| *** rpedde is now known as rpedde_away | 00:00 | |
| *** cloudchimp has quit IRC | 00:01 | |
| *** pcm_ has quit IRC | 00:02 | |
| *** dontalton has quit IRC | 00:02 | |
| *** davidha has quit IRC | 00:02 | |
| *** davidha has joined #openstack-dev | 00:04 | |
| *** eglynn has quit IRC | 00:11 | |
| *** winston-d_ has joined #openstack-dev | 00:13 | |
| *** Nachi has joined #openstack-dev | 00:14 | |
| *** Nachi is now known as nati_ueno_ | 00:14 | |
| ayoung | jaypipes, If I use the sql alchemy test for Is Null (blah.field == none) , I get a Pep 8 error. Is there away around this? | 00:15 |
|---|---|---|
| *** henrynash has quit IRC | 00:15 | |
| *** nati_ueno has quit IRC | 00:15 | |
| jaypipes | ayoung: lol, no idea! :( | 00:16 |
| ayoung | jaypipes, I want to suppress just that warning on just one file | 00:16 |
| ayoung | E711 | 00:16 |
| *** AlanClark has quit IRC | 00:17 | |
| *** AlanClark has joined #openstack-dev | 00:17 | |
| *** sacharya has joined #openstack-dev | 00:19 | |
| jog0 | ayoung: in nova we just ignore all E711s | 00:21 |
| *** olaph has quit IRC | 00:22 | |
| *** reed has quit IRC | 00:23 | |
| *** olaph has joined #openstack-dev | 00:23 | |
| *** vipul|away is now known as vipul | 00:24 | |
| *** zodiak has quit IRC | 00:25 | |
| *** sandywalsh has quit IRC | 00:27 | |
| *** bswartz has joined #openstack-dev | 00:28 | |
| *** markwash has quit IRC | 00:33 | |
| openstackgerrit | A change was merged to openstack/quantum: Imported Translations from Transifex https://review.openstack.org/22921 | 00:37 |
| clarkb | ayoung: I think you are supposed to say if blah.field is None? | 00:37 |
| clarkb | you use identity against None not equality | 00:37 |
| ayoung | clarkb, but sqlalchemy doesn;t like that | 00:37 |
| ayoung | clarkb, I think they overload an operator or something | 00:38 |
| clarkb | :( | 00:38 |
| *** digitalsanctum has quit IRC | 00:38 | |
| *** Mandell has quit IRC | 00:40 | |
| *** CaptTofu has quit IRC | 00:41 | |
| *** ilblackdragon has quit IRC | 00:41 | |
| matiu | So I made this small change on top of that xml parsing security bug | 00:41 |
| matiu | https://review.openstack.org/#/c/22926/ | 00:41 |
| matiu | it's supposed to change http 500 errs to 400 errs | 00:41 |
| *** sandywalsh has joined #openstack-dev | 00:41 | |
| *** zodiak has joined #openstack-dev | 00:42 | |
| *** jsindy has joined #openstack-dev | 00:42 | |
| *** Tross has joined #openstack-dev | 00:42 | |
| matiu | if anyone has any insight into that, a quick comment would be most apreciated | 00:42 |
| *** dolphm has joined #openstack-dev | 00:44 | |
| *** dhellmann has quit IRC | 00:46 | |
| *** anniec_ has quit IRC | 00:47 | |
| *** anniec has joined #openstack-dev | 00:48 | |
| *** pabelanger has joined #openstack-dev | 00:49 | |
| *** epim has quit IRC | 00:50 | |
| *** jkyle has joined #openstack-dev | 00:51 | |
| jkyle | reading the "setting up a development environment" is devstack what most do their work against? (as suggested in the doc) | 00:51 |
| matiu | jkyle, yeah, devstack is easiest | 00:52 |
| clarkb | jkyle: its what is used to do integration testing. The unittests for all projects can be run in isolation in their own virtualenvs | 00:52 |
| matiu | there's a vagrant devstack somewhere, that's apparently good for macs | 00:52 |
| *** Mandell has joined #openstack-dev | 00:53 | |
| jkyle | I have hardware and vms, if devstack is what's tested against. I'll roll with that. | 00:53 |
| jkyle | looks like learning the gerrit workflow will be more work than installing the dev environment hehe | 00:55 |
| clarkb | jkyle: if you pick a random change, eg https://review.openstack.org/#/c/22886/, you can see the tests that are run against it by jenkins in the comments | 00:55 |
| clarkb | jkyle: its not too bad, especially with git review | 00:56 |
| jkyle | clarkb: was more of a comment on how easy setting up a dev environment is ;) | 00:56 |
| clarkb | ah :) | 00:56 |
| *** mestery has quit IRC | 00:59 | |
| *** garyTh has quit IRC | 01:00 | |
| *** alexxu has joined #openstack-dev | 01:03 | |
| *** Tross has quit IRC | 01:04 | |
| *** spzala has quit IRC | 01:05 | |
| *** jkyle has quit IRC | 01:08 | |
| *** anteaya has left #openstack-dev | 01:10 | |
| *** bing_bu has joined #openstack-dev | 01:12 | |
| *** jsindy has quit IRC | 01:12 | |
| *** esp1 has quit IRC | 01:14 | |
| *** ilblackdragon has joined #openstack-dev | 01:17 | |
| dolphm | ayoung: i think we got out of sync on the impersonation attribute -- i saw it somewhere as "impersonation", so i ran with that, but you have "impersonate" in the impl? | 01:21 |
| dolphm | ayoung: i'd prefer something that has an antonym (e.g. enable/disable) -- but we at least need to straighten out impersonation / impersonate | 01:22 |
| *** sthaha has joined #openstack-dev | 01:24 | |
| openstackgerrit | A change was merged to openstack/nova: Fix incorrect zookeeper group name https://review.openstack.org/22917 | 01:25 |
| openstackgerrit | A change was merged to openstack-infra/devstack-gate: Add a missing space inside test brackets. https://review.openstack.org/22901 | 01:26 |
| openstackgerrit | A change was merged to openstack/nova: Change nova-dhcpbridge FLAGFILE to a list of files https://review.openstack.org/22746 | 01:26 |
| openstackgerrit | A change was merged to openstack/nova: Enable VM DHCP request to reach DHCP agent https://review.openstack.org/22546 | 01:26 |
| *** yidclare has quit IRC | 01:31 | |
| *** pabelanger has quit IRC | 01:31 | |
| *** mrodden has quit IRC | 01:36 | |
| *** cp16net is now known as cp16net|away | 01:37 | |
| *** cp16net|away is now known as cp16net | 01:37 | |
| *** dims has quit IRC | 01:40 | |
| *** jkyle has joined #openstack-dev | 01:44 | |
| *** Tross has joined #openstack-dev | 01:45 | |
| *** pabelanger has joined #openstack-dev | 01:46 | |
| *** sacharya has quit IRC | 01:46 | |
| openstackgerrit | A change was merged to openstack/quantum: Add default state_path to quantum.conf https://review.openstack.org/22849 | 01:47 |
| ayoung | dolphm, I just went through and made it all impersonation | 01:47 |
| dolphm | ayoung: oh okay | 01:48 |
| ayoung | dolphm, do_impersonation/don't_do_impersonation?\ | 01:48 |
| *** bdpayne has quit IRC | 01:48 | |
| ayoung | impersonate/don't_impersonate | 01:48 |
| dolphm | ayoung: i'm getting an unspecified validationerror in my test -- not sure why, see comment on 26 | 01:48 |
| ayoung | its awkward no matter what | 01:48 |
| ayoung | dolphm, 26? | 01:49 |
| dolphm | ayoung: patchset 26 | 01:49 |
| dolphm | ayoung: cloaked/uncloaked | 01:49 |
| *** mestery has joined #openstack-dev | 01:49 | |
| ayoung | dolphm, My guess is that it is the roles. What are you passing in for roles? | 01:50 |
| dolphm | ayoung: a list of objects with just id's | 01:50 |
