now known as morganfainberg | 18:15 | |
| *** Mandell has quit IRC | 18:16 | |
| *** jtomasek has quit IRC | 18:18 | |
| *** dmakogon_ has joined #openstack-dev | 18:18 | |
| *** jtomasek has joined #openstack-dev | 18:18 | |
| *** nachi has left #openstack-dev | 18:19 | |
| dmakogon_ | hey, guys, we seeing bug in devstack, this bug failing gerrit build of dashboard | 18:19 |
| *** freedomhui has quit IRC | 18:20 | |
| *** kbrierly has joined #openstack-dev | 18:20 | |
| *** dina_belova has joined #openstack-dev | 18:21 | |
| *** sarob has quit IRC | 18:23 | |
| *** diogogmt has joined #openstack-dev | 18:23 | |
| *** sarob has joined #openstack-dev | 18:23 | |
| *** mkollaro has quit IRC | 18:24 | |
| *** dina_belova has quit IRC | 18:25 | |
| *** sarob has quit IRC | 18:28 | |
| *** jayg|g0n3 is now known as jayg | 18:28 | |
| *** briancurtin has joined #openstack-dev | 18:31 | |
| *** dina_belova has joined #openstack-dev | 18:31 | |
| *** jasdeepH has quit IRC | 18:34 | |
| *** sdake_ has quit IRC | 18:35 | |
| *** dina_belova has quit IRC | 18:36 | |
| *** aelkikhia has quit IRC | 18:39 | |
| *** SergeyLukjanov has joined #openstack-dev | 18:40 | |
| *** nubbie has quit IRC | 18:41 | |
| *** vkmc has quit IRC | 18:42 | |
| *** ffio_ has quit IRC | 18:43 | |
| *** Mandell has joined #openstack-dev | 18:46 | |
| *** Alexei_987 has quit IRC | 18:47 | |
| *** jasdeepH has joined #openstack-dev | 18:47 | |
| *** tjones has quit IRC | 18:50 | |
| *** sarob has joined #openstack-dev | 18:51 | |
| *** nachi has joined #openstack-dev | 18:51 | |
| nachi | henrynash: ping | 18:53 |
| *** openstackstatus has quit IRC | 18:53 | |
| *** bdpayne has quit IRC | 18:54 | |
| nachi | ayoung: ping | 18:54 |
| *** bdpayne has joined #openstack-dev | 18:55 | |
| *** networkstatic has quit IRC | 18:56 | |
| *** donaldh has joined #openstack-dev | 18:58 | |
| nachi | henrynash, ayoung: I am trying to update the keystone/common/config.py for registering a new auth plugin. I found https://github.com/openstack/keystone/blob/master/keystone/common/config.py#L220 is incorrect. The code tries to register default value for password plugin as keystone.auth.plugins.token.Token. Is this correct? | 19:00 |
| *** dina_belova has joined #openstack-dev | 19:01 | |
| *** salv-orlando has quit IRC | 19:01 | |
| *** yolanda has quit IRC | 19:06 | |
| *** cthulhup has joined #openstack-dev | 19:07 | |
| *** aeperezt has quit IRC | 19:08 | |
| *** aeperezt has joined #openstack-dev | 19:10 | |
| *** dstanek has joined #openstack-dev | 19:11 | |
| *** cthulhup has quit IRC | 19:12 | |
| *** noslzzp has quit IRC | 19:12 | |
| *** sandywalsh has quit IRC | 19:14 | |
| morganfainberg | nachi: it looks like token and password are inverted | 19:17 |
| nachi | morganfainberg: ok. | 19:17 |
| morganfainberg | nachi: if you look at the https://github.com/openstack/keystone/blob/master/keystone/common/config.py#L222 it tries to use Password for token | 19:18 |
| morganfainberg | i don't know if this actually would cause problems on the back end, but it definitely looks "off" | 19:18 |
| nachi | morganfainberg: yes. I will change it. thanks for checking for me | 19:19 |
