| jamielennox | un-deprecate? where's this coming from? | 00:02 |
|---|---|---|
| morganfainberg | jamielennox, http://lists.openstack.org/pipermail/openstack-dev/2014-March/031016.html | 00:03 |
| *** nkinder has quit IRC | 00:03 | |
| *** Trozz has joined #openstack-keystone | 00:03 | |
| jamielennox | morganfainberg: yea - i definetly get his point | 00:04 |
| jamielennox | does that need to be in for rc? | 00:05 |
| *** nkinder has joined #openstack-keystone | 00:06 | |
| morganfainberg | jamielennox, that is the point of the discussion | 00:06 |
| *** topol has quit IRC | 00:08 | |
| *** henrynash has quit IRC | 00:10 | |
| gyee | bknudson, morganfainberg, dstanek, you guys recall why we removed gettextutils from httpd/keystone.py? | 00:16 |
| gyee | k, maybe we just never add it there | 00:19 |
| morganfainberg | gyee, somethinG about lazy loading? | 00:19 |
| morganfainberg | gyee, not sure | 00:19 |
| gyee | yes, running under apache yield a bunch warnings related to _() | 00:19 |
| *** nkinder has quit IRC | 00:23 | |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: Discover should support other services https://review.openstack.org/72878 | 00:32 |
| *** morganfainberg is now known as morganfainberg_Z | 00:33 | |
| *** gokrokve has joined #openstack-keystone | 00:42 | |
| derek_c | openstackclient is giving me a "ERROR: openstackclient.shell Exception raised: Authorization failed: The resource could not be found. (HTTP 404)" | 00:53 |
| stevemar | derek_c, what did you issue? | 00:53 |
| derek_c | when I look into keystone's log, I see something like: "POST http://10.0.2.15:5000/v2.0/auth/tokens HTTP/1.0" 404 93 | 00:53 |
| derek_c | I mean, I'm just trying to run "openstack" | 00:53 |
| derek_c | not even with any commands | 00:53 |
| stevemar | derek_c, do you have any env vars set up? like username or password? | 00:54 |
| derek_c | stevemar: I'm pretty sure I sourced the OpenStack RC file downloaded from horizon | 00:55 |
| derek_c | the thing is it says 404, which is strange | 00:55 |
| derek_c | and also, I thought openstackclient uses v3 APIs by default? | 00:55 |
| gyee | jammielennox, gimme a shout when you have a chance, its about https://review.openstack.org/#/c/81977/ | 00:56 |
| dolphm | morganfainberg_Z: dstanek: the way i see it, we've achieved the most important goal -- communicating that the v2 API is going away | 00:56 |
| stevemar | derek_c, it would depend on what keystone endpoints are being used, if you used devstack it still sets v2.0 endpoints :\ | 00:56 |
| dolphm | morganfainberg_Z: dstanek: i didn't expect the K release to be completely without v2 anyway, so i don't think it's much of a dent in terms of realistic timing, either | 00:57 |
| stevemar | jamielennox, whats the deal with: https://github.com/openstack/python-keystoneclient/blob/master/doc/source/releases.rst should we give up on it? do a mass update - how do we figure out what we included? | 00:57 |
| derek_c | stevemar: hmm.. surely there must be a way to configure it to use V3 endpoint though? | 00:58 |
| jamielennox | stevemar: lol - no idea | 00:58 |
| jamielennox | ummm | 00:58 |
| stevemar | jamielennox, it's so very out of date | 00:58 |
| jamielennox | stevemar: yea - is it something that we can autogenerate do you think? | 00:59 |
| derek_c | stevemar: ah I see | 00:59 |
| jamielennox | the other option is that we just start enforcing in reviews that people add to the file when they add a feature and dolphm just seals it off before a release | 00:59 |
| derek_c | OS_IDENTITY_API_VERSION | 00:59 |
| stevemar | derek_c, http://paste.openstack.org/show/74307/ | 01:00 |
| openstackgerrit | Jamie Lennox proposed a change to openstack/keystone: Isolate backend loading https://review.openstack.org/74293 | 01:02 |
| openstackgerrit | Jamie Lennox proposed a change to openstack/keystone: Make Pecan the root routing framework https://review.openstack.org/65428 | 01:02 |
| stevemar | jamielennox, ugh, there isn't much info in blueprints | 01:02 |
| jamielennox | stevemar: there has been a few but in many cases one blueprint covered a lot of work | 01:02 |
| jamielennox | eg auth-plugins | 01:02 |
| stevemar | jamielennox, there isn't a releases file in keystone, maybe we don't need one in keystonelcient either? | 01:04 |
| *** theocean154 has joined #openstack-keystone | 01:04 | |
| jamielennox | stevemar: i'm happy enough to drop it consiering that it's useless anyway | 01:07 |
| openstackgerrit | Jamie Lennox proposed a change to openstack/keystone: Make Pecan the root routing framework https://review.openstack.org/65428 | 01:10 |
| *** marcoemorais1 has quit IRC | 01:20 | |
| openstackgerrit | Steve Martinelli proposed a change to openstack/python-keystoneclient: Remove releases.rst from keystone docs https://review.openstack.org/82962 | 01:24 |
| openstackgerrit | Steve Martinelli proposed a change to openstack/python-keystoneclient: Remove releases.rst from keystone docs https://review.openstack.org/82962 | 01:25 |
| stevemar | derek_c, try using v2.0, openstack user list should work | 01:27 |
| stevemar | it's possible that there is a bug where if just 'openstack' is issued, it comes up at garbage | 01:27 |
| stevemar | since it's not really a command... | 01:27 |
| stevemar | if you need help, try --help or -h | 01:27 |
| *** jamiec has joined #openstack-keystone | 01:34 | |
| ayoung | Why o why do I still need to do a full install to have a VM running Fedora on my local machine? I must figure out how to install a cloud image locally | 01:34 |
| *** devlaps1 has quit IRC | 01:36 | |
| *** theocean154 has quit IRC | 01:36 | |
| *** derek_c has quit IRC | 01:41 | |
| ayoung | jamielennox, yum search openssl | grep python .... | 01:43 |
| ayoung | m2crypto and python3-pyOpenSSL | 01:43 |
| ayoung | Just another reason to go Full Py3 | 01:43 |
| *** nkinder has joined #openstack-keystone | 01:43 | |
| jamielennox | sure - but there are plenty of better ones | 01:45 |
| jamielennox | the problem with trying to change a framework incrementally is everything ends up looking very much like the old onee | 01:48 |
| *** topol has joined #openstack-keystone | 01:49 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: revert deprecation of v2 API https://review.openstack.org/82963 | 01:54 |
| ayoung | jamielennox, I was just looking at what was available without having to become a package maintainer | 02:01 |
| ayoung | jamielennox, what do you suggest? | 02:01 |
| jamielennox | ayoung: what are you trying to do/ | 02:01 |
| ayoung | jamielennox, getus a crypto library | 02:01 |
| jamielennox | oh - i don't think either do what you want | 02:01 |
| ayoung | jamielennox, if we stop doing Eventlet, we can actually use a library | 02:02 |
| jamielennox | assuming you're still talking cms | 02:02 |
| ayoung | nope. But m2 would support the pysmime....we'd still need to own that | 02:02 |
| ayoung | and openssl probably would need to be extended | 02:02 |
| ayoung | could we do m2 for python27 and pyopenssl for py3? | 02:02 |
| ayoung | or maybe there is py27 support in the opensssl approach, we just don't ship | 02:03 |
| *** derek_c has joined #openstack-keystone | 02:03 | |
| jamielennox | ayoung: so pyopenssl is fairly limited | 02:03 |
| jamielennox | mostly just ssl | 02:03 |
| jamielennox | i don't think they have much in raw cert handling | 02:03 |
| jamielennox | m2 is much bigger and more likely to want that sort of thing | 02:04 |
| ayoung | but py 27 only | 02:04 |
| jamielennox | but i don't know if we want it | 02:04 |