| ayoung | I haven't changed to role {id} yet. THat is coming next | 01:50 |
| *** jkyle has quit IRC | 01:50 | |
| *** zeriouz has joined #openstack-dev | 01:50 | |
| ayoung | dolphm, OK...I'll tackle that next. I was just working through "expires" to "expires_at" and need to do a pep8 cleanup | 01:51 |
| dolphm | ayoung: cool | 01:52 |
| *** sacharya has joined #openstack-dev | 01:53 | |
| dolphm | ayoung: ping me when you have updates ... i'm fleshing out a test | 01:53 |
| dolphm | ayoung: autopersonate :P | 01:53 |
| *** bing_bu has quit IRC | 01:55 | |
| *** markmcclain has quit IRC | 01:56 | |
| *** amerine has quit IRC | 01:57 | |
| *** nati_ueno_ has quit IRC | 02:00 | |
| *** vkmc has quit IRC | 02:01 | |
| *** bing_bu has joined #openstack-dev | 02:01 | |
| ayoung | dolphm, what should I be using for time. I keep getting offsets | 02:02 |
| *** yaguang has joined #openstack-dev | 02:03 | |
| dolphm | ayoung: what does token expires use? | 02:03 |
| dolphm | ayoung: that should be utc | 02:03 |
| *** bknudson has quit IRC | 02:03 | |
| ayoung | timeutils.parse_isotime("1999-02-18T10:10:00Z") | 02:03 |
| ayoung | Gives me | 02:03 |
| ayoung | datetime: 1999-02-18 10:10:00+00:00 | 02:04 |
| *** otherwiseguy has quit IRC | 02:04 | |
| *** READ10 has quit IRC | 02:04 | |
| dolphm | ayoung: http://paste.openstack.org/raw/32493/ | 02:05 |
| ayoung | dolphm, the difference is that token does utc + time delta | 02:05 |
| dolphm | ayoung: i don't understand, utcnow() returns a naive utc object | 02:06 |
| ayoung | so it is all internally consistant, I think | 02:06 |
| *** salv-orlando has quit IRC | 02:06 | |
| ayoung | Maybe I should leave off the Z... lets see. | 02:06 |
| *** pixelbeat has quit IRC | 02:06 | |
| ayoung | Nope...no difference | 02:07 |
| ayoung | dolphm, there is something wonky even with the token code. | 02:07 |
| *** andrewbogott is now known as andrewbogott_afk | 02:07 | |
| *** zeriouz has quit IRC | 02:08 | |
| ayoung | dolphm, in the patch note what I did: https://review.openstack.org/#/c/20289/26/keystone/token/backends/kvs.py | 02:08 |
| ayoung | that was just getting a token and passing it back in, no trusts involved | 02:08 |
| *** andrewbogott_afk is now known as andrewbogott | 02:08 | |
| *** mrodden has joined #openstack-dev | 02:12 | |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: Improve error message for missing endpoint https://review.openstack.org/22744 | 02:15 |
| dolphm | ayoung: why would ref['expires'] ever be None? | 02:16 |
| *** andrewbogott is now known as andrewbogott_afk | 02:16 | |
| ayoung | dolphm, I'm not sure | 02:16 |
| *** ladquin has quit IRC | 02:16 | |
| ayoung | dolphm, that was from the original. | 02:17 |
| dolphm | ayoung: can you fix the todo instead of that patch? can test with something like assertNotIn('Z', expires) | 02:17 |
| ayoung | dolphm, maybe tokens didn't expire at one point and they left that in | 02:18 |
| *** pabelanger has quit IRC | 02:18 | |
| dolphm | ayoung: tokens have always expired :( that looks like a hack to get around a bug ... and makes me nervous | 02:18 |
| ayoung | dolphm, at a minimum, we can treat a token with no Expiry as expired. | 02:19 |
| *** kagan has quit IRC | 02:19 | |
| dolphm | ayoung: +1 | 02:21 |
| *** kgriffs has joined #openstack-dev | 02:21 | |
| *** Ryan_Lane has quit IRC | 02:22 | |
| *** enikanorov has quit IRC | 02:25 | |
| *** zeriouz has joined #openstack-dev | 02:29 | |
| *** cloudchimp has joined #openstack-dev | 02:31 | |
| *** AlanClark has quit IRC | 02:32 | |
| *** dolphm has quit IRC | 02:33 | |
| *** AlanClark has joined #openstack-dev | 02:33 | |
| *** arbrandes has quit IRC | 02:33 | |
| *** esp1 has joined #openstack-dev | 02:40 | |
| *** stevemar has joined #openstack-dev | 02:42 | |
| *** spzala has joined #openstack-dev | 02:43 | |
| fungi | i need to link https://bugs.launchpad.net/nova/+bug/1073569 to nova stable/essex... can't figure out the right dance in lp to make that happen (maybe i'm just not holding my teeth right?) | 02:47 |
| uvirtbot | Launchpad bug 1073569 in nova "Jenkins jobs fail because of incompatibility between sqlalchemy-migrate and the newest sqlalchemy-0.8.0b1" [Undecided,Fix committed] | 02:47 |
| *** adjohn has quit IRC | 02:48 | |
| *** esp1 has left #openstack-dev | 02:48 | |
| *** pabelanger has joined #openstack-dev | 02:49 | |
| *** jcmartin has quit IRC | 02:49 | |
| *** zeriouz has quit IRC | 02:52 | |
| *** koolhead17 has joined #openstack-dev | 02:52 | |
| *** markwash has joined #openstack-dev | 02:54 | |
| *** zeriouz has joined #openstack-dev | 02:55 | |
| *** anniec has quit IRC | 03:00 | |
| *** yaguang has quit IRC | 03:03 | |
| *** dims has joined #openstack-dev | 03:03 | |
| *** kgriffs has quit IRC | 03:04 | |
| *** kgriffs has joined #openstack-dev | 03:05 | |
| *** cp16net is now known as cp16net|away | 03:05 | |
| *** jamespage has quit IRC | 03:05 | |
| *** jamespage has joined #openstack-dev | 03:06 | |
| *** danwent has quit IRC | 03:06 | |
| *** koolhead17 has quit IRC | 03:06 | |
| *** novas0x2a|laptop has quit IRC | 03:07 | |
| *** dims has quit IRC | 03:08 | |
| *** cp16net|away is now known as cp16net | 03:08 | |
| *** kgriffs has quit IRC | 03:08 | |
| *** kgriffs has joined #openstack-dev | 03:09 | |
| *** dims has joined #openstack-dev | 03:10 | |
| *** sacharya1 has joined #openstack-dev | 03:10 | |
| *** kgriffs has quit IRC | 03:12 | |
| *** sacharya has quit IRC | 03:12 | |
| *** kgriffs has joined #openstack-dev | 03:13 | |
| *** roampune has quit IRC | 03:13 | |
| *** melwitt has quit IRC | 03:14 | |
| *** annegentle has joined #openstack-dev | 03:19 | |
| *** annegentle has quit IRC | 03:26 | |
| *** annegentle has joined #openstack-dev | 03:27 | |
| *** CaptTofu has joined #openstack-dev | 03:30 | |
| *** Tross has quit IRC | 03:30 | |
| *** tomoe_ has quit IRC | 03:31 | |
| *** tomoe_ has joined #openstack-dev | 03:31 | |
| *** tomoe_ has quit IRC | 03:36 | |
| *** annegentle has quit IRC | 03:37 | |
| *** esp has joined #openstack-dev | 03:40 | |
| *** AlanClark has quit IRC | 03:41 | |
| *** AlanClark has joined #openstack-dev | 03:41 | |
| *** CaptTofu has quit IRC | 03:43 | |
| *** kgriffs has quit IRC | 03:43 | |
| *** martine has joined #openstack-dev | 03:45 | |
| *** jcmartin has joined #openstack-dev | 03:48 | |
| *** esp has left #openstack-dev | 03:53 | |
| *** amerine has joined #openstack-dev | 03:55 | |
| *** adjohn has joined #openstack-dev | 03:58 | |
| *** jog0 has quit IRC | 04:01 | |
| *** koolhead17 has joined #openstack-dev | 04:02 | |
| mordred | gongysh: hey! I can't see your comment on the delete file on https://review.openstack.org/#/c/21350/21 | 04:03 |
| *** adjohn has quit IRC | 04:04 | |
| *** yaguang has joined #openstack-dev | 04:10 | |
| *** jsindy has joined #openstack-dev | 04:10 | |