| ayoung | yeah, that looks reversed. Bet I did that. | 19:19 |
| *** donaldh has quit IRC | 19:20 | |
| morganfainberg | nachi: sure thing. | 19:20 |
| ayoung | actually, It might have been henrynash that did that | 19:20 |
| ayoung | but still my fault | 19:21 |
| ayoung | nope, was there before me...lets see | 19:22 |
| nachi | ayoung: i think gyee added like that | 19:22 |
| *** adjohn has joined #openstack-dev | 19:23 | |
| ayoung | nachi, yes he did | 19:24 |
| ayoung | nachi, in commit 9f812939 | 19:24 |
| nachi | ayoung: ok. | 19:24 |
| ayoung | nachi, he owes you a beer, or your beverage of choise | 19:24 |
| ayoung | choice | 19:24 |
| ayoung | nachi, you coming to Hong Kong? | 19:24 |
| nachi | a beer would be nice | 19:24 |
| *** vipul is now known as vipul-away | 19:24 | |
| *** dprince has quit IRC | 19:25 | |
| henrynash | nachi: looks like those two lines are swapped? | 19:25 |
| ayoung | nachi, please file that as a bug. Ping me and I'll set it as confirmed | 19:25 |
| nachi | ayoung: ok | 19:26 |
| *** sandywalsh has joined #openstack-dev | 19:26 | |
| *** vipul-away is now known as vipul | 19:26 | |
| *** colinmcnamara has quit IRC | 19:28 | |
| henrynash | nachi: no, I think it was always like that…I (and young) just converted it to the new format | 19:30 |
| *** adjohn_ has joined #openstack-dev | 19:30 | |
| henrynash | nachi: if you want to raise it and assign to me, happy to fix | 19:31 |
| *** novas0x2a|laptop has joined #openstack-dev | 19:32 | |
| *** henrynash has quit IRC | 19:32 | |
| *** adjohn has quit IRC | 19:32 | |
| nachi | henrynash: OK. https://bugs.launchpad.net/keystone/+bug/1213216 | 19:32 |
| uvirtbot | Launchpad bug 1213216 in keystone "Default values of token and password are inverted in the code keystone/common/config.py " [Undecided,New] | 19:32 |
| *** dolphm has joined #openstack-dev | 19:33 | |
| *** alexb_ has quit IRC | 19:33 | |
| *** lexinator has quit IRC | 19:35 | |
| *** alexb_ has joined #openstack-dev | 19:36 | |
| *** herndon has joined #openstack-dev | 19:41 | |
| *** adjohn_ has quit IRC | 19:43 | |
| ttx | ayoung, dolphm: I don't think there was consensus around having unsupported "experimental features" in stable/supported branches | 19:44 |
| dolphm | ttx: i wanted to follow up with you -- is there any precedent in the release process for that? | 19:44 |
| ttx | at least not until there is a code framework do make all sorts of bells and whistles go off if you do | 19:44 |
| ttx | dolphm: not that I know of | 19:45 |
| ttx | The past "experimental features" were fully suppored | 19:45 |
| ttx | supported* | 19:45 |
| ttx | you were just discouraged (by various means, including setting the default off) to use them | 19:45 |
| ttx | but if a security issue is found in them we'd still have to fix it | 19:46 |
| *** vipul is now known as vipul-away | 19:46 | |
| ttx | dolphm: "feature branches" are what we have for unsupported code | 19:46 |
| *** zul has quit IRC | 19:47 | |
| ttx | so the kds could be landed now in a feature branch if that makes maintaining the future master landing easier | 19:47 |
| dolphm | ttx: i doubt ayoung / simo would be satisfied with that | 19:48 |