| ayoung | then there is pycrypto | 02:04 |
| jamielennox | i like the idea of contributing to the cryptography library | 02:04 |
| jamielennox | i was wondering if i'd have the chance to do the nss backend for that | 02:05 |
| jamielennox | i think that's your best bet | 02:05 |
| ayoung | so we do pycrypto, and support openssl and NSS long term. | 02:05 |
| jamielennox | though they still haven't defined a certificate | 02:05 |
| jamielennox | pycrypto != cryptography | 02:05 |
| jamielennox | https://pypi.python.org/pypi/cryptography | 02:06 |
| ayoung | Maybe we just cut straight to nss...which still doesn't do CMS in python. | 02:06 |
| ayoung | How bad is M2? | 02:07 |
| ayoung | https://github.com/martinpaljak/M2Crypto | 02:08 |
| ayoung | https://github.com/martinpaljak/M2Crypto/commit/b03e643979bcdbdf0260696dcfee5ad59bacb6c0 | 02:09 |
| jamielennox | i'm really not sure | 02:11 |
| jamielennox | https://github.com/martinpaljak/M2Crypto/blob/master/M2Crypto/SMIME.py | 02:12 |
| jamielennox | i can't see how you'd actually use that interface though | 02:12 |
| ayoung | https://github.com/martinpaljak/M2Crypto/blob/master/M2Crypto/SMIME.py#L207 | 02:16 |
| ayoung | https://github.com/martinpaljak/M2Crypto/blob/master/M2Crypto/SMIME.py#L191 | 02:16 |
| ayoung | probably x = SMIME(); x. set_x509_store(); | 02:17 |
| ayoung | something like that...doesn't look horrible | 02:18 |
| stevemar | ayoung, mind if i push a new patch of the initialization scripts? | 02:21 |
| ayoung | stevemar, not in the slightest | 02:21 |
| jamielennox | ayoung: i'm going to have a stab and say the M2crypto writer was probably not an experience python user | 02:25 |
| ayoung | jamielennox, thats ok. Experienced Python people we have access too. | 02:25 |
| ayoung | to | 02:25 |
| jamielennox | i saw a tweet or something similar recently about how a good C programmer can make any language look like C | 02:26 |
| *** harlowja is now known as harlowja_away | 02:27 | |
| ayoung | jamielennox, if the only tool you have is duck tape, everything looks like a duck | 02:27 |
| jamielennox | it's been a while since i've done openssl and C but i'd almost say we're better off filling the gap with our own library unti cryptography catches up | 02:28 |
| ayoung | I'm guessing "no" | 02:28 |
| ayoung | I'm guessing we are better off fixing M2 | 02:29 |
| ayoung | I'd rather have something that works but is uglyt | 02:29 |
| jamielennox | fixing as in py3/ | 02:29 |
| ayoung | that too | 02:29 |
| ayoung | right now I am just trying to get the unit tests to run | 02:29 |
| *** zhiyan_ is now known as zhiyan | 02:32 | |
| *** mberlin has joined #openstack-keystone | 02:50 | |
| *** mberlin1 has quit IRC | 02:52 | |
| *** esmute has joined #openstack-keystone | 02:54 | |
| *** gyee has quit IRC | 02:56 | |
| *** devlaps has joined #openstack-keystone | 03:03 | |
| *** derek_c has quit IRC | 03:13 | |
| ayoung | jamielennox, lets see how this goes https://github.com/martinpaljak/M2Crypto/pull/20 | 03:19 |
| jamielennox | ayoung: so that's not a good sign | 03:20 |
| ayoung | jamielennox, I think that someone moved the directory | 03:20 |
| ayoung | but, yeah. | 03:21 |
| ayoung | OTOH hand, once I fixed it, they all ran | 03:21 |
| jamielennox | ayoung: yea the very last commit moved a bunch of tests | 03:21 |
| jamielennox | like 3 days ago | 03:21 |
| ayoung | My guess is that this is a code base we can work with | 03:22 |
| ayoung | jamielennox, I'm guessing a liberal use of six and we can get this py33 | 03:24 |
| ayoung | Anyway, I'm a head to bed. Thought you might find it interesting | 03:25 |
| *** ayoung has quit IRC | 03:26 | |
| *** gokrokve has quit IRC | 03:37 | |
| *** gokrokve_ has joined #openstack-keystone | 03:38 | |
| *** gokrokve_ has quit IRC | 03:42 | |
| *** wchrisj has quit IRC | 03:43 | |
| *** browne has quit IRC | 03:48 | |
| *** gokrokve has joined #openstack-keystone | 04:07 | |
| *** gokrokve has quit IRC | 04:13 | |
| *** david-lyle has joined #openstack-keystone | 04:16 | |
| *** dolphm has quit IRC | 04:25 | |
| *** dolphm has joined #openstack-keystone | 04:27 | |
| stevemar | is there a correct order in deleting keystone data? | 04:50 |
| stevemar | i would think it's users, projects, roles, domains, endpoints, services | 04:51 |
| *** YorikSar has quit IRC | 05:03 | |
| *** amcrn has quit IRC | 05:08 | |
| *** topol has quit IRC | 05:14 | |
| openstackgerrit | Steve Martinelli proposed a change to openstack/python-keystoneclient: Example Initialization scripts https://review.openstack.org/82687 | 05:15 |
| *** ChanServ sets mode: +o dolphm | 05:19 | |
| *** marcoemorais has joined #openstack-keystone | 05:20 | |
| *** devlaps has quit IRC | 05:43 | |
| *** stevemar has quit IRC | 05:58 | |
| openstackgerrit | Jenkins proposed a change to openstack/keystone: Imported Translations from Transifex https://review.openstack.org/78525 | 06:01 |
| *** gokrokve has joined #openstack-keystone | 06:08 | |
| *** bvandenh has quit IRC | 06:13 | |
| *** gokrokve has quit IRC | 06:13 | |
| *** YorikSar has joined #openstack-keystone | 06:14 | |
| *** saju_m has joined #openstack-keystone | 06:19 | |
| *** amcrn has joined #openstack-keystone | 06:51 | |
| *** derek_c has joined #openstack-keystone | 06:53 | |
| *** zoresvit has joined #openstack-keystone | 07:12 | |
| *** saju_m has quit IRC | 07:13 | |
| *** derek_c has quit IRC | 07:18 | |
| *** jaosorior has joined #openstack-keystone | 07:25 | |
| *** henrynash has joined #openstack-keystone | 07:28 | |
| *** marcoemorais has quit IRC | 07:28 | |
| openstackgerrit | Jamie Lennox proposed a change to openstack/keystone: Make Pecan the root routing framework https://review.openstack.org/65428 | 07:35 |
| *** henrynash has quit IRC | 07:40 | |
| openstackgerrit | A change was merged to openstack/keystone: Uses generator expressions instead of filter https://review.openstack.org/70816 | 07:55 |
| *** flaper87|afk is now known as flaper87 | 07:57 | |
| *** zoresvit has quit IRC | 08:03 | |
| *** bvandenh has joined #openstack-keystone | 08:07 | |
| *** gokrokve has joined #openstack-keystone | 08:08 | |
| *** gokrokve has quit IRC | 08:12 | |
| *** henrynash has joined #openstack-keystone | 08:28 | |
| *** chandankumar_ has joined #openstack-keystone | 08:42 | |
| *** henrynash has quit IRC | 08:50 | |
| *** leseb has joined #openstack-keystone | 08:54 | |
| *** henrynash has joined #openstack-keystone | 08:55 | |
| *** marcoemorais has joined #openstack-keystone | 08:57 | |
| *** marcoemorais has quit IRC | 09:01 | |
| openstackgerrit | Marcos FermÃn Lobo proposed a change to openstack/keystone: Unimplemented get roles by group for project list https://review.openstack.org/76470 | 09:11 |
| *** Gippa has joined #openstack-keystone | 09:13 | |
| *** gokrokve has joined #openstack-keystone | 09:13 | |
| *** gokrokve has quit IRC | 09:13 | |
| *** gokrokve has joined #openstack-keystone | 09:14 | |
| *** gokrokve has quit IRC | 09:16 | |
| *** gokrokve has joined #openstack-keystone | 09:17 | |
| *** gokrokve has quit IRC | 09:21 | |
| *** andreaf has joined #openstack-keystone | 09:22 | |
| *** Trozz has quit IRC | 09:28 | |
| *** Trozz has joined #openstack-keystone | 09:29 | |
| *** Trozz_ has joined #openstack-keystone | 09:30 | |
| *** Trozz_ has quit IRC | 09:30 | |
| *** Trozz has quit IRC | 09:30 | |
| *** Trozz has joined #openstack-keystone | 09:31 | |
| *** Trozz_ has joined #openstack-keystone | 09:31 | |
| *** mich5225 has joined #openstack-keystone | 09:32 | |