| *** zzs has left #openstack-dev | 04:14 | |
| *** jsindy has quit IRC | 04:21 | |
| *** bdpayne has joined #openstack-dev | 04:22 | |
| *** topol has joined #openstack-dev | 04:23 | |
| *** networkstatic has joined #openstack-dev | 04:24 | |
| *** jimfehlig has quit IRC | 04:25 | |
| *** harlowja has quit IRC | 04:26 | |
| *** bdpayne has quit IRC | 04:28 | |
| *** Tross has joined #openstack-dev | 04:29 | |
| *** bdpayne has joined #openstack-dev | 04:30 | |
| *** Mandell has quit IRC | 04:30 | |
| *** esp has joined #openstack-dev | 04:32 | |
| *** esp has left #openstack-dev | 04:32 | |
| openstackgerrit | A change was merged to openstack-dev/devstack: Devstack support for nicira plugin https://review.openstack.org/21279 | 04:32 |
| *** zeriouz has quit IRC | 04:33 | |
| *** anniec has joined #openstack-dev | 04:35 | |
| *** jkyle has joined #openstack-dev | 04:36 | |
| *** osphy has quit IRC | 04:38 | |
| *** soody has quit IRC | 04:39 | |
| *** nunosantos has quit IRC | 04:42 | |
| *** zeriouz has joined #openstack-dev | 04:46 | |
| *** pabelanger has quit IRC | 04:49 | |
| *** markmcclain has joined #openstack-dev | 04:50 | |
| *** anniec has quit IRC | 04:51 | |
| *** vipul is now known as vipul|away | 04:54 | |
| *** danwent has joined #openstack-dev | 04:54 | |
| *** jkyle has quit IRC | 05:02 | |
| *** navid has joined #openstack-dev | 05:05 | |
| *** tomoe_ has joined #openstack-dev | 05:12 | |
| *** rushiagr has joined #openstack-dev | 05:12 | |
| *** stevemar has quit IRC | 05:13 | |
| rushiagr | hi folks | 05:13 |
| rushiagr | I currently see that in devstack you only can see info level logging | 05:13 |
| *** armaan has joined #openstack-dev | 05:13 | |
| rushiagr | any way to enable debug level messages? or maybe can someone give me the link to the log file (assuming that debug messages atleast goes into the log file)? | 05:14 |
| openstackgerrit | A change was merged to openstack/swift: Fix descriptions in sample configs. https://review.openstack.org/22906 | 05:14 |
| openstackgerrit | A change was merged to openstack/quantum: Fixes import reorder nits https://review.openstack.org/22829 | 05:14 |
| *** Ryan_Lane has joined #openstack-dev | 05:14 | |
| *** markwash has quit IRC | 05:16 | |
| *** nati_ueno has joined #openstack-dev | 05:17 | |
| *** nati_ueno has quit IRC | 05:18 | |
| *** nati_ueno has joined #openstack-dev | 05:18 | |
| *** aeperezt has quit IRC | 05:19 | |
| *** martine has quit IRC | 05:19 | |
| openstackgerrit | A change was merged to openstack/glance: Fix migration 016 for sqlite. https://review.openstack.org/22731 | 05:20 |
| openstackgerrit | A change was merged to openstack/glance: Refactor data migration tests https://review.openstack.org/21362 | 05:20 |
| *** burris has quit IRC | 05:23 | |
| *** Tross has quit IRC | 05:24 | |
| *** burris has joined #openstack-dev | 05:24 | |
| *** koolhead17 has quit IRC | 05:25 | |
| *** bdpayne has quit IRC | 05:26 | |
| *** Nachi has joined #openstack-dev | 05:31 | |
| *** buzztroll_ has quit IRC | 05:34 | |
| *** nati_ueno has quit IRC | 05:35 | |
| *** amerine has quit IRC | 05:35 | |
| *** crandquist has joined #openstack-dev | 05:41 | |
| *** amotoki has joined #openstack-dev | 05:42 | |
| *** enikanorov has joined #openstack-dev | 05:43 | |
| *** Mandell has joined #openstack-dev | 05:47 | |
| *** Tross has joined #openstack-dev | 05:47 | |
| *** Nachi has quit IRC | 05:50 | |
| *** nati_ueno has joined #openstack-dev | 05:51 | |
| *** Nachi has joined #openstack-dev | 05:54 | |
| *** AlanClark has quit IRC | 05:57 | |
| *** AlanClark has joined #openstack-dev | 05:58 | |
| *** nati_ueno has quit IRC | 05:58 | |
| *** sacharya1 has quit IRC | 06:00 | |
| *** almaisan-away is now known as al-maisan | 06:00 | |
| *** al-maisan is now known as almaisan-away | 06:02 | |
| *** garyk has quit IRC | 06:03 | |
| *** vipul|away is now known as vipul | 06:06 | |
| openstackgerrit | A change was merged to openstack/nova: Fix hacking.py to handle 'cannot import x' https://review.openstack.org/22947 | 06:07 |
| *** jcmartin has quit IRC | 06:09 | |
| *** hattwick has quit IRC | 06:19 | |
| *** Ryan_Lane has quit IRC | 06:20 | |
| *** vipul is now known as vipul|away | 06:21 | |
| openstackgerrit | A change was merged to openstack/swift: Adding speed limit options for DB auditor https://review.openstack.org/22286 | 06:22 |
| *** markwash has joined #openstack-dev | 06:23 | |
| *** adjohn has joined #openstack-dev | 06:25 | |
| *** topol has quit IRC | 06:25 | |
| *** markwash has quit IRC | 06:27 | |
| *** spzala has quit IRC | 06:30 | |
| *** vipul|away is now known as vipul | 06:37 | |
| *** garyk has joined #openstack-dev | 06:45 | |
| *** zaitcev has quit IRC | 06:45 | |
| *** mrodden1 has joined #openstack-dev | 06:46 | |
| *** mrodden has quit IRC | 06:48 | |
| *** dev_sa has joined #openstack-dev | 06:56 | |
| *** vipul is now known as vipul|away | 06:57 | |
| *** ritzcarltn has joined #openstack-dev | 07:01 | |
| *** ritzcarltn has quit IRC | 07:02 | |
| *** ritzcarltn has joined #openstack-dev | 07:03 | |
| *** gargya has joined #openstack-dev | 07:03 | |
| *** AlanClark has quit IRC | 07:05 | |
| *** AlanClark has joined #openstack-dev | 07:05 | |
| *** ritzcarltn has joined #openstack-dev | 07:08 | |
| *** techlife has quit IRC | 07:10 | |
| *** techlife has joined #openstack-dev | 07:11 | |
| *** gargya has quit IRC | 07:12 | |
| *** ritzcarltn has quit IRC | 07:13 | |
| *** ritzcarlt has joined #openstack-dev | 07:13 | |
| *** ritzcarltn has joined #openstack-dev | 07:14 | |
| *** morganfainberg has quit IRC | 07:15 | |
| *** amerine has joined #openstack-dev | 07:15 | |
| *** ritzcarltn has joined #openstack-dev | 07:17 | |
| *** ritzcarltn has joined #openstack-dev | 07:17 | |
| *** davidha has quit IRC | 07:18 | |
| *** zyluo has joined #openstack-dev | 07:18 | |
| zyluo | Does anybody know the link to the new CLA form? | 07:19 |
| *** yaguang has quit IRC | 07:20 | |
| *** avishay has joined #openstack-dev | 07:21 | |
| openstackgerrit | A change was merged to openstack/quantum: Enables packetfilter ext in NEC plugin based on its driver config https://review.openstack.org/22950 | 07:25 |
| openstackgerrit | A change was merged to openstack/quantum: Set default api_extensions_path for NEC plugin https://review.openstack.org/22949 | 07:27 |
| *** mindpixel has joined #openstack-dev | 07:29 | |
| *** ritzcarltn has quit IRC | 07:33 | |
| *** ritzcarltn has joined #openstack-dev | 07:37 | |
| *** adalbas has quit IRC | 07:39 | |
| *** mdomsch has quit IRC | 07:39 | |
| *** romcheg has quit IRC | 07:39 | |
| *** yuanz has quit IRC | 07:39 | |
| *** dubsquared has quit IRC | 07:39 | |
| *** ayoung has quit IRC | 07:39 | |
| *** sulrich has quit IRC | 07:39 | |
| *** locke105 has quit IRC | 07:39 | |