| dolphm | ttx: but i find "experimental" to be a completely meaningless weasel word | 19:48 |
| dolphm | ttx: especially when it means backporting private security fixes to a stable branch, anyway | 19:49 |
| ttx | dolphm: I see no point in having this in keystone, unused, but still having to maintain it with security patches. | 19:49 |
| dolphm | ttx: agree | 19:49 |
| ttx | much better to highlight it as a feature that everyone would use by the end of Icehouse | 19:49 |
| *** shardy is now known as shardy_afk | 19:50 | |
| ttx | by landing it early in the cycle, people can talk about integration at the design summit and actually complete it within the cycle | 19:50 |
| *** noslzzp has joined #openstack-dev | 19:50 | |
| *** tmclaugh[work] has quit IRC | 19:51 | |
| *** psedlak has joined #openstack-dev | 19:52 | |
| *** bswartz has quit IRC | 19:54 | |
| *** jruzicka has quit IRC | 19:55 | |
| *** noslzzp has quit IRC | 19:55 | |
| *** dolphm has quit IRC | 19:56 | |
| *** vipul-away is now known as vipul | 19:57 | |
| *** dolphm has joined #openstack-dev | 19:59 | |
| *** drewlander has quit IRC | 20:00 | |
| *** MIDENN_ has joined #openstack-dev | 20:01 | |
| *** anniec has quit IRC | 20:01 | |
| *** henrynash has joined #openstack-dev | 20:02 | |
| *** alop has quit IRC | 20:04 | |
| *** anniec has joined #openstack-dev | 20:04 | |
| *** otherwiseguy has quit IRC | 20:04 | |
| *** lcheng has quit IRC | 20:04 | |
| *** mdenny has quit IRC | 20:04 | |
| *** vaml has joined #openstack-dev | 20:05 | |
| *** lcheng has joined #openstack-dev | 20:05 | |
| *** henrynash has quit IRC | 20:05 | |
| *** kushal has quit IRC | 20:06 | |
| *** jdurgin has quit IRC | 20:06 | |
| stevemar-afk | dolphm: classic dolph, over-writing my changes | 20:07 |
| *** stevemar-afk is now known as stevemar | 20:07 | |
| *** alop has joined #openstack-dev | 20:07 | |
| *** galstrom is now known as galstrom_zzz | 20:08 | |
| *** gyee has joined #openstack-dev | 20:08 | |
| *** psedlak has quit IRC | 20:08 | |
| *** jdurgin has joined #openstack-dev | 20:09 | |
| *** cococafe has quit IRC | 20:09 | |
| *** tjones has joined #openstack-dev | 20:11 | |
| *** bswartz has joined #openstack-dev | 20:13 | |
| *** adjohn has joined #openstack-dev | 20:14 | |
| *** xBsd has quit IRC | 20:14 | |
| morganfainberg | stevemar: i don't think there is anything else i can add to the oauth stuff, doing one more pass, but i wont approve so ayoung can review it if he wants. | 20:16 |
| stevemar | morganfainberg: m'alrighty | 20:16 |
| *** sarob has quit IRC | 20:17 | |
| stevemar | ayoung: pressure is on | 20:17 |
| ayoung | morganfainberg, looking now | 20:17 |
| morganfainberg | ayoung: cool. | 20:17 |
| *** nayward has quit IRC | 20:17 | |
| *** sarob has joined #openstack-dev | 20:17 | |
| morganfainberg | ooh henrynash's change went in. nice. | 20:18 |
| *** krtaylor has joined #openstack-dev | 20:18 | |
| ayoung | 64. stevemar you may now have the record. Let me check | 20:19 |
| stevemar | ayoung: i do, yours was 59 | 20:19 |
| ayoung | Yes it was. Well done. | 20:20 |
| ayoung | stevemar, however, in mine, the tests were in a second review, so you really should count the two together... roughly 120 | 20:21 |