| *** Trozz has quit IRC | 09:33 | |
| *** Trozz_ has quit IRC | 09:33 | |
| *** mich5225 has quit IRC | 09:33 | |
| *** Trozz has joined #openstack-keystone | 09:33 | |
| *** Trozz has quit IRC | 09:39 | |
| *** Trozz has joined #openstack-keystone | 09:40 | |
| *** Trozz has quit IRC | 09:40 | |
| *** Trozz has joined #openstack-keystone | 09:43 | |
| *** gokrokve has joined #openstack-keystone | 09:44 | |
| *** Trozz has quit IRC | 09:45 | |
| *** marcoemorais has joined #openstack-keystone | 09:57 | |
| *** marcoemorais has quit IRC | 10:01 | |
| *** chandankumar_ has quit IRC | 10:22 | |
| *** Trozz has joined #openstack-keystone | 10:31 | |
| *** Trozz has quit IRC | 10:34 | |
| *** Trozz has joined #openstack-keystone | 10:41 | |
| *** Trozz has quit IRC | 10:42 | |
| *** Trozz has joined #openstack-keystone | 10:44 | |
| *** Gippa has quit IRC | 10:44 | |
| *** Trozz has quit IRC | 10:45 | |
| *** Gippa has joined #openstack-keystone | 10:48 | |
| *** marcoemorais has joined #openstack-keystone | 10:58 | |
| *** marcoemorais1 has joined #openstack-keystone | 11:00 | |
| *** marcoemorais1 has quit IRC | 11:00 | |
| *** marcoemorais has quit IRC | 11:00 | |
| *** marcoemorais has joined #openstack-keystone | 11:00 | |
| *** marcoemorais has quit IRC | 11:05 | |
| *** henrynash has quit IRC | 11:09 | |
| *** leseb has quit IRC | 11:14 | |
| *** leseb has joined #openstack-keystone | 11:14 | |
| *** leseb has quit IRC | 11:19 | |
| *** saju_m has joined #openstack-keystone | 11:21 | |
| *** Trozz has joined #openstack-keystone | 11:23 | |
| *** Trozz has quit IRC | 11:23 | |
| *** Trozz has joined #openstack-keystone | 11:29 | |
| *** Trozz has quit IRC | 11:36 | |
| thiagop | Hello everyone. I was investigating the use of domains on Horizon and suddenly bumped into a problem: When I'm using API v3 on Horizon, authenticate myself and send my token to Nova, Nova doesn't works using API v3 and send the token to be validated by v2.0 port | 11:45 |
| *** gokrokve has quit IRC | 11:45 | |
| *** gokrokve has joined #openstack-keystone | 11:46 | |
| thiagop | Talking to some colleagues here we thought of a solution that seems reasonable, but I can't view all the implications so, here it goes... | 11:46 |
| thiagop | We could create a way to distinguish a token v2 from a v3 and use it to create a port that could receive any token of any version and use the correct validation procedure for it | 11:48 |
| thiagop | what could be problematic on doing that? | 11:48 |
| *** gokrokve has quit IRC | 11:50 | |
| *** leseb has joined #openstack-keystone | 11:56 | |
| *** gokrokve has joined #openstack-keystone | 11:57 | |
| *** leseb has quit IRC | 12:00 | |
| *** marcoemorais has joined #openstack-keystone | 12:01 | |
| *** Trozz has joined #openstack-keystone | 12:02 | |
| *** raildo has joined #openstack-keystone | 12:03 | |
| *** Trozz has quit IRC | 12:03 | |
| *** wchrisj has joined #openstack-keystone | 12:04 | |
| *** marcoemorais has quit IRC | 12:05 | |
| *** topol has joined #openstack-keystone | 12:16 | |
| *** dims_ has quit IRC | 12:18 | |
| dolphm | thiagop: you should be able to configure auth_token to talk to v3 by setting auth_version = 3 in it's config | 12:18 |
| thiagop | dolphm: on Nova? | 12:18 |
| dolphm | thiagop: in nova's paste config for auth_token, yes | 12:18 |
| dolphm | thiagop: the v3 api can validate all v2 tokens, and v2 can validate default-domain tokens | 12:19 |
| *** dims_ has joined #openstack-keystone | 12:21 | |
| thiagop | dolphm: I'll try that, thanks! | 12:26 |
| *** david-lyle has quit IRC | 12:27 | |
| *** leseb has joined #openstack-keystone | 12:27 | |
| *** leseb has quit IRC | 12:43 | |
| dstanek | dolphm: i just saw a message from gyee (from last night) about a possible bug in httpd/keystone.py. do you happen to know if there is a review for it? i don't see one from gyee | 12:45 |
| dolphm | dstanek: i definitely don't have a bug report | 12:45 |
| dstanek | dolphm: i'll make one with the details i have to far | 12:46 |
| dolphm | dstanek: what's the gist? | 12:46 |
| dstanek | warnings about _() when running under apache | 12:46 |
| dolphm | dstanek: warnings? | 12:47 |
| dstanek | i haven't tried it yet so i don't know if it's a complete fail or just something logged | 12:47 |
| dolphm | dstanek: not the failure we saw in keystone-all i assume | 12:47 |
| dstanek | i don't think so: "gyee | yes, running under apache yield a bunch warnings related to _()" | 12:47 |
| dstanek | is devstack setup to run under apache? | 12:48 |
| dolphm | well that's vague | 12:48 |
| dolphm | dstanek: you can switch to apache with... | 12:48 |
| dstanek | if i can't reproduce i won't worry about it unless gyee says it's actually an issue | 12:48 |
| dolphm | there's a function called "is_apache_enabled_service" somewhere... | 12:51 |
| dolphm | dstanek: APACHE_ENABLED_SERVICES = key | 12:52 |
| dolphm | i think | 12:52 |
| dolphm | swift is also supported there | 12:52 |
| *** gokrokve has quit IRC | 12:53 | |
| *** gokrokve has joined #openstack-keystone | 12:53 | |
| dstanek | stacking now | 12:57 |
| *** gokrokve has quit IRC | 12:57 | |
| openstackgerrit | Marcos FermÃn Lobo proposed a change to openstack/keystone: Unimplemented get roles by group for project list https://review.openstack.org/76470 | 12:58 |
| *** leseb has joined #openstack-keystone | 13:00 | |
| * dolphm python-keystoneclient 0.7.0 is gating | 13:01 | |
| *** marcoemorais has joined #openstack-keystone | 13:01 | |
| *** bknudson has joined #openstack-keystone | 13:03 | |
| *** marcoemorais has quit IRC | 13:06 | |
| *** gokrokve has joined #openstack-keystone | 13:07 | |
| * dolphm https://pypi.python.org/pypi/python-keystoneclient/0.7.0 | 13:17 | |
| *** zhiyan is now known as zhiyan_ | 13:19 | |
| *** zhiyan_ is now known as zhiyan | 13:19 | |
| *** ayoung_ has joined #openstack-keystone | 13:21 | |
| dolphm | graphing the last year of load on the gate is interesting http://graphite.openstack.org/render/?from=-1year&width=1920&height=135&margin=0&hideLegend=true&hideAxes=true&hideGrid=true&target=color(stats.gauges.zuul.pipeline.gate.current_changes,%20%27000000%27)&bgcolor=ffffff&1395839942651 | 13:22 |
| *** ayoung_ is now known as ayoung | 13:22 | |
| *** bknudson has quit IRC | 13:23 | |
| dolphm | the 60 hour gate times is reflected by the big giant spike on the right | 13:23 |
| *** topol has quit IRC | 13:24 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Avoid using .values() on the indexed columns https://review.openstack.org/82793 | 13:26 |
| *** joesavak has joined #openstack-keystone | 13:27 | |
| *** leseb has quit IRC | 13:31 | |
| *** leseb has joined #openstack-keystone | 13:34 | |
| dstanek | dolphm: i can definitely see how translations would not work, but i can't reproduce yet | 13:35 |
| *** lbragstad has quit IRC | 13:38 | |
| *** stevemar has joined #openstack-keystone | 13:47 | |
| *** stevemar has quit IRC | 13:48 | |
| *** stevemar has joined #openstack-keystone | 13:48 | |
| *** nkinder has quit IRC | 13:49 | |
| stevemar | dolphm, what about the access and stats extension, weren't those marked deprecated, and need to be reverted? or because they are contrib, we still go ahead with it? | 13:53 |
| dstanek | ah, so i don't actually the the .mo translations installed | 13:55 |
| *** marcoemorais has joined #openstack-keystone | 14:02 | |
| *** browne has joined #openstack-keystone | 14:03 | |
| *** marcoemorais has quit IRC | 14:07 | |
| *** zuqiang has quit IRC | 14:14 | |
| *** topol has joined #openstack-keystone | 14:16 | |