| *** dwilson has quit IRC | 07:39 | |
| *** maurosr has quit IRC | 07:39 | |
| *** adam_g has quit IRC | 07:39 | |
| *** ahale has quit IRC | 07:39 | |
| *** mrodden1 has quit IRC | 07:40 | |
| *** ritzcarltn has quit IRC | 07:40 | |
| *** yuanz has joined #openstack-dev | 07:40 | |
| *** ritzcarltn has joined #openstack-dev | 07:40 | |
| *** ahale has joined #openstack-dev | 07:41 | |
| *** dubsquared has joined #openstack-dev | 07:41 | |
| *** yaguang has joined #openstack-dev | 07:42 | |
| *** mrodden has joined #openstack-dev | 07:42 | |
| *** ayoung has joined #openstack-dev | 07:42 | |
| *** mdomsch has joined #openstack-dev | 07:43 | |
| *** dwilson has joined #openstack-dev | 07:43 | |
| *** blamar_ has joined #openstack-dev | 07:44 | |
| *** mrunge has joined #openstack-dev | 07:44 | |
| *** blamar has quit IRC | 07:45 | |
| *** blamar_ is now known as blamar | 07:45 | |
| *** henrynash has joined #openstack-dev | 07:47 | |
| *** flaper87 has joined #openstack-dev | 07:48 | |
| *** adalbas has joined #openstack-dev | 07:48 | |
| *** romcheg has joined #openstack-dev | 07:51 | |
| *** sulrich has joined #openstack-dev | 07:51 | |
| *** locke105 has joined #openstack-dev | 07:51 | |
| *** maurosr has joined #openstack-dev | 07:51 | |
| *** adam_g has joined #openstack-dev | 07:51 | |
| gongysh | mordred: pong | 07:52 |
| *** gargya has joined #openstack-dev | 07:55 | |
| *** reidrac has joined #openstack-dev | 07:55 | |
| *** thouveng has joined #openstack-dev | 07:55 | |
| *** markmcclain has quit IRC | 07:56 | |
| mordred | gongysh: I couldn't see your comment on the deleted file | 07:56 |
| mordred | gongysh: gerrit is not showing it to me | 07:57 |
| gongysh | I have replied in the review msg. | 07:57 |
| mordred | oh, great! | 07:57 |
| *** buzztroll_ has joined #openstack-dev | 07:57 | |
| gongysh | mordred: I cannot see it myself. I just see there is one comment. :) | 07:58 |
| *** rafaduran has joined #openstack-dev | 07:58 | |
| mordred | gongysh: k. replied myself- tl;dr - it was a test of code copied from oslo | 07:58 |
| mordred | which is tested in oslo | 07:58 |
| *** yolanda has joined #openstack-dev | 08:01 | |
| gongysh | mordred: ok, +2 from me. | 08:01 |
| mordred | gongysh: yippee!! (this set of patches is killing me) :) | 08:01 |
| gongysh | mordred: but you are alive, right? :) | 08:01 |
| mordred | gongysh: barely :) | 08:02 |
| gongysh | mordred: that is a good patch we need, hope with testr, we will reduce case running time a lot. | 08:02 |
| mordred | gongysh: me too. I still have a few more patches before we can do parallel -but they're getting closer | 08:03 |
| *** eglynn has joined #openstack-dev | 08:05 | |
| *** zyluo has quit IRC | 08:07 | |
| *** jbr_zzz has left #openstack-dev | 08:10 | |
| *** zoresvit has joined #openstack-dev | 08:11 | |
| *** zoresvit has quit IRC | 08:15 | |
| *** davidha has joined #openstack-dev | 08:16 | |
| *** ondergetekende has joined #openstack-dev | 08:21 | |
| *** mmagr has joined #openstack-dev | 08:23 | |
| *** flepied has joined #openstack-dev | 08:26 | |
| *** Mandell has quit IRC | 08:28 | |
| *** yaguang has quit IRC | 08:29 | |
| *** BobBall has quit IRC | 08:30 | |
| *** BobBall has joined #openstack-dev | 08:30 | |
| *** BobBall has quit IRC | 08:32 | |
| *** zoresvit has joined #openstack-dev | 08:32 | |
| *** BobBall has joined #openstack-dev | 08:32 | |
| *** afazekas has joined #openstack-dev | 08:36 | |
| *** jgallard has joined #openstack-dev | 08:39 | |
| *** networkstatic has quit IRC | 08:40 | |
| *** almaisan-away is now known as al-maisan | 08:41 | |
| *** BobBall has quit IRC | 08:42 | |
| *** BobBall_ has joined #openstack-dev | 08:42 | |
| *** pixelbeat has joined #openstack-dev | 08:42 | |
| *** adjohn has quit IRC | 08:46 | |
| *** zing has joined #openstack-dev | 08:46 | |
| openstackgerrit | A change was merged to openstack/nova: if reset fails, display the command that failed https://review.openstack.org/22888 | 08:47 |
| *** yaguang has joined #openstack-dev | 08:52 | |
| *** henrynash has quit IRC | 08:54 | |
| *** winston-d_ has quit IRC | 08:55 | |
| *** salv-orlando has joined #openstack-dev | 08:56 | |
| *** zoresvit has quit IRC | 08:58 | |
| *** gael has joined #openstack-dev | 08:59 | |
| *** wwallnrr__ has quit IRC | 08:59 | |
| *** jpich has joined #openstack-dev | 09:00 | |
| *** zoresvit has joined #openstack-dev | 09:01 | |
| *** zoresvit has quit IRC | 09:05 | |
| *** corXi has joined #openstack-dev | 09:06 | |
| *** boris-42 has joined #openstack-dev | 09:06 | |
| *** trapni has joined #openstack-dev | 09:09 | |
| *** trapni has joined #openstack-dev | 09:09 | |
| *** zeriouz has quit IRC | 09:11 | |
| *** derekh has joined #openstack-dev | 09:14 | |
| *** trapnii has joined #openstack-dev | 09:15 | |
| *** trapni has quit IRC | 09:16 | |
| *** adjohn has joined #openstack-dev | 09:17 | |
| *** markmc has joined #openstack-dev | 09:17 | |
| *** zoresvit has joined #openstack-dev | 09:19 | |
| *** AlanClark has quit IRC | 09:22 | |
| *** AlanClark has joined #openstack-dev | 09:23 | |
| *** johnthetubaguy1 has joined #openstack-dev | 09:23 | |
| *** darraghb has joined #openstack-dev | 09:24 | |
| *** adjohn has quit IRC | 09:26 | |
| *** psedlak has joined #openstack-dev | 09:31 | |
| *** yaguang has quit IRC | 09:31 | |
| *** maroh has joined #openstack-dev | 09:35 | |
| *** trapniii has joined #openstack-dev | 09:41 | |
| *** johnthetubaguy has quit IRC | 09:42 | |
| *** bing_bu has quit IRC | 09:43 | |
| *** bing_bu has joined #openstack-dev | 09:44 | |
| *** trapnii has quit IRC | 09:45 | |
| *** bing_bu has quit IRC | 09:45 | |
| *** henrynash has joined #openstack-dev | 09:46 | |
| *** danpb has joined #openstack-dev | 09:48 | |
| *** Nachi has quit IRC | 09:52 | |
| *** adjohn has joined #openstack-dev | 09:52 | |
| *** yaguang has joined #openstack-dev | 09:53 | |
| *** iartarisi has joined #openstack-dev | 09:56 | |
| *** adjohn has quit IRC | 09:58 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Add glossary definitions for additional terms https://review.openstack.org/22651 | 10:00 |
| maroh | garyk : hi | 10:01 |
| garyk | maroh: hi | 10:01 |
| maroh | gary, i'd like to test multi-agent patch | 10:02 |
| openstackgerrit | A change was merged to openstack/ceilometer: Remove nose wrapper script https://review.openstack.org/22891 | 10:02 |
| maroh | garyk, do you already filled the wiki page you spoked about yesterday? | 10:02 |
| *** trapniii has quit IRC | 10:02 | |
| garyk | maroh: not yet. on my to do list | 10:03 |
| garyk | maroh: you having problems with ovs and devstack? | 10:03 |
| maroh | i didn't try yet | 10:03 |
| maroh | garyk, so let's try first ;-) | 10:05 |
| *** jbr_ has joined #openstack-dev | 10:09 | |
| *** trapniii has joined #openstack-dev | 10:16 | |