| *** gordc has joined #openstack-dev | 20:21 | |
| stevemar | ayoung: yowza | 20:21 |
| morganfainberg | ayoung: wow. | 20:21 |
| morganfainberg | ayoung: thats a lot of patchsets. | 20:22 |
| ayoung | stevemar, https://review.openstack.org/#/c/23057/ | 20:22 |
| stevemar | ayoung: i imagine finding out the initial token revocation rules was not fun | 20:22 |
| ayoung | Actually, 94 | 20:22 |
| ayoung | 35 for the tests | 20:22 |
| ayoung | so you can still make it | 20:23 |
| stevemar | i was hoping to tie you at 59 :) | 20:23 |
| *** jbresnah has joined #openstack-dev | 20:24 | |
| morganfainberg | stevemar: looks good to me. | 20:25 |
| stevemar | woo hoo | 20:25 |
| stevemar | 2 outta 3 | 20:25 |
| *** vkmc has joined #openstack-dev | 20:25 | |
| stevemar | ayoung: your turn | 20:25 |
| stevemar | ayoung: c'mon, let me have a relaxing weekend :) (just kidding) | 20:26 |
| *** macjack has quit IRC | 20:26 | |
| morganfainberg | stevemar: oh.. if this means you get to relax this weekend… maybe i should find something for you to fix :P | 20:26 |
| stevemar | morganfainberg: no take backs! | 20:26 |
| stevemar | school yard rules | 20:26 |
| stevemar | morganfainberg: i forget... i know we can review WIP patches, but should we? | 20:28 |
| *** boden has quit IRC | 20:28 | |
| *** lexinator has joined #openstack-dev | 20:28 | |
| morganfainberg | stevemar: my view on that is if it is something you've been asked to look at, sure. but likely, i'd ask the person in irc. | 20:29 |
| stevemar | morganfainberg: good call | 20:29 |
| morganfainberg | stevemar: WIP and Draft aren't going to merge for the most part w/o a subsequent patchset. | 20:29 |
| stevemar | yeah | 20:29 |
| *** lexinator1 has joined #openstack-dev | 20:31 | |
| *** lexinator1 has quit IRC | 20:31 | |
| *** lexinator1 has joined #openstack-dev | 20:32 | |
| *** dolphm has quit IRC | 20:32 | |
| *** woodspa__ is now known as woodspa | 20:34 | |
| *** lexinator has quit IRC | 20:35 | |
| *** jonesld has quit IRC | 20:35 | |
| *** sdake_ has joined #openstack-dev | 20:36 | |
| *** sdake_ has quit IRC | 20:36 | |
| *** sdake_ has joined #openstack-dev | 20:36 | |
| *** sdake_ has quit IRC | 20:36 | |
| *** sdake_ has joined #openstack-dev | 20:36 | |
| stevemar | lbragstad: for https://review.openstack.org/#/c/41970/3 you went back to sending notifications in the controller? | 20:37 |
| *** otherwiseguy has joined #openstack-dev | 20:37 | |
| *** xBsd has joined #openstack-dev | 20:39 | |
| ayoung | stevemar, so, I have to admit that I am working on something that I think is slightly higher priority first...but I'll try to get to oauth shortly | 20:41 |
| *** sarob has quit IRC | 20:42 | |
| stevemar | ayoung: that's cool | 20:42 |
| morganfainberg | ayoung: i put a +2 on it, with a note i was holding on approval, should i just make that a +1 so no one gets to it before you? | 20:42 |
| morganfainberg | just was thinking about that and it prob. makes sense to make that a +1 instead. | 20:43 |
| morganfainberg | stevemar: ok, well, i +1'd it ;) | 20:45 |
| *** rods has quit IRC | 20:45 | |
| *** SumitNaiksatam has quit IRC | 20:47 | |