| *** lbragstad has joined #openstack-keystone | 14:21 | |
| marekd | stevemar: hi. Just want to confirm: In the tests like here https://review.openstack.org/#/c/60820/19/keystoneclient/tests/v3/test_oauth1.py what is really checked is not the interaction with the server but rather the fact if keystoneclient correctly creates the requests? | 14:25 |
| openstackgerrit | A change was merged to openstack/keystone: Imported Translations from Transifex https://review.openstack.org/78525 | 14:32 |
| dolphm | stevemar: the reasoning behind deprecating those was different, i don't think there's an issue there | 14:36 |
| stevemar | dolphm, cool | 14:36 |
| stevemar | dolphm, just wanted to make sure we're covered | 14:36 |
| stevemar | marekd, yes, the test is to make sure the request is correctly formed, the headers, body and url | 14:37 |
| stevemar | marekd, the keystoneclient tests don't interact with the keystone server, if you want that, you can look at: https://github.com/openstack/keystone/blob/master/keystone/tests/test_keystoneclient.py#L111 | 14:38 |
| marekd | stevemar: and the way back? | 14:39 |
| stevemar | marekd, how does keystoneclient verify what was received? | 14:40 |
| *** chandan_kumar has quit IRC | 14:42 | |
| marekd | stevemar: attributes, structure etc. It's not for validating server side, rather validating whether you wrote the code that correctly parses the response... | 14:42 |
| marekd | stevemar: and this response you can *somehow* create, by hand.. | 14:42 |
| *** nkinder has joined #openstack-keystone | 14:42 | |
| *** chandankumar_ has joined #openstack-keystone | 14:44 | |
| marekd | stevemar: stevemar what's the point in creating such fake URL (I assume they are fake URL..) like here for instance? | 14:44 |
| marekd | stevemar: https://github.com/openstack/python-keystoneclient/blob/master/keystoneclient/tests/v3/test_groups.py#L41 | 14:45 |
| *** chandankumar_ has quit IRC | 14:46 | |
| *** chandan_kumar has joined #openstack-keystone | 14:47 | |
| *** chandan_kumar has quit IRC | 14:48 | |
| *** chandan_kumar has joined #openstack-keystone | 14:48 | |
| *** thedodd has joined #openstack-keystone | 14:48 | |
| stevemar | marekd, the point of the fake url is to make sure your new class is actually invoking that url. | 14:48 |
| marekd | stevemar: ah, ok :-) | 14:49 |
| marekd | stevemar: it was kinda unclear to me :-) | 14:49 |
| stevemar | marekd, so in that case, managers.lits(user=user_id) should trigger the url GET /users/user_id | 14:49 |
| marekd | stevemar: ok, triger URL and get some response? | 14:50 |
| stevemar | sort of | 14:50 |
| marekd | stevemar: because my initial understanding was: httpretty acts as a fake server, you basically pass to it what should be responded (just pure to-be-jsonized-dictionary), and call keystoneclient method, that will hit that URL, get response, parse it etc. | 14:51 |
| stevemar | you can't verify the server response, but you can verify that the class created the object with all the data | 14:51 |
| stevemar | yeah, you're not far off | 14:51 |
| stevemar | theres a lot of behind-the-scenes code going on in those tests, so it's a bit weird | 14:52 |
| marekd | stevemar: Already noticed it :-) | 14:52 |
| *** marcoemorais has joined #openstack-keystone | 14:52 | |
| stevemar | https://github.com/openstack/python-keystoneclient/blob/master/keystoneclient/tests/v3/test_auth.py | 14:52 |
| marekd | stevemar: especially 'default' test_{create,delete,get} methods :-) | 14:52 |
| stevemar | yeah :\ | 14:53 |
| stevemar | marekd, you can use stub_url to pass in a json parameter to act as the respose body | 14:53 |
| *** david-lyle has joined #openstack-keystone | 14:53 | |
| stevemar | marekd, in this case: https://github.com/openstack/python-keystoneclient/blob/master/keystoneclient/tests/v3/test_auth.py#L90 the response is checked | 14:54 |
| stevemar | err, not actually checked, you know what i mean | 14:54 |
| marekd | yep | 14:54 |
| stevemar | stub_auth is just stub_url but checks /auth specific stuff | 14:55 |
| *** chandan_kumar has quit IRC | 14:55 | |
| stevemar | https://github.com/openstack/python-keystoneclient/blob/master/keystoneclient/tests/v3/utils.py#L140 | 14:55 |
| marekd | yep, i was looking at it. | 14:55 |
| *** chandan_kumar has joined #openstack-keystone | 14:56 | |
| marekd | hmmm, stub_auth is a general method for *all* calls where authentication is required? | 14:56 |
| marekd | stevemar: erm...stupid me. | 14:57 |
| *** marcoemorais has quit IRC | 14:57 | |
| stevemar | marekd, yes, generic call for all auth related calls, you don't have to use it, but it's handy | 14:57 |
| marekd | because when I am calling test that adds IdP now what I get is Unauthorized error. | 14:59 |
| marekd | RESP BODY: {"error": {"message": "The request you have made requires authentication.", "code": 401, "title": "Unauthorized"}} | 14:59 |
| marekd | so i don't know if the solution for that is adding some extra parameter, like X-Auth-Token in kwargs in stub_uri() | 14:59 |
| *** bvandenh has quit IRC | 15:02 | |
| *** chandan_kumar has quit IRC | 15:03 | |
| *** vhoward has joined #openstack-keystone | 15:07 | |
| *** chandan_kumar has joined #openstack-keystone | 15:09 | |
| stevemar | marekd, show me some code in a pastie | 15:12 |
| marekd | stevemar: ok | 15:13 |
| *** wchrisj has quit IRC | 15:18 | |
| *** Gippa has quit IRC | 15:20 | |
| marekd | stevemar: ok, nvm, looks like i found where the problem was :-) | 15:24 |
| stevemar | yay | 15:24 |
| *** jsavak has joined #openstack-keystone | 15:30 | |
| *** bknudson has joined #openstack-keystone | 15:30 | |
| *** joesavak has quit IRC | 15:33 | |
| openstackgerrit | Marcos FermÃn Lobo proposed a change to openstack/keystone: Unimplemented get roles by group for project list https://review.openstack.org/76470 | 15:39 |
| *** saju_m has quit IRC | 15:43 | |
| dolphm | isn't there a bug that s/v2.0/v3/ https://review.openstack.org/#/c/75731/ can be attached to? | 15:44 |
| dolphm | that would have been in 0.7.0 if it was tracked ;( | 15:45 |
| dolphm | maybe https://bugs.launchpad.net/python-keystoneclient/+bug/1260916 | 15:45 |
| uvirtbot | Launchpad bug 1260916 in python-openstackclient "Setting --os-identity-api-version=3 still uses v2" [Undecided,Invalid] | 15:45 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/python-keystoneclient: Adds to Keystone to convert V2 endpoints to V3 https://review.openstack.org/75731 | 15:46 |
| *** joesavak has joined #openstack-keystone | 15:47 | |
| *** marcoemorais has joined #openstack-keystone | 15:49 | |
| *** jsavak has quit IRC | 15:51 | |
| *** wchrisj has joined #openstack-keystone | 15:52 | |
| *** jsavak has joined #openstack-keystone | 15:59 | |
| *** joesavak has quit IRC | 16:02 | |
| *** wchrisj has quit IRC | 16:05 | |
| *** wchrisj has joined #openstack-keystone | 16:07 | |
| ayoung | dolphm, https://review.openstack.org/#/c/75731/ thank you | 16:09 |
| ayoung | might be the killer feature for KC | 16:09 |
| dolphm | ayoung: i don't know about "killer" but it would be nice for icehouse :P | 16:10 |
| ayoung | dolphm, its one of the reasons we had to remove the deprecation from V2... | 16:10 |
| ayoung | it allows v2/v3 interop. That is big | 16:10 |
| openstackgerrit | David Stanek proposed a change to openstack/keystone: Enable lazy translations in httpd/keystone.py https://review.openstack.org/83113 | 16:10 |
| *** wchrisj has quit IRC | 16:12 | |
| dolphm | dstanek: know if that was a regression during icehouse? | 16:12 |