| *** aloga is now known as ojete | 10:16 | |
| *** arosen has quit IRC | 10:16 | |
| *** ojete is now known as aloga | 10:17 | |
| *** arosen has joined #openstack-dev | 10:20 | |
| *** yaguang has quit IRC | 10:20 | |
| *** mindpixel has quit IRC | 10:21 | |
| *** zoresvit has quit IRC | 10:26 | |
| *** armaan has quit IRC | 10:29 | |
| *** zoresvit has joined #openstack-dev | 10:32 | |
| *** AlanClark has quit IRC | 10:34 | |
| *** hattwick has joined #openstack-dev | 10:35 | |
| *** masumotok has joined #openstack-dev | 10:37 | |
| *** johnthetubaguy1 has quit IRC | 10:41 | |
| *** armaan has joined #openstack-dev | 10:45 | |
| *** johnthetubaguy has joined #openstack-dev | 10:47 | |
| *** yamahata has joined #openstack-dev | 10:48 | |
| *** vkmc has joined #openstack-dev | 10:53 | |
| *** amerine has quit IRC | 10:57 | |
| *** danwent has quit IRC | 10:58 | |
| *** ritzcarltn has quit IRC | 10:59 | |
| *** Ritz has joined #openstack-dev | 11:00 | |
| *** mkollaro has joined #openstack-dev | 11:02 | |
| *** jgallard has quit IRC | 11:03 | |
| *** jgallard has joined #openstack-dev | 11:05 | |
| *** pcm_ has joined #openstack-dev | 11:10 | |
| openstackgerrit | A change was merged to openstack/quantum: Use testtools instead of unittest or unittest2. https://review.openstack.org/21350 | 11:31 |
| *** larsbutler has joined #openstack-dev | 11:32 | |
| *** yaguang has joined #openstack-dev | 11:33 | |
| *** larsbutler has quit IRC | 11:34 | |
| *** larsbutler_ has joined #openstack-dev | 11:34 | |
| *** tomoe_ has quit IRC | 11:36 | |
| *** larsbutler_ is now known as larsbutler | 11:37 | |
| *** davidh_ has joined #openstack-dev | 12:04 | |
| *** mohits has joined #openstack-dev | 12:05 | |
| *** davidha has quit IRC | 12:06 | |
| *** rkukura has quit IRC | 12:08 | |
| *** mkollaro has quit IRC | 12:11 | |
| *** trapniii has quit IRC | 12:14 | |
| *** trapni has joined #openstack-dev | 12:16 | |
| *** trapni has joined #openstack-dev | 12:16 | |
| *** adjohn has joined #openstack-dev | 12:24 | |
| *** mohits has quit IRC | 12:25 | |
| *** ilblackd_ has joined #openstack-dev | 12:27 | |
| *** adjohn has quit IRC | 12:29 | |
| *** ilblackdragon has quit IRC | 12:30 | |
| *** gongysh has quit IRC | 12:36 | |
| *** zoresvit has quit IRC | 12:38 | |
| *** gongysh has joined #openstack-dev | 12:39 | |
| *** jbresnah has joined #openstack-dev | 12:39 | |
| *** buzztroll_ has quit IRC | 12:40 | |
| *** krtaylor has quit IRC | 12:41 | |
| *** yaguang has quit IRC | 12:42 | |
| *** espen__ has quit IRC | 12:49 | |
| *** navid has left #openstack-dev | 12:52 | |
| *** mohits has joined #openstack-dev | 12:54 | |
| *** jgallard has quit IRC | 12:54 | |
| *** zoresvit has joined #openstack-dev | 12:56 | |
| *** kashyap has quit IRC | 12:56 | |
| *** sandywalsh has quit IRC | 12:56 | |
| *** krtaylor has joined #openstack-dev | 12:57 | |
| *** navid has joined #openstack-dev | 12:57 | |
| *** martine has joined #openstack-dev | 13:00 | |
| *** zoresvit has quit IRC | 13:01 | |
| *** tomoe_ has joined #openstack-dev | 13:01 | |
| *** wowdd1 has joined #openstack-dev | 13:01 | |
| *** soody has joined #openstack-dev | 13:03 | |
| *** amotoki has quit IRC | 13:04 | |
| *** mindpixel has joined #openstack-dev | 13:06 | |
| *** olaph has quit IRC | 13:08 | |
| *** sandywalsh has joined #openstack-dev | 13:09 | |
| *** olaph has joined #openstack-dev | 13:10 | |
| *** soody has quit IRC | 13:11 | |
| *** alexxu has quit IRC | 13:12 | |
| *** jbresnah_ has joined #openstack-dev | 13:12 | |
| *** jbresnah has quit IRC | 13:13 | |
| *** Yada has joined #openstack-dev | 13:13 | |
| *** markvoelker has joined #openstack-dev | 13:14 | |
| *** zoresvit has joined #openstack-dev | 13:15 | |
| *** Saurabh has joined #openstack-dev | 13:17 | |
| Saurabh | need help with debugging dhcp-agent. Could anyone help me out with using pdb for dhcp agent? | 13:18 |
| Saurabh | when I try strace I get this error AttributeError: 'NoneType' object has no attribute 'rpartition' | 13:21 |
| *** adjohn has joined #openstack-dev | 13:24 | |
| *** alunduil has quit IRC | 13:27 | |
| *** adjohn has quit IRC | 13:29 | |
| *** digitalsanctum has joined #openstack-dev | 13:29 | |
| *** arbrandes has joined #openstack-dev | 13:34 | |
| *** dolphm has joined #openstack-dev | 13:36 | |
| *** imsplitbit has joined #openstack-dev | 13:37 | |
| *** anteaya has joined #openstack-dev | 13:39 | |
| *** armaan has quit IRC | 13:40 | |
| ayoung | henrynash, can you take a look at https://review.openstack.org/#/c/22889/? It is a prereq for Trusts. | 13:43 |
| *** gongysh has quit IRC | 13:43 | |
| *** anniec has joined #openstack-dev | 13:52 | |
| *** doude has joined #openstack-dev | 13:52 | |
| aloga | mikal: ping | 13:52 |
| *** edmund has joined #openstack-dev | 13:55 | |
| *** gongysh has joined #openstack-dev | 13:56 | |
| henrynash | ayoung: made one commend - minor…but probably worth correcting…then will +2 | 13:57 |
| ayoung | you are correct. | 13:57 |
| ayoung | Will fix. | 13:57 |
| henrynash | ayoung: ok | 13:57 |
| *** mohits has quit IRC | 14:00 | |
| *** dprince has joined #openstack-dev | 14:00 | |
| dolphm | ayoung: docstr is wrong too | 14:00 |
| dolphm | ayoung: or at least, it's not at all indicative of what's actually being asserted | 14:01 |
| ayoung | dolphm, ok | 14:01 |
| henrynash | dolphm: made comments to your concerns on https://review.openstack.org/#/c/22789/ - let me know what you think | 14:02 |
| *** kbringard has joined #openstack-dev | 14:03 | |
| *** johnthetubaguy has quit IRC | 14:03 | |
| ayoung | dolphm, are we close? | 14:04 |
| *** johnthetubaguy has joined #openstack-dev | 14:04 | |
| dolphm | ayoung: see comments on trusts | 14:04 |
| ayoung | dolphm, yeah, those are both easy fixes (I think) . So nothing major? | 14:05 |
| dolphm | ayoung: don't know yet -- haven't finished writing the test i have in my head because of those failures | 14:06 |
| dolphm | ayoung: +2 when my test passes :) will simply be exercising every new call in the spec, with and without impersonation | 14:07 |
| *** yaguang has joined #openstack-dev | 14:10 | |
| ayoung | dolphm, so, On the roles, right now I have it implemented with role_id only. In order to do anything more, I would have to do additional sql from the controller, and I was too tired/stupid to do that last night without messing it up. | 14:10 |
| *** annegentle has joined #openstack-dev | 14:10 | |
| dolphm | ayoung: we'll get it today | 14:11 |
| *** topol has joined #openstack-dev | 14:12 | |
| *** ladquin has joined #openstack-dev | 14:15 | |
| *** anniec has quit IRC | 14:16 | |
| *** anniec has joined #openstack-dev | 14:17 | |