| *** jonesld has joined #openstack-dev | 20:47 | |
| roaet | any thoughts of devstack on ubuntu 13.10? | 20:48 |
| *** msmedved has quit IRC | 20:49 | |
| *** pabelanger has joined #openstack-dev | 20:49 | |
| *** salv-orlando has joined #openstack-dev | 20:50 | |
| *** msmedved has joined #openstack-dev | 20:50 | |
| *** rfolco has quit IRC | 20:53 | |
| *** msmedved has quit IRC | 20:53 | |
| *** jonesld has quit IRC | 20:54 | |
| *** alop has quit IRC | 20:54 | |
| *** ajyounge has quit IRC | 20:55 | |
| *** ajyounge has joined #openstack-dev | 20:56 | |
| *** litong has quit IRC | 20:59 | |
| ayoung | morganfainberg, +2 is fine. If another core wants to put their name on the final approval, I'm OK with that. Just not going to do it without one last look | 21:03 |
| annegentle | on a devstack install, should I get a service catalog back? | 21:03 |
| ayoung | annegentle, only if you ask for it | 21:03 |
| annegentle | mine's empty on a /tokens/ call. Howcome? | 21:04 |
| ayoung | annegentle, do you have anything in the database? | 21:04 |
| *** SumitNaiksatam has joined #openstack-dev | 21:04 | |
| ayoung | Could be a bad install | 21:04 |
| annegentle | ayoung: does devstack do a database or a file catalog? | 21:04 |
| ayoung | annegentle, db | 21:04 |
| annegentle | ayoung: yeah I'm sure it's a drifted install at this point so I can just redo | 21:04 |
| *** jasondotstar has quit IRC | 21:04 | |
| annegentle | ayoung: just making sure | 21:04 |
| *** mrodden has quit IRC | 21:04 | |
| lbragstad | stevemar: yeah, for now.. I responded with my reasoning here: https://review.openstack.org/#/c/41970/1/keystone/identity/controllers.py | 21:05 |
| ayoung | lbragstad, what do you need from context? | 21:05 |
| *** carl_baldwin has quit IRC | 21:05 | |
| ayoung | We dropped context as it was never used. I would prefer to pass in an explicit value that is fetched from context. We were doing contexts of {} all over the place. | 21:06 |
| ayoung | twas dumb | 21:06 |
| stevemar | ayoung: notifications frameworks just needs it apparently | 21:06 |
| lbragstad | ayoung: it is required by the notifier API in Oslo https://github.com/openstack/oslo-incubator/blob/master/openstack/common/notifier/api.py#L95 | 21:06 |
| ayoung | lbragstad, stevemar, that is not an answer | 21:06 |
| ayoung | that is like abuse...passed on from generation to generation | 21:07 |
| *** donaldh has joined #openstack-dev | 21:07 | |
| mordred | ayoung: we're caling them project and not tenants now right? | 21:07 |
| *** sdake has quit IRC | 21:07 | |
| stevemar | mordred: yep | 21:07 |
| ayoung | mordred, yeah, they never really were tenants | 21:07 |
| *** sdake has joined #openstack-dev | 21:07 | |
| *** sdake has quit IRC | 21:07 | |
| *** sdake has joined #openstack-dev | 21:07 | |
| mordred | ayoung: then why did we call them tenants? | 21:08 |
| ayoung | mordred, I think it was optimism | 21:08 |
| mordred | heh | 21:08 |
| mordred | what if we just called them julia? | 21:08 |
| ayoung | they were going for multitenancy, but never quite got the implementation to match the real req | 21:08 |
| stevemar | romeo wouldn't be happy | 21:08 |
| mordred | stevemar: well, he's dead, so nyeh | 21:08 |