| dolphm | dstanek: stable/havana just has a gettextutils.install('keystone') | 16:13 |
| dstanek | dolphm: it slipped in on patch set 27 https://review.openstack.org/#/c/58766/ | 16:13 |
| dstanek | dolphm: httpd/keystone.py should have had the same change as bin/keystone-* | 16:15 |
| ayoung | dstanek, +A | 16:15 |
| ayoung | dolphm, anything gating RC1 at the moment? | 16:16 |
| dolphm | ayoung: that ^ now | 16:16 |
| dolphm | dstanek: thanks! | 16:16 |
| dolphm | ayoung: deprecation revert hasn't landed yet eitehr | 16:16 |
| ayoung | dolphm, looking | 16:16 |
| dolphm | ayoung: nor has the .values() removal from migration 43 | 16:17 |
| dolphm | ayoung: so, three patches to land until milestone-proposed | 16:17 |
| ayoung | dolphm, that one is progressing, I just checked | 16:17 |
| ayoung | dolphm, so nothing that still needs review, then? | 16:18 |
| openstackgerrit | A change was merged to openstack/identity-api: Add service name to service catalog https://review.openstack.org/81975 | 16:18 |
| dolphm | ayoung: i don't think so, we're just waiting on the electrons to shuffle about | 16:18 |
| dstanek | dolphm: i thought electrons were pretty fast. i think we are actually waiting for paint to dry | 16:20 |
| ayoung | dolphm, cool....change of subjects. jamielennox and I were looking at Crypto libraries last night. It looks like someone has picked up M2 as the maintainer. It seems to be the most full-featured of the set. Does SMIME, for example. Its Py26 and 27 only. | 16:20 |
| ayoung | I don;t think there is anything out there we can use without putting some effort in | 16:20 |
| ayoung | I suspect M2 is the best candidate | 16:20 |
| ayoung | I submitted this last night https://github.com/martinpaljak/M2Crypto/pull/20 | 16:21 |
| dolphm | dstanek: depends on if you're talking AC or DC :P | 16:21 |
| dolphm | ayoung: what is barbican using? | 16:22 |
| ayoung | Which indicates he's playing it a little fast and loose on commits, since his previous obviously broke all of the unit tests | 16:22 |
| dolphm | ayoung: and is m2 in openstack/requirements already? | 16:22 |
| ayoung | dolphm, I suspect that they are using pycrypto. m2 used to be, not sure if it was removed | 16:22 |
| *** leseb has quit IRC | 16:22 | |
| ayoung | https://github.com/openstack/requirements/blob/master/global-requirements.txt nope, it is gone | 16:23 |
| ayoung | pycrypto>=2.6 is in there | 16:23 |
| dstanek | ayoung: i think they do use pycrypto | 16:23 |
| ayoung | I suspect, though, that the amount of work we would have to put into pycrypto is much greater. | 16:23 |
| dolphm | ayoung: maybe a good cross-project session - crypto libraries in the community and reasons behind using/avoiding each | 16:24 |
| ayoung | Still doing research, of course | 16:24 |
| ayoung | dolphm, yeah, that is what I was thinking | 16:24 |
| dolphm | ayoung: start a mailing list discussion? | 16:24 |
| ayoung | Need to do some homework for that. | 16:24 |
| openstackgerrit | Jenkins proposed a change to openstack/keystone: Updated from global requirements https://review.openstack.org/82231 | 16:26 |
| *** joesavak has joined #openstack-keystone | 16:29 | |
| *** jsavak has quit IRC | 16:31 | |
| *** leseb has joined #openstack-keystone | 16:32 | |
| openstackgerrit | Jenkins proposed a change to openstack/python-keystoneclient: Updated from global requirements https://review.openstack.org/79695 | 16:33 |
| openstackgerrit | A change was merged to openstack/keystone: Avoid using .values() on the indexed columns https://review.openstack.org/82793 | 16:34 |
| openstackgerrit | Ilya Pekelny proposed a change to openstack/keystone: Sync test_migrations https://review.openstack.org/80618 | 16:35 |
| openstackgerrit | Ilya Pekelny proposed a change to openstack/keystone: Comparision of database models and migrations. https://review.openstack.org/80630 | 16:35 |
| *** devlaps has joined #openstack-keystone | 16:36 | |
| *** wchrisj has joined #openstack-keystone | 16:36 | |
| *** sneezewort has joined #openstack-keystone | 16:41 | |
| sneezewort | Do you guys have any good in-depth documentation of keystone endpoints? | 16:43 |
| dolphm | sneezewort: referring to the catalog of other services, or to keystone's own API endpoints? | 16:47 |
| *** marekd is now known as marekd|away | 16:48 | |
| sneezewort | dolphm, The API endpoints. | 16:49 |
| *** lbragstad has quit IRC | 16:49 | |
| dolphm | sneezewort: :5000/v2.0/ implements the "public service" described here https://github.com/openstack/identity-api/tree/master/openstack-identity-api/v2.0/src/docbkx | 16:50 |
| *** jaosorior has quit IRC | 16:50 | |
| dolphm | sneezewort: :35357/v2.0/ implements the "admin" interface described the corresponding wadl/xsd above | 16:51 |
| dolphm | sneezewort: /v3/ is implements https://github.com/openstack/identity-api/blob/master/openstack-identity-api/v3/src/markdown/identity-api-v3.md | 16:51 |
| dolphm | sneezewort: extensions define their own resources on top of the common apis | 16:52 |
| *** gokrokve has quit IRC | 16:55 | |
| *** chandan_kumar has quit IRC | 16:57 | |
| *** harlowja_away is now known as harlowja | 16:58 | |
| *** gokrokve_ has joined #openstack-keystone | 16:58 | |
| openstackgerrit | Thierry Carrez proposed a change to openstack/keystone: Open Juno development https://review.openstack.org/83132 | 16:59 |
| *** gpizza has joined #openstack-keystone | 17:03 | |
| dolphm | soon ^ | 17:03 |
| *** wchrisj has quit IRC | 17:05 | |
| *** leseb has quit IRC | 17:10 | |
| openstackgerrit | A change was merged to openstack/keystone: revert deprecation of v2 API https://review.openstack.org/82963 | 17:12 |
| *** gpizza has quit IRC | 17:13 | |
| *** amcrn has quit IRC | 17:15 | |
| sneezewort | dolphm, Thanks! | 17:15 |
| *** gyee has joined #openstack-keystone | 17:40 | |
| *** gokrokve_ has quit IRC | 17:41 | |
| gyee | dolphm, thanks for https://bugs.launchpad.net/bugs/1297922 | 17:46 |
| uvirtbot | Launchpad bug 1297922 in keystone "Internationalization is broken when using httpd/keystone.py" [Medium,In progress] | 17:46 |
| *** leseb has joined #openstack-keystone | 17:48 | |
| dolphm | gyee: you should have reported it as a bug ASAP! | 17:49 |
| *** lbragstad has joined #openstack-keystone | 17:50 | |
| *** morganfainberg_Z is now known as morganfainberg | 17:50 | |
| *** pcargnel_ has joined #openstack-keystone | 17:53 | |
| morganfainberg | mornin' | 17:54 |
| dstanek | gyee: you also said something about warnings for _() - where were you seeing that? | 17:54 |
| *** amcrn has joined #openstack-keystone | 17:55 | |
| dolphm | morganfainberg: o/ | 17:56 |
| morganfainberg | dolphm, how goes? (saw the removal of V2 deprecation) | 17:56 |
| morganfainberg | dolphm, i'm wondering if we can roll up all of V2 into a middleware that translates v3... | 17:57 |
| dolphm | morganfainberg: waiting on one last bug fix (i18n+httpd) and we'll open for juno | 17:57 |
| gyee | dstanek, a follow HP dude noticed them in the logs | 17:57 |
| morganfainberg | dolphm, +++++++ | 17:57 |
| morganfainberg | dolphm, YAY! | 17:57 |
| dolphm | morganfainberg: and yes, i'd like to do that specifically for the basic auth flow | 17:57 |
| morganfainberg | dolphm, that means it's time to crank on getting revoke and ephemeral tokens moving. | 17:57 |
| *** pcargnel_ has quit IRC | 17:57 | |
| dolphm | morganfainberg: ++ | 17:57 |
| *** saju_m has joined #openstack-keystone | 17:58 | |
| dstanek | gyee: do you have access to those logs? i'd be interested to see if my fix actually fixes that issue | 17:59 |