| ayoung | dolphm, so if I get a list with role names in them, to translate them to role Ids, I need to query the roles. Do you think it is a safer assumption that the list of roles will be short and we should query the whole list once and then hold it in memory, or do a look up for each individual role? I think the whole list. | 14:18 |
| dolphm | ayoung: not sure what you mean? cache the list of all roles in memory? | 14:19 |
| ayoung | dolphm, yeah. | 14:19 |
| ayoung | dolphm, not across calls | 14:19 |
| ayoung | just for a single create_trust call | 14:19 |
| dolphm | ayoung: ah, i bet it would be short | 14:19 |
| openstackgerrit | A change was merged to openstack/nova: Minor code optimization in _compute_topic https://review.openstack.org/22688 | 14:19 |
| ayoung | dolphm, that's my take as well.. | 14:19 |
| dolphm | ayoung: call list_roles rather than get_role_by_name several times? | 14:20 |
| ayoung | correct | 14:20 |
| dolphm | that works | 14:20 |
| *** ondergetekende has quit IRC | 14:20 | |
| *** ondergetekende has joined #openstack-dev | 14:20 | |
| *** jergerber has joined #openstack-dev | 14:20 | |
| ayoung | henrynash, you are referring to : TODO(henry-nash) do we need to put the whole object | 14:22 |
| ayoung | I can take that out now, right? | 14:22 |
| henrynash | ayoung: yep…you solved that issue by flattening the object! | 14:22 |
| *** doude has quit IRC | 14:22 | |
| *** jaypipes has quit IRC | 14:23 | |
| *** zoresvit has quit IRC | 14:24 | |
| *** kgriffs has joined #openstack-dev | 14:24 | |
| *** rkukura has joined #openstack-dev | 14:29 | |
| *** bknudson has joined #openstack-dev | 14:29 | |
| *** mindpixel has quit IRC | 14:30 | |
| dolphm | ayoung: do i have to request the same project scope as defined in the trust, or does the trust imply the scope automatically? | 14:31 |
| *** stevemar has joined #openstack-dev | 14:32 | |
| ayoung | dolphm, when asking for the token? Everything is defined by the trust | 14:34 |
| ayoung | you should not have to specify scoped other than the trust | 14:34 |
| dolphm | ayoung: k | 14:35 |
| topol | dolphm, what is the invisible_to_admin project used for? Is this documented somewhere? | 14:35 |
| ayoung | dolphm, can I push https://review.openstack.org/#/c/22889/10 | 14:35 |
| dolphm | topol: i've always wondered what that was, i have no idea | 14:35 |
| ayoung | topol, I think it is just a demonstration of the capability done by devstack. It isn't something meant to be there in a live deploy | 14:36 |
| dolphm | ayoung: done | 14:36 |
| ayoung | whew | 14:36 |
| *** READ10 has joined #openstack-dev | 14:37 | |
| devananda | annegentle: ping | 14:38 |
| annegentle | devananda: heyo | 14:38 |
| *** YorikSar has quit IRC | 14:38 | |
| *** mohits has joined #openstack-dev | 14:39 | |
| devananda | annegentle: hi! have a few minutes for nova doc questions? | 14:39 |
| *** Q310 has joined #openstack-dev | 14:40 | |
| *** Qten has quit IRC | 14:40 | |
| annegentle | devananda: sure | 14:40 |
| devananda | annegentle: more specifically, i need to write all the docs for baremetal driver. can you point me to resources/examples/etc of where to put them, formatting, etc? | 14:41 |
| *** zoresvit has joined #openstack-dev | 14:41 | |
| *** jgallard has joined #openstack-dev | 14:42 | |
| annegentle | devananda: the audience you're writing for is operators who want to do this, right? | 14:42 |
| devananda | annegentle: yep | 14:43 |
| annegentle | devananda: You'll want to add it to the Compute Admin manual | 14:43 |
| annegentle | devananda: this page has a ton of info https://wiki.openstack.org/wiki/Documentation/HowTo | 14:43 |
| annegentle | devananda: this week I'm otherwise occupied with the book sprint, but you can ask any questions in #openstack-doc | 14:43 |
| *** CaptTofu has joined #openstack-dev | 14:43 | |
| annegentle | devananda: I'd say, put it in a new section in this chapter http://docs.openstack.org/folsom/openstack-compute/admin/content/ch_openstack-compute-automated-installations.html | 14:44 |
| annegentle | devananda: the source file is https://github.com/openstack/openstack-manuals/blob/master/doc/src/docbkx/openstack-compute-admin/computeautomation.xml | 14:45 |
| annegentle | devananda: editor options are described here: https://wiki.openstack.org/wiki/Documentation/HowTo#Editor_Options | 14:46 |
| *** primeministerp has joined #openstack-dev | 14:46 | |
| devananda | annegentle: hm, so there's the general "how to do a baremetal deploy" -- what' the environment, what should/shouldn't be runnig, etc -- and then there's "how to deploy openstack using openstack" | 14:47 |
| annegentle | devananda: that's probably all you can intake for now. Another option is to write it all in asciidoc or markdown or RST and we'll get it into docbook | 14:47 |
| devananda | annegentle: automation is more the second. right now, i'm aiming to document the former first ;) | 14:47 |
| annegentle | devananda: ah, okay | 14:47 |
| annegentle | devananda: so perhaps you're better off with a new standalone book | 14:48 |
| devananda | it's a bit of a different beast than just using nova-compute, since there are specific network bits for pxe booting, etc, that need to be explained | 14:48 |
| annegentle | devananda: the compute admin guide is getting bloated anyway | 14:48 |
| *** nunosantos has joined #openstack-dev | 14:49 | |
| devananda | annegentle: new book seems fine. | 14:49 |
| devananda | annegentle: i'll dig into those links and see how far i get. thanks! | 14:49 |
| *** nunosantos has quit IRC | 14:50 | |
| *** nunosantos has joined #openstack-dev | 14:51 | |
| *** davidh_ has quit IRC | 14:51 | |
| *** davidha has joined #openstack-dev | 14:52 | |
| *** mtreinish has joined #openstack-dev | 14:52 | |
| annegentle | devananda: you might like the HA guide, https://github.com/openstack/openstack-manuals/tree/master/doc/src/docbkx/openstack-ha | 14:52 |
| annegentle | devananda: as a model | 14:52 |
| ayoung | dolphm, I might just be getting the hang of this python thing. http://www.fpaste.org/fcxp/ | 14:55 |
| *** adjohn has joined #openstack-dev | 14:55 | |
| dolphm | ayoung: pep8 line 2 | 14:55 |
| ayoung | dolphm, heh | 14:55 |
| dolphm | ayoung: raise exception.ValidationError(attribute='id or name', target='roles') | 14:56 |
| dolphm | ayoung: couple more pep8 violations | 14:56 |
| ayoung | dolphm, I'll get them before posting | 14:57 |
| *** radez_g0n3 is now known as radez | 14:57 | |
| ayoung | dolphm, so on validating the trust dictionary on create, would it be | 14:57 |
| ayoung | except KeyError as e: | 14:58 |
| ayoung | raise exception.ValidationError(attribute=e.args[0], | 14:58 |
| ayoung | target='trust') | 14:58 |
| *** zykes- has quit IRC | 14:58 | |
| dolphm | ayoung: i have no idea what KeyError args look like? | 14:59 |
| ayoung | dolphm, it is a tuple | 14:59 |
| dolphm | ayoung: of? | 14:59 |
| ayoung | dolphm, the key names. | 14:59 |
| dolphm | ayoung: how would there be more than one key name? | 14:59 |