| mordred | :) | 21:08 |
| jgriffith | dtroyer: around? | 21:08 |
| ayoung | stevemar, I think he means Julia Child | 21:08 |
| stevemar | ayoung: how could you know that :P | 21:09 |
| ayoung | stevemar, I work on OpenStack. OpenStack is used by the NSA....draw your own conclusions | 21:09 |
| stevemar | ayoung: haha, well done sir, well done. | 21:10 |
| ayoung | stevemar, so I see driver.notify(context, msg) | 21:10 |
| stevemar | ayoung: yeah, i was writing out that reason, but got distracted | 21:10 |
| lbragstad | https://github.com/openstack/oslo-incubator/blob/master/openstack/common/notifier/api.py#L143 | 21:10 |
| ayoung | if not context: | 21:10 |
| ayoung | context = req_context.get_admin_context() | 21:10 |
| *** changbl has quit IRC | 21:11 | |
| ayoung | what does it actually need out of the context? | 21:11 |
| ayoung | https://github.com/openstack/oslo-incubator/blob/master/openstack/common/notifier/log_notifier.py | 21:11 |
| ayoung | nothing there | 21:11 |
| ayoung | rpc.notify(context, topic, message) | 21:12 |
| lbragstad | https://github.com/openstack/oslo-incubator/blob/master/openstack/common/rpc/impl_kombu.py#L852 | 21:12 |
| *** flaperboon is now known as flaper87|afk | 21:12 | |
| lbragstad | which leads into: https://github.com/openstack/oslo-incubator/blob/master/openstack/common/rpc/amqp.py#L602 | 21:13 |
| ayoung | https://github.com/openstack/oslo-incubator/blob/master/openstack/common/rpc/__init__.py#L226 | 21:13 |
| ayoung | https://github.com/openstack/oslo-incubator/blob/master/openstack/common/rpc/amqp.py#L294 | 21:13 |
| lbragstad | yep | 21:14 |
| *** adalbas has quit IRC | 21:14 | |
| dtroyer | jgriffith: yeah, lost in a sea of IRC tabs... | 21:14 |
| ayoung | lbragstad, um, don't send that | 21:14 |
| ayoung | context = {} | 21:14 |
| *** grepory has quit IRC | 21:14 | |
| lbragstad | just hand it an empty context? | 21:14 |
| ayoung | you are going to be sending the whole http request context on each log message? | 21:15 |
| jgriffith | dtroyer: haha | 21:15 |
| ayoung | lbragstad for this kind of notification, I think so | 21:15 |
| jgriffith | dtroyer: what's the secret to getting unstack.sh/stack.sh to work again? | 21:15 |
| ayoung | lbragstad this is in addition to the actual message, no? | 21:15 |
| jgriffith | dtroyer: I've learned that I pretty much need to delete /opt/stack and reclone | 21:15 |
| lbragstad | yeah. everything else specific to the resource being modified should be in the payload | 21:15 |
| dtroyer | jgriffith: cross your toes and put one hand in the air | 21:16 |
| lbragstad | built independently from the context | 21:16 |
| ayoung | lbragstad then, yes, empty context. | 21:16 |
| jgriffith | dtroyer: Ha! | 21:16 |
| dtroyer | jgriffith: or try adding RECLONE=yes to localrc | 21:16 |
| lbragstad | ayoung: cool, I'll respin that | 21:16 |
| stevemar | lbragstad, do as the man says | 21:16 |
| jgriffith | dtroyer: indeed, doing that but now hitting: http://paste.openstack.org/show/44351/ | 21:16 |
| lbragstad | ayoung: stevemar thanks guys | 21:16 |
| ayoung | NP | 21:16 |
| jgriffith | dtroyer: err... wait, not adding reclone to localrc but deleting /opt/stack should essential be equivalent no? | 21:16 |