| gyee | dstanek, sure, I'll attach them to the bug | 18:00 |
| dstanek | gyee: my fix was targeting the problem of always translating to the default locale | 18:00 |
| openstackgerrit | Morgan Fainberg proposed a change to openstack/keystone: Use in-memory SQLite for testing https://review.openstack.org/82917 | 18:01 |
| openstackgerrit | Morgan Fainberg proposed a change to openstack/keystone: Use in-memory SQLite for sql migration tests https://review.openstack.org/82918 | 18:01 |
| morganfainberg | dolphm, ^ everything should be in place for way better testing once we jump into juno. | 18:01 |
| morganfainberg | dolphm, I think... I think... we're starting in a much better place than before. | 18:02 |
| morganfainberg | dolphm, also, the transient error that we occassionally see in multiple workers I've seen a couple times now. I think it's a teardown issue because it's something wrong with the FAKELDAP impl. | 18:03 |
| gyee | dstanek, I bet my beer money you fixed it, its the same as https://github.com/openstack/keystone/blob/master/bin/keystone-all#L37 | 18:03 |
| gyee | but I'll reverify | 18:03 |
| ayoung | morganfainberg, this makes me happy as a little schoolgirl! https://review.openstack.org/#/c/82917/2 | 18:07 |
| morganfainberg | ayoung, :) | 18:07 |
| ayoung | dstanek, we holding off on +A on https://review.openstack.org/#/c/81720/7 due to gate? | 18:08 |
| morganfainberg | ayoung, wait for juno | 18:08 |
| ayoung | morganfainberg, ++ | 18:08 |
| morganfainberg | ayoung, :) we're close and no reason to chance anything. | 18:09 |
| ayoung | J0 should be cut at the time of the summit! | 18:09 |
| gyee | don't mess up the karma now | 18:09 |
| ayoung | ++ | 18:09 |
| dstanek | ayoung: yup - i probably should have put that in there to that effect | 18:09 |
| ayoung | gyee, wanna see something? | 18:09 |
| ayoung | https://github.com/martinpaljak/M2Crypto/pull/20 gyee | 18:10 |
| ayoung | gyee, not 100% sure, but I think M2 might be the right way forward. Doing my homework now. It would need some love...specially in a python3 sort of way | 18:10 |
| gyee | ayoung, yeah, I agree | 18:11 |
| morganfainberg | ayoung, ugh another pull request based lib. | 18:11 |
| * morganfainberg dislikes pull-requests | 18:11 | |
| ayoung | gyee, do you guys have an in house maintainer for M2? | 18:11 |
| morganfainberg | gerrit workflow is better imo | 18:11 |
| ayoung | morganfainberg, meh...its what github does. Better than we had in the past | 18:11 |
| gyee | ayoung, not officially afaik | 18:12 |
| morganfainberg | ayoung, true, but still. | 18:12 |
| gyee | ayoung, lemme kick off an internal email to find out | 18:12 |
| morganfainberg | ayoung, gerrit > github | 18:12 |
| ayoung | gyee, hows about unofficially? Someone we could lean on? | 18:12 |
| ayoung | nice | 18:12 |
| gyee | ayoung, I can help | 18:12 |
| morganfainberg | ayoung, you thinking of resurrecting M2 (inc. for py33?) | 18:12 |
| ayoung | morganfainberg, contemplating. not committed | 18:12 |
| morganfainberg | it looks maintained-ish | 18:12 |
| morganfainberg | now | 18:12 |
| ayoung | morganfainberg, resurrected, I would say | 18:12 |
| morganfainberg | lol | 18:13 |
| ayoung | morganfainberg, its more full featured than any of the other openssl libraries | 18:13 |
| ayoung | so question is how hard would Py3 be? | 18:13 |
| gyee | m2 is just a openssl wrap | 18:13 |
| ayoung | Cuz whatever we would chose would require work | 18:13 |
| morganfainberg | ayoung, iirc it was a large refactor to make it py3 happy | 18:13 |
| morganfainberg | ayoung, which was why it was discounted (along with, more importantly, that it wasn't maintained) | 18:14 |
| dstanek | late lunch time! | 18:14 |
| *** dstanek is now known as dstanek_afk | 18:14 | |
| ayoung | morganfainberg, I suspect that is true, but I suspect that with what I've seen in Key Client that is going to be true of any libraray | 18:14 |
| ayoung | p2/3 interop is not trivial | 18:14 |
| ayoung | and we are going to need both for a while. | 18:14 |
| morganfainberg | ayoung, worse when you're doing c-bindings | 18:15 |
| ayoung | Not a Py3 rewrite | 18:15 |
| ayoung | morganfainberg, one possibility is M2 for py27 and openssl for Py3 | 18:15 |
| gyee | morganfainberg, more scary than that, code is generated from the SWIG | 18:15 |
| gyee | so we have to make sure the generated code is p3 happy | 18:16 |
| gyee | py3 I mean | 18:16 |
| ayoung | is SWIG capable of producing Py3 code? | 18:16 |
| morganfainberg | gyee, swig... euuuuuwwww | 18:16 |
| gyee | yep swig :) | 18:16 |
| morganfainberg | ayoung, https://bitbucket.org/hpk42/tox/pull-request/86/support-optional-env-variable-substitution/diff once that is in, and we move to that vintage of tox (some work) we can convert 100% of run_tests over to using tox | 18:17 |
| ayoung | morganfainberg, they all suck, just in different ways | 18:17 |
| ayoung | NICE | 18:18 |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: Adds to Keystone to convert V2 endpoints to V3 https://review.openstack.org/75731 | 18:24 |
| openstackgerrit | Jenkins proposed a change to openstack/keystone: Updated from global requirements https://review.openstack.org/82231 | 18:27 |
| openstackgerrit | A change was merged to openstack/keystone: Enable lazy translations in httpd/keystone.py https://review.openstack.org/83113 | 18:30 |
| dolphm | woo | 18:30 |
| *** amerine has joined #openstack-keystone | 18:32 | |
| openstackgerrit | Jenkins proposed a change to openstack/python-keystoneclient: Updated from global requirements https://review.openstack.org/79695 | 18:33 |
| *** dstanek_afk has quit IRC | 18:36 | |
| openstackgerrit | Steve Martinelli proposed a change to openstack/keystone: Remove _delete_tokens function from federation controller https://review.openstack.org/83169 | 18:41 |
| stevemar | dolphm, ^ | 18:42 |
| *** zhiyan is now known as zhiyan_ | 18:52 | |
| *** jamielennox is now known as jamielennox|away | 18:54 | |
| *** pcargnel has joined #openstack-keystone | 18:55 | |
| *** thedodd has quit IRC | 19:05 | |
| *** jaosorior has joined #openstack-keystone | 19:06 | |
| *** wchrisj has joined #openstack-keystone | 19:14 | |
| ayoung | morganfainberg, so, in our tox setup, the requirements.txt file is processed to install the reqs into the tox venv. What makes that happen? | 19:15 |
| ayoung | [testenv] | 19:16 |
| ayoung | usedevelop = True | 19:16 |
| ayoung | install_command = pip install -U {opts} {packages} | 19:16 |
| ayoung | 19:16 | |
| morganfainberg | ayoung, if you look in tox.ini https://github.com/openstack/keystone/blob/master/tox.ini#L10 | 19:21 |
| ayoung | morganfainberg, yeah...still some magix in there | 19:21 |
| ayoung | install_command = pip install -U {opts} {packages} | 19:22 |
| ayoung | what builds {packages} | 19:22 |
| morganfainberg | ayoung, http://tox.readthedocs.org/en/latest/config.html#confval-install_command=ARGV | 19:22 |
| morganfainberg | it's a tox substitution | 19:23 |
| *** leseb has quit IRC | 19:24 | |
| morganfainberg | dolphm, minutes... | 19:25 |
| dolphm | morganfainberg: 17! | 19:25 |
| morganfainberg | :) | 19:25 |
| openstackgerrit | A change was merged to openstack/keystone: Open Juno development https://review.openstack.org/83132 | 19:27 |
| ayoung | w00t! | 19:27 |
| morganfainberg | dolphm, that was less than 10 | 19:27 |
| morganfainberg | less than 17 | 19:27 |
| morganfainberg | ayoung, i think this means we need to get working on Ephemeral tokens and getting revoke up to snuff. | 19:31 |