| ayoung | dolphm, I think it gets args from the base class. That is just where it stores them | 15:00 |
| *** adjohn has quit IRC | 15:00 | |
| ayoung | dolphm, but I've confirmed that is where they go. I've a test for it now. | 15:00 |
| dolphm | ayoung: cool | 15:00 |
| dolphm | the signature looks fine otherwise | 15:00 |
| *** YorikSar has joined #openstack-dev | 15:01 | |
| *** nunosantos has quit IRC | 15:02 | |
| *** ijw1 has joined #openstack-dev | 15:04 | |
| *** aeperezt has joined #openstack-dev | 15:04 | |
| *** zykes- has joined #openstack-dev | 15:04 | |
| *** dolphm has quit IRC | 15:04 | |
| *** jimfehlig has joined #openstack-dev | 15:05 | |
| *** ndipanov has quit IRC | 15:05 | |
| *** ijw has quit IRC | 15:06 | |
| *** YorikSar has quit IRC | 15:06 | |
| *** ndipanov has joined #openstack-dev | 15:06 | |
| *** YorikSar has joined #openstack-dev | 15:07 | |
| *** CaptTofu has quit IRC | 15:08 | |
| *** CaptTofu has joined #openstack-dev | 15:09 | |
| *** al-maisan is now known as almaisan-away | 15:11 | |
| *** primeministerp has quit IRC | 15:11 | |
| *** sthaha has quit IRC | 15:11 | |
| *** rnirmal has joined #openstack-dev | 15:12 | |
| *** zing has quit IRC | 15:13 | |
| *** blamar has quit IRC | 15:13 | |
| *** zing has joined #openstack-dev | 15:13 | |
| *** blamar has joined #openstack-dev | 15:15 | |
| *** john5223 has joined #openstack-dev | 15:15 | |
| *** dolphm has joined #openstack-dev | 15:16 | |
| *** psedlak has quit IRC | 15:16 | |
| *** pabelanger has joined #openstack-dev | 15:16 | |
| *** riskable has joined #openstack-dev | 15:16 | |
| *** mindpixel has joined #openstack-dev | 15:18 | |
| *** cloudchimp has joined #openstack-dev | 15:18 | |
| openstackgerrit | A change was merged to openstack/keystone: flatten payload for policy https://review.openstack.org/22889 | 15:20 |
| *** CaptTofu has quit IRC | 15:21 | |
| *** afazekas has quit IRC | 15:21 | |
| *** CaptTofu has joined #openstack-dev | 15:23 | |
| *** AlanClark has joined #openstack-dev | 15:23 | |
| *** eharney has joined #openstack-dev | 15:24 | |
| *** adjohn has joined #openstack-dev | 15:26 | |
| *** annegentle has quit IRC | 15:26 | |
| *** adjohn has quit IRC | 15:27 | |
| *** adjohn has joined #openstack-dev | 15:27 | |
| *** gordc has joined #openstack-dev | 15:29 | |
| *** NobodyCam has quit IRC | 15:31 | |
| *** jsindy has joined #openstack-dev | 15:31 | |
| openstackgerrit | A change was merged to openstack/quantum: Make sure all db accesses use subtransaction https://review.openstack.org/22979 | 15:32 |
| *** CaptTofu_ has joined #openstack-dev | 15:32 | |
| *** garyTh has joined #openstack-dev | 15:32 | |
| *** CaptTofu has quit IRC | 15:32 | |
| *** alunduil has joined #openstack-dev | 15:32 | |
| ayoung | dolphm, just updated the trusts patch with a cleaned up version of that change. It had a few side effects, I think maybe flushing out some bugs. Let me know now if you can't get your test working and I'll work through it with you. | 15:33 |
| dolphm | ayoung: cool | 15:33 |
| *** alunduil has quit IRC | 15:33 | |
| *** alunduil has joined #openstack-dev | 15:34 | |
| *** epim has joined #openstack-dev | 15:35 | |
| *** blamar has quit IRC | 15:38 | |
| *** sacharya has joined #openstack-dev | 15:38 | |
| *** opilotte has quit IRC | 15:38 | |
| dolphm | ayoung: I got Expecting to find id in trust while trying to POST /trusts | 15:38 |
| *** otherwiseguy has joined #openstack-dev | 15:38 | |
| *** blamar has joined #openstack-dev | 15:38 | |
| *** digitalsanctum has quit IRC | 15:38 | |
| *** Gordonz has joined #openstack-dev | 15:39 | |
| *** Gordonz has quit IRC | 15:39 | |
| topol | dolphm, do we provide more documentation on the role migration beyond what is found in keystone.conf.sample? | 15:39 |
| topol | # Role for migrating membership relationships | 15:39 |
| topol | # During a SQL upgrade, the following values will be used to create a new role | 15:39 |
| topol | # that will replace records in the user_tenant_membership table with explicit | 15:39 |
| topol | # role grants. After migration, the member_role_id will be used in the API | 15:39 |
| topol | # add_user_to_project, and member_role_name will be ignored. | 15:39 |
| *** digitalsanctum has joined #openstack-dev | 15:40 | |
| *** Gordonz has joined #openstack-dev | 15:40 | |
| dolphm | ayoung: ^ | 15:40 |
| dolphm | topol: i'm not aware of any other docs atm | 15:40 |
| *** jkordish_ has joined #openstack-dev | 15:41 | |
| *** jkordish has quit IRC | 15:42 | |
| topol | dolphm, are we planning to cover this in more detail ? | 15:42 |
| *** gongysh has quit IRC | 15:42 | |
| dolphm | topol: i think we need a full folsom->grizzly migration guide | 15:42 |
| topol | dolphm, +1000 :-) | 15:43 |
| *** john5223 has quit IRC | 15:43 | |
| dolphm | topol: are you familiar with grenade? | 15:44 |
| topol | dolphm, the ones people throw at me when Im at work? | 15:44 |
| dolphm | topol: yep :) https://github.com/nebula/grenade | 15:44 |
| *** navid has left #openstack-dev | 15:45 | |
| dolphm | topol: there's already a workaround for a migration bug we have | 15:45 |
| *** zeriouz has joined #openstack-dev | 15:45 | |
| *** jimfehlig has quit IRC | 15:46 | |
| bknudson | anybody else seen this, from run_tests.sh --update: No distributions at all found for oslo-config from http://tarballs.openstack.org/oslo-config/oslo-config-2013.1b4.tar.gz ... | 15:46 |
| *** zaneb has joined #openstack-dev | 15:46 | |
| *** gargya has quit IRC | 15:46 | |
| topol | dolphm, so no Im not familiar with grenade. Whats its purpose? | 15:46 |
| dolphm | topol: primary goal is to test the upgrade process, but at the same time it happens to produce a decent starting point for human-readable upgrade docs :) | 15:47 |
| dolphm | dtroyer: /salute | 15:47 |
| topol | dolphm, OK. good to know | 15:48 |
| *** jsindy has quit IRC | 15:48 | |
| *** jimfehlig has joined #openstack-dev | 15:49 | |
| *** zbitter has quit IRC | 15:49 | |
| *** CaptTofu_ has quit IRC | 15:50 | |
| aloga | dolphm: awesome | 15:50 |
| aloga | (the grenade stuff) | 15:50 |
| dolphm | bknudson: i'm pretty sure there's a fix floating around for that -- i got around it by pip installing it and passing -N to run_tests.sh | 15:53 |
| *** bswartz has quit IRC | 15:53 | |
| dolphm | bknudson: (i.e. setting up my own venv) | 15:53 |
| bknudson | dolphm: I'll give it a shot. | 15:53 |
| *** mrodden has quit IRC | 15:54 | |
| *** gargya has joined #openstack-dev | 15:55 | |
| *** dhellmann has joined #openstack-dev | 15:55 | |
| *** john5223 has joined #openstack-dev | 15:55 | |
| *** dhellmann has quit IRC | 15:56 | |
| *** dhellmann has joined #openstack-dev | 15:57 | |
| *** edmund has quit IRC | 15:57 | |
| *** koolhead17 has joined #openstack-dev | 15:57 | |
| *** sacharya has quit IRC | 15:57 | |
| ayoung | dolphm, hmmm... | 15:57 |
| *** rushiagr has quit IRC | 15:58 | |
| ayoung | dolphm, which backend? | 15:58 |
| dolphm | ayoung: sql i guess, but i'm assuming that error is at the controller layer? | 15:59 |
| dolphm | ayoung: do you raise ValidationError anywhere else? | 15:59 |
| ayoung | dolphm, no. But I also don't check for 'id' | 15:59 |
| dolphm | ayoung: i'm guessing your e.args[0] is actually referring to some other 'id', but you're catch all is hardcoded to mention 'trusts' | 16:00 |
| dolphm | your* | 16:00 |
| ayoung | dolphm, the only collection inside there is the roles, and that also does a validation error | 16:00 |
| dolphm | ayoung: run the request i pasted | 16:01 |
| *** reidrac has quit IRC | 16:01 | |
| ayoung | dolphm, I realize I had a start of the role code here as well: | 16:01 |
| ayoung | _fill_in_roles | 16:02 |
| openstackgerrit | A change was merged to openstack/nova: Allow zk driver be imported without zookeeper https://review.openstack.org/22918 | 16:04 |
| *** pnavarro has joined #openstack-dev | 16:04 | |
| *** datsun180b has joined #openstack-dev | 16:04 | |
| *** ijw1 is now known as ijw | 16:08 | |
| *** burris has quit IRC | 16:08 | |
| *** burris has joined #openstack-dev | 16:08 | |
| *** yaguang has quit IRC | 16:08 | |
| *** devoid has joined #openstack-dev | 16:09 | |
| *** mrodden has joined #openstack-dev | 16:09 | |
| *** CaptTofu has joined #openstack-dev | 16:10 | |
| *** markmcclain has joined #openstack-dev | 16:11 | |
| *** mrodden1 has joined #openstack-dev | 16:12 | |
| *** sstent_ has joined #openstack-dev | 16:12 | |
| *** networkstatic has joined #openstack-dev | 16:12 | |
| *** markmcclain has quit IRC | 16:13 | |
| *** sstent_ has joined #openstack-dev | 16:13 | |
| *** mrodden has quit IRC | 16:13 | |
| zul | mordred: ping | 16:13 |
| *** sstent_ has quit IRC | 16:14 | |
| *** sstent_ has joined #openstack-dev | 16:14 | |
| *** jsindy has joined #openstack-dev | 16:14 | |
| *** john5223 has quit IRC | 16:15 | |
| *** anniec has quit IRC | 16:15 | |
| ayoung | dolphm, OK, reproduced | 16:16 |
| *** markmcclain has joined #openstack-dev | 16:16 | |
| *** ociuhandu has joined #openstack-dev | 16:17 | |
| *** lorin11 has joined #openstack-dev | 16:17 | |
| dolphm | ayoung: awesome | 16:17 |
| *** ociuhandu has quit IRC | 16:18 | |
| *** ociuhandu has joined #openstack-dev | 16:18 | |
| ayoung | dolphm, it is from the clean role list function. | 16:18 |
| *** esheffield has joined #openstack-dev | 16:19 | |
| *** networkstatic has quit IRC | 16:19 | |
| *** edmund has joined #openstack-dev | 16:20 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Fixes import order nits https://review.openstack.org/22956 | 16:21 |
| ayoung | dolphm, so the problem is with the sql backend, when getting called from the controller. I'll post afix shortly | 16:21 |
| dolphm | ayoung: thanks | 16:21 |
| *** boris-42 has quit IRC | 16:22 | |
| *** gasbakid has joined #openstack-dev | 16:23 | |
| *** gasbakid_ has joined #openstack-dev | 16:23 | |
| *** jaypipes has joined #openstack-dev | 16:23 | |
| dolphm | ayoung: POST /trusts {"trust": {}} returns a 403 instead of a 400 | 16:23 |
| *** gasbakid has quit IRC | 16:24 | |
| ayoung | dolphm, ah, probably because of Policy denying it. | 16:24 |
| dolphm | ayoung: that's what i was thinking | 16:24 |
| dolphm | ayoung: lower priority, but captured in a test for later | 16:24 |
| dolphm | ayoung: Expecting to find roles in trust. <-- roles shouldn't be required (ability to delegate nothing) | 16:25 |
| ayoung | dolphm, intersting conundrum, which takes priority, validation or policy. I'm almosted tempted to fix that in the policy layer | 16:25 |
| *** cp16net is now known as cp16net|away | 16:26 | |
| ayoung | dolphm, meh. Trust with no roles is probably a mistake | 16:26 |
| dolphm | ayoung: i think it makes sense to raise 400 from policy if the user is expecting an attribute that wasn't provided | 16:26 |
| dolphm | ayoung: might be weird for users to debug, but it'd be their own fault | 16:26 |
| dolphm | ayoung: i like the idea, because it makes sense to set impersonation=True but without roles | 16:26 |
| ayoung | dolphm, agreed. I'll open a ticket for that | 16:26 |
| dolphm | ayoung: so, you can get an unscoped token as me | 16:27 |
| *** rushiagr has joined #openstack-dev | 16:27 | |
| ayoung | dolphm, OK, I can remove the check that roles must have something in them. | 16:27 |
| dolphm | ayoung: don't expect the attribute at all, either | 16:27 |
| *** digitalsanctum has quit IRC | 16:27 | |
| lorin11 | When is the "AUDIT" level of logging used? | 16:28 |
| ayoung | dolphm, the token code raises an exception if there are no roles, | 16:28 |
| *** bswartz has joined #openstack-dev | 16:28 | |
| dolphm | ayoung: empty should be cool too though | 16:28 |
| ayoung | feature add? | 16:28 |
| *** jsindy has quit IRC | 16:28 | |
| *** gargya has quit IRC | 16:28 | |
| dolphm | ayoung: why would the token code fail without roles? that's what an unscoped token is | 16:28 |
| dolphm | ayoung: oh, if a project_id was provided, then roles should be provided | 16:28 |
| dolphm | ayoung: they go hand-in-hand | 16:28 |
| dolphm | ayoung: i'll amend the spec on that | 16:28 |
| *** digitalsanctum has joined #openstack-dev | 16:29 | |
| ayoung | dolphm, https://bugs.launchpad.net/keystone/+bug/1133435 | 16:29 |
| uvirtbot | Launchpad bug 1133435 in keystone "policy should return a 400 if a required filed is missing" [Undecided,New] | 16:29 |
| *** ondergetekende has quit IRC | 16:29 | |
| *** john5223 has joined #openstack-dev | 16:30 | |
| *** danwent has joined #openstack-dev | 16:30 | |
| dolphm | ayoung: filed -> field | 16:30 |
| *** wowdd1 has quit IRC | 16:30 | |
| ayoung | dolphm, which file? | 16:30 |
| dolphm | ayoung: your bug title | 16:30 |
| ayoung | got it | 16:33 |
| mordred | zul: pong | 16:33 |
| *** mrodden has joined #openstack-dev | 16:33 | |
| zul | mordred/jamespage: just opened https://bugs.launchpad.net/oslo/+bug/1133433 we wanted to talk to you about it first | 16:34 |
| uvirtbot | Launchpad bug 1133433 in oslo "Running python setup.py sdist for quantum creates the wrong version." [Undecided,New] | 16:34 |
| zul | jamespage: ^^^ | 16:34 |
| mordred | zul: oh gosh | 16:35 |
| mordred | zul: not more version code | 16:35 |
| *** bdpayne has joined #openstack-dev | 16:35 | |
| mordred | is quantum using it differently than glance and nova and friends? | 16:35 |
| zul | mordred: yoru favorite! | 16:35 |
| *** trapni has quit IRC | 16:35 | |
| *** mrodden1 has quit IRC | 16:35 | |
| mordred | zul: lemme look real quick... | 16:36 |
| zul | mordred: i think so it seems to be stripping off the g3-38 bits | 16:36 |
| zul | mordred: k | 16:36 |
| *** gyee has joined #openstack-