| jgriffith | dtroyer: even doing 'rm -rf devstack' and reclone still getting htis | 21:17 |
| *** colinmcnamara has joined #openstack-dev | 21:17 | |
| dtroyer | jgriffith: RECLONE only affects the repos, deleting /opt/stack cleans out /opt/stack/data and friends. sounds like something in unstack isn't removing something that needs to be removed. | 21:17 |
| dtroyer | jgriffith: try ./clean.sh though | 21:18 |
| *** vkmc has quit IRC | 21:18 | |
| jgriffith | dtroyer: bummer... was hoping you'd say "oh yeah.. just do this" :) | 21:18 |
| *** mrodden has joined #openstack-dev | 21:18 | |
| dtroyer | jgriffith: it calls unstack.sh then gets all ruthless on things like rabbitmq and other stuff…it might get the data dirs you need | 21:18 |
| jgriffith | dtroyer: k, I'll keep messing with it, if I figure anything out you'll be one of the first to hear ;) | 21:18 |
| jgriffith | dtroyer: trying now | 21:18 |
| dtroyer | jgriffith: I haven't seen that particular error, but it reeks of a cert cache dir not getting cleaned out or overwritten as it should | 21:19 |
| jgriffith | dtroyer: oh.. indeed, that's a bit more in depth | 21:19 |
| dtroyer | jgriffith: I'm not too familiar with the accrc bits | 21:19 |
| *** donaldh has quit IRC | 21:19 | |
| jgriffith | dtroyer: yeah, I tried tracing through *where* those end up living but no luck | 21:20 |
| jgriffith | dtroyer: ok, well, trying w/ clean.sh now... maybe I'll get lucky :) | 21:20 |
| *** radez is now known as radez_g0n3 | 21:20 | |
| *** rods has joined #openstack-dev | 21:22 | |
| *** dina_belova has quit IRC | 21:25 | |
| *** adjohn has quit IRC | 21:25 | |
| *** adjohn has joined #openstack-dev | 21:25 | |
| *** Aj0 has joined #openstack-dev | 21:28 | |
| *** rwsu has quit IRC | 21:29 | |
| *** adalbas has joined #openstack-dev | 21:29 | |
| *** rwsu has joined #openstack-dev | 21:33 | |
| *** anniec has quit IRC | 21:36 | |
| *** anniec has joined #openstack-dev | 21:37 | |
| *** adalbas has quit IRC | 21:38 | |
| *** hemanth has joined #openstack-dev | 21:40 | |
| *** hemanth has quit IRC | 21:40 | |
| *** markwash has joined #openstack-dev | 21:43 | |
| *** spzala has quit IRC | 21:47 | |
| *** mdomsch has quit IRC | 21:52 | |
| *** jrgcombr has quit IRC | 21:53 | |
| *** jrgcombr has joined #openstack-dev | 21:53 | |
| *** eharney has quit IRC | 21:55 | |
| *** esheffield has quit IRC | 21:55 | |
| *** ms_brown has quit IRC | 21:57 | |
| *** egallen has joined #openstack-dev | 22:03 | |
| *** ftcjeff has quit IRC | 22:04 | |
| *** bknudson has quit IRC | 22:06 | |
| *** kbringard has quit IRC | 22:07 | |
| *** aeperezt has quit IRC | 22:08 | |
| *** jv__ has joined #openstack-dev | 22:09 | |
| *** jv__ has quit IRC | 22:10 | |
| *** gordc has quit IRC | 22:11 | |
| *** woodspa has quit IRC | 22:12 | |
| *** jtomasek has quit IRC | 22:13 | |
| *** neelashah has quit IRC | 22:14 | |
| *** kbrierly has quit IRC | 22:15 | |
| *** blamar has joined #openstack-dev | 22:16 | |
| *** rharwood has quit IRC | 22:19 | |
| *** jvrbanac has quit IRC | 22:20 | |
| *** SergeyLukjanov has quit IRC | 22:25 | |
| *** dina_belova has joined #openstack-dev | 22:25 | |