| morganfainberg | ayoung, cause... that should be ready before summit imo | 19:32 |
| *** leseb has joined #openstack-keystone | 19:32 | |
| ayoung | w00t woot! | 19:33 |
| ayoung | morganfainberg, Time to start on Juno. | 19:33 |
| *** leseb has quit IRC | 19:34 | |
| dolphm | i need to go unblock things | 19:34 |
| dolphm | all the things | 19:34 |
| *** andreaf has quit IRC | 19:34 | |
| *** pcargnel has quit IRC | 19:44 | |
| *** dstanek_afk has joined #openstack-keystone | 19:52 | |
| dolphm | bknudson: restore? https://review.openstack.org/#/c/70630/ | 19:53 |
| dolphm | dstanek_afk: restore? https://review.openstack.org/#/c/74447/ | 19:53 |
| *** dstanek_afk is now known as dstanek | 19:53 | |
| dolphm | bknudson: restore? https://review.openstack.org/#/c/75708/ | 19:53 |
| dolphm | jamielennox|away: restore? https://review.openstack.org/#/c/78127/ | 19:53 |
| dstanek | dolphm: restored | 19:54 |
| dolphm | morganfainberg: definitely restore ASAP https://review.openstack.org/#/c/78169/ | 19:54 |
| dolphm | morganfainberg: i also need to restore my empty migrations patch | 19:54 |
| dolphm | dstanek: thanks | 19:54 |
| dolphm | dstanek: same https://review.openstack.org/#/c/73368/ | 19:55 |
| dolphm | dstanek: https://review.openstack.org/#/c/74623/ | 19:55 |
| dolphm | bknudson: https://review.openstack.org/#/c/80401/ | 19:56 |
| bknudson | dolphm: ? | 19:56 |
| dstanek | all the things in that patchset are now restored - looks like i have quite a bit of rebasing ahead of me | 19:56 |
| dolphm | bknudson: restore! | 19:57 |
| bknudson | dolphm: https://review.openstack.org/#/c/80401/ is not abandoned | 19:57 |
| bknudson | it will probably be abandoned tomorrow | 19:57 |
| dolphm | bknudson: my mistake! | 20:00 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Add placeholders for reserved migrations https://review.openstack.org/70153 | 20:07 |
| bknudson | is there a v3 api to get the list of extensions? | 20:07 |
| dolphm | bknudson: no - think extensions should inject themselves as links from existing resources | 20:07 |
| dolphm | bknudson: even if that's from /v3/ as OS-FEDERATION might do | 20:08 |
| *** thedodd has joined #openstack-keystone | 20:08 | |
| dolphm | this should land before anything else in juno: https://review.openstack.org/#/c/70153/ | 20:09 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Add placeholders for reserved migrations https://review.openstack.org/70153 | 20:10 |
| dolphm | bknudson: just fixed a typo ^ s/JUno/Juno/ | 20:10 |
| *** saju_m has quit IRC | 20:11 | |
| bknudson | /v2.0/extensions has OS-FEDERATION | 20:12 |
| bknudson | http://api.openstack.org/api-ref-identity.html#identity-v3-ext -- says /v3/extensions :( | 20:13 |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: Add 'methods' to all v3 test tokens https://review.openstack.org/74955 | 20:18 |
| openstackgerrit | Steve Martinelli proposed a change to openstack/keystone: Remove _delete_tokens function from federation controller https://review.openstack.org/83169 | 20:21 |
| *** derek_c has joined #openstack-keystone | 20:22 | |
| *** joesavak has quit IRC | 20:29 | |
| *** saju_m has joined #openstack-keystone | 20:30 | |
| openstackgerrit | Steve Martinelli proposed a change to openstack/python-keystoneclient: Example Initialization scripts https://review.openstack.org/82687 | 20:34 |
| *** david-lyle has quit IRC | 20:34 | |
| *** devlaps has quit IRC | 20:35 | |
| *** saju_m has quit IRC | 20:37 | |
| dolphm | 51 bugs fixed during RC, nice work everyone :) | 20:38 |
| morganfainberg | dolphm, i'll bring that one back soon | 20:41 |
| dolphm | morganfainberg: https://review.openstack.org/#/c/70153/ | 20:41 |
| *** devlaps has joined #openstack-keystone | 20:41 | |
| *** nkinder has quit IRC | 20:42 | |
| morganfainberg | dolphm, so we're going wiht that for sure this time? | 20:42 |
| morganfainberg | +2 | 20:43 |
| dolphm | morganfainberg: might as well have room for the possibility, right? | 20:43 |
| morganfainberg | dolphm, +2'd | 20:43 |
| morganfainberg | dolphm, about to propose the removal of the assignment proxy | 20:44 |
| dolphm | morganfainberg: IN_MEM_DB_CONN_STRING - but why didn't you abbreviate STRING too? | 20:48 |
| morganfainberg | dolphm, :P | 20:49 |
| morganfainberg | dolphm, what it's a descriptive (ish) variable name | 20:49 |
| *** topol has quit IRC | 20:50 | |
| dolphm | morganfainberg: IN_MEM_SQL_DB_CONN_STR <- i would have used those extra three characters for something more productive | 20:50 |
| morganfainberg | dolphm, oh yes! TEST_RUN_CONCURRENCY=8 tox -epy27 -- --subunit | subunit-2to1 | tools/colorizer.py | 20:50 |
| morganfainberg | thats right run concurrency of 8 | 20:50 |
| dolphm | yeah i can't match that | 20:50 |
| bknudson | the concurrency run often fails for me | 20:51 |
| morganfainberg | dolphm, mbp 13 only have i5? | 20:51 |
| bknudson | usually in fakeldap | 20:51 |
| bknudson | but I also saw it fail once on the keystoneclient tests | 20:51 |
| morganfainberg | bknudson, that is the bug i need to smash out before we enable it for everything | 20:51 |
| dolphm | morganfainberg: mine has an i7 | 20:51 |
| dolphm | morganfainberg: i guess i could do 8 with threading | 20:51 |
| morganfainberg | dolphm, 8 should be good. | 20:51 |
| morganfainberg | bknudson, the ksc one is really infrequent compared to the fakeldap one | 20:52 |
| morganfainberg | bknudson, but i haven't seen the issue enough to really solve it. | 20:52 |
| morganfainberg | bknudson, it's an attribute error, so something wrong w/ setup/teardown? | 20:53 |
| morganfainberg | bknudson, i'm planning on poking at it as I can to try and get us to using concurrency by default | 20:53 |
| bknudson | morganfainberg: I think it was git complaining | 20:53 |
| morganfainberg | bknudson, the keystoneclient ones? | 20:53 |
| bknudson | morganfainberg: yes | 20:53 |
| morganfainberg | bknudson, yeah, i think we need to collapse the two keystoneclient files and that will likely go away | 20:54 |
| morganfainberg | bknudson, i _think_ | 20:54 |
| *** leseb has joined #openstack-keystone | 20:54 | |
| morganfainberg | bknudson, the fakeldap one, i'm not really sure about | 20:55 |
| morganfainberg | bknudson, ooh interestingly i have a failure happening consistently now. | 20:55 |
| morganfainberg | bknudson, maybe i can fix it. | 20:55 |
| bknudson | morganfainberg can fix it. | 20:56 |
| morganfainberg | bknudson, lol | 20:56 |
| dstanek | dolphm: your fix to replace len(idx.columns.values()) fixes a bug in Py3, win win | 20:59 |
| morganfainberg | dstanek, yay! | 21:00 |
| dstanek | morganfainberg: i should have something today for fixing the tests under python3 - i've been poking at it in between other things | 21:01 |
| morganfainberg | dstanek, sounds good | 21:01 |
| dstanek | i've haven't really had to do much to get tests working - right now though testr just dies and i don't know how many succeed or fail | 21:02 |
| dstanek | i get path the cert creation without errors because i see that output just fine | 21:02 |
| morganfainberg | dolphm, do we want a "removed as of Juno" bp? | 21:02 |
| morganfainberg | dolphm, since we ahve a deprecated as of | 21:03 |
| openstackgerrit | Morgan Fainberg proposed a change to openstack/keystone: Remove assignment proxy methods/controllers https://review.openstack.org/83219 | 21:05 |
| morganfainberg | bknudson, i think i foudn the issue w/ fakeldap. | 21:08 |
| morganfainberg | bknudson, looking into it now. | 21:08 |
| *** wchrisj has quit IRC | 21:11 | |