| *** jimfehlig has quit IRC | 22:25 | |
| *** blamar has quit IRC | 22:25 | |
| *** sarob has joined #openstack-dev | 22:28 | |
| *** jmontemayor has quit IRC | 22:28 | |
| *** sarob has quit IRC | 22:28 | |
| *** anniec has quit IRC | 22:28 | |
| *** sarob has joined #openstack-dev | 22:29 | |
| *** dina_belova has quit IRC | 22:30 | |
| *** wu_wenxiang has joined #openstack-dev | 22:33 | |
| *** stevemar has quit IRC | 22:34 | |
| *** rcleere has quit IRC | 22:35 | |
| *** wu_wenxiang has quit IRC | 22:35 | |
| *** dina_belova has joined #openstack-dev | 22:36 | |
| *** alunch has quit IRC | 22:36 | |
| *** moted has quit IRC | 22:38 | |
| *** moted has joined #openstack-dev | 22:39 | |
| *** dina_belova has quit IRC | 22:40 | |
| *** FunnyLookinHat has quit IRC | 22:45 | |
| *** sudorandom has quit IRC | 22:45 | |
| *** anniec has joined #openstack-dev | 22:48 | |
| *** pmathews has quit IRC | 22:49 | |
| *** xBsd has quit IRC | 22:54 | |
| *** Mandell has quit IRC | 22:55 | |
| *** Mandell has joined #openstack-dev | 22:55 | |
| *** edmund has quit IRC | 22:55 | |
| *** rnirmal has quit IRC | 22:56 | |
| *** nati_uen_ has joined #openstack-dev | 22:56 | |
| *** jasdeepH has quit IRC | 22:57 | |
| *** rods has quit IRC | 22:57 | |
| *** eharney has joined #openstack-dev | 22:57 | |
| *** eharney has quit IRC | 22:57 | |
| *** eharney has joined #openstack-dev | 22:57 | |
| *** fabio__ has quit IRC | 22:58 | |
| *** nati_ueno has quit IRC | 22:59 | |
| *** jasdeepH has joined #openstack-dev | 23:02 | |
| *** vaml has quit IRC | 23:04 | |
| *** jasondotstar has joined #openstack-dev | 23:04 | |
| *** stevemar has joined #openstack-dev | 23:06 | |
| *** fabio has joined #openstack-dev | 23:06 | |
| *** datsun180b has quit IRC | 23:06 | |
| *** fabio has quit IRC | 23:06 | |
| *** stevemar has quit IRC | 23:11 | |
| *** herndon has quit IRC | 23:11 | |
| *** zul has joined #openstack-dev | 23:16 | |
| *** gyee has quit IRC | 23:16 | |
| *** vladimir3p has joined #openstack-dev | 23:19 | |
| *** colinmcnamara has quit IRC | 23:21 | |
| *** tjones has quit IRC | 23:23 | |
| *** alexxu has joined #openstack-dev | 23:25 | |
| *** colinmcnamara has joined #openstack-dev | 23:26 | |
| *** pcm_ has quit IRC | 23:30 | |
| *** otherwiseguy has quit IRC | 23:30 | |
| *** jasdeepH has quit IRC | 23:31 | |
| *** danwent has quit IRC | 23:32 | |
| *** alunch has joined #openstack-dev | 23:32 | |
| *** tjones has joined #openstack-dev | 23:32 | |
| *** dstanek has quit IRC | 23:33 | |
| *** pcm_ has joined #openstack-dev | 23:34 | |
| *** pcm_ has quit IRC | 23:34 | |
| *** diogogmt has quit IRC | 23:35 | |
| *** alexb_ has quit IRC | 23:36 | |
| *** dina_belova has joined #openstack-dev | 23:36 | |
| *** alunch has quit IRC | 23:37 | |
| *** vipul is now known as vipul-away | 23:38 | |
| *** terryh has quit IRC | 23:41 | |
| *** dina_belova has quit IRC | 23:41 | |
| *** alunch has joined #openstack-dev | 23:43 | |
| *** vipul-away is now known as vipul | 23:53 | |
| *** Aj0 has quit IRC | 23:53 | |
| *** anniec has quit IRC | 23:56 | |
| *** anniec has joined #openstack-dev | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!