| bknudson | morganfainberg: is this the error you see - UnexpectedMethodCallError(self, None) | 21:11 |
| morganfainberg | bknudson, yeah it's a mox stubout issue | 21:11 |
| bknudson | seems odd to use mox on a fake. | 21:12 |
| morganfainberg | bknudson, it looks like we may not do stubout cleanly in the test cases. | 21:12 |
| morganfainberg | bknudson, it's to stubout the connection in the ldap lib | 21:12 |
| morganfainberg | iirc | 21:12 |
| morganfainberg | so we don't _actually_ connect to an ldap server | 21:12 |
| bknudson | the fake stuff doesn't connect to an ldap server | 21:12 |
| morganfainberg | bknudson, oh it doesn't? | 21:13 |
| morganfainberg | hmm. | 21:13 |
| bknudson | no, it's a fake | 21:13 |
| morganfainberg | lol | 21:13 |
| morganfainberg | sure | 21:13 |
| morganfainberg | also, we're not clearing the handlers out in common_ldap between test cases | 21:13 |
| morganfainberg | not sure if we care about that | 21:13 |
| bknudson | might as well clean up after ourselves | 21:13 |
| morganfainberg | bknudson, that is my thought | 21:14 |
| bknudson | it'll come back to bite us sometime. | 21:14 |
| morganfainberg | bknudson, i'll add an addCleanup for that | 21:14 |
| bknudson | could have a fixture | 21:14 |
| morganfainberg | bknudson, converting to a fixture ++, but i think i'm going to focus on cleanup needed to solve this issue first | 21:15 |
| morganfainberg | bknudson, if it works, it might be easier to convert to a fixture cleanly | 21:15 |
| *** dstanek has quit IRC | 21:26 | |
| *** dstanek has joined #openstack-keystone | 21:28 | |
| morganfainberg | bknudson, is this normal mox usage https://github.com/openstack/keystone/blob/master/keystone/tests/test_backend_ldap.py#L947 bind conn to a self.CreateMockAnything then rebind it to the fakeldap.FakeLdap object? | 21:28 |
| bknudson | morganfainberg: This looks like it's supposed to be testing that chase_referrals is called with True.. | 21:32 |
| bknudson | seems like it wouldn't be difficult to have FakeLdap record that it was called with chase_referrals=True and expose a function that returns if chase_referrals is true or not | 21:33 |
| bknudson | I'd have to reimplement the test case to see how difficult it is. | 21:33 |
| morganfainberg | bknudson, that was my thought | 21:33 |
| morganfainberg | bknudson, the way i see it we could remove mox completely here pretty easily | 21:35 |
| *** jamielennox|away is now known as jamielennox | 21:35 | |
| *** leseb has quit IRC | 21:57 | |
| *** nkinder has joined #openstack-keystone | 22:00 | |
| morganfainberg | bknudson, i think i have a fix ready to go for the mox stuff in the backend. about to post it | 22:02 |
| morganfainberg | bknudson, test_backend_ldap that is | 22:02 |
| *** derek_c has quit IRC | 22:03 | |
| *** bknudson has quit IRC | 22:09 | |
| *** dstanek has quit IRC | 22:19 | |
| jamielennox | like 3 days ago | 22:35 |
| jamielennox | bah - up + enter | 22:35 |
| openstackgerrit | Jamie Lennox proposed a change to openstack/keystone: Make service catalog include service name https://review.openstack.org/78127 | 22:38 |
| jamielennox | dolphm: restored ^ | 22:39 |
| *** thedodd has quit IRC | 22:45 | |
| *** sneezewort has quit IRC | 22:48 | |
| openstackgerrit | Jamie Lennox proposed a change to openstack/keystone: Isolate backend loading https://review.openstack.org/74293 | 22:49 |
| openstackgerrit | Jamie Lennox proposed a change to openstack/keystone: Make Pecan the root routing framework https://review.openstack.org/65428 | 22:49 |
| *** dstanek has joined #openstack-keystone | 22:49 | |
| openstackgerrit | Morgan Fainberg proposed a change to openstack/keystone: Cleanup ldap tests (mox and reset values) https://review.openstack.org/83235 | 22:55 |
| *** harlowja has quit IRC | 22:58 | |
| *** jaosorior has quit IRC | 23:00 | |
| *** harlowja has joined #openstack-keystone | 23:03 | |
| *** bknudson has joined #openstack-keystone | 23:05 | |
| dstanek | turns out paste isn't py3 friendly yet | 23:09 |
| morganfainberg | dstanek, doh! | 23:09 |
| *** Trozz has joined #openstack-keystone | 23:10 | |
| *** dims_ has quit IRC | 23:12 | |
| *** flaper87 is now known as flaper87|afk | 23:18 | |
| *** dims_ has joined #openstack-keystone | 23:30 | |
| *** Trozz has quit IRC | 23:40 | |
| jamielennox | morganfainberg: i'm not sure if it's going to be feasible to use cfg.DeprecatedOpt in auth_token | 23:41 |
| jamielennox | a couple of things: | 23:41 |
| morganfainberg | jamielennox, figured i'd ask before hand though | 23:42 |
| jamielennox | 1st there is a mix of arguments from paste and CONF which have to co-exist and i can't figure out how to know if identity_uri was the default or provided | 23:42 |
| jamielennox | 2nd if i remove the options from the CONF then other services that do testing by setting the CONF will fail because the options will no longer be registered | 23:43 |
| morganfainberg | jamielennox, there is the dumb way to tell default | 23:43 |
| morganfainberg | jamielennox, set the default external to the opt definition, and compart | 23:43 |
| morganfainberg | compare* | 23:43 |
| jamielennox | right - but the default is http://localhost:35357 which is what eg devstack will actually set | 23:43 |
| jamielennox | so the default is fairly | 23:44 |
| jamielennox | useful | 23:44 |
| morganfainberg | jamielennox, i'll rescind my -1, | 23:46 |
| morganfainberg | jamielennox, just figured i'd ask. | 23:46 |
| jamielennox | morganfainberg: that's ok - i'll put in a big comment as to why i do it | 23:46 |
| morganfainberg | well, rescind the -1 for the opt, but this looks like it justifies a BP | 23:46 |
| morganfainberg | so we can track correlated patches / work | 23:47 |
| jamielennox | morganfainberg: yep - i'm doing that | 23:47 |
| morganfainberg | cool | 23:47 |
| openstackgerrit | Jamie Lennox proposed a change to openstack/python-keystoneclient: Tests should use identity_uri by default https://review.openstack.org/77492 | 23:54 |
| openstackgerrit | Jamie Lennox proposed a change to openstack/python-keystoneclient: Replace auth fragements with identity_uri https://review.openstack.org/77491 | 23:54 |
| ayoung | jamielennox, I'm not sure I am so hot on your organizational scheme direction in https://review.openstack.org/#/c/65428/17 for v2/v3. I'd rather keep the code for the modules together. | 23:55 |
| ayoung | Not so much this patch, as where you seem to be headed | 23:55 |
| openstackgerrit | Jamie Lennox proposed a change to openstack/python-keystoneclient: Rename request_uri to identity_uri https://review.openstack.org/77748 | 23:56 |
| morganfainberg | ayoung, about 10mins frmo having sqlite in-mem in juno | 23:57 |
| jamielennox | ayoung: i'm not as convinced on pecan as i was - it's a bit less easy without WSME but i think the organization is much better | 23:57 |
| morganfainberg | also this should fix one of the bigger transient errors for making concurrent tests the default https://review.openstack.org/#/c/83235/ | 23:57 |
| *** Mario_ has joined #openstack-keystone | 23:58 | |
| jamielennox | the information in eg keystone/token should be how to interact with the token - this doesn't need to be only from the /token URL | 23:58 |
| Mario_ | hello | 23:58 |
| jamielennox | every time i have to find the /auth/tokens route i need to try and remember if that lives in the /auth /tokens or some other directory | 23:59 |
| jamielennox | it's splitting the backend from the control layer | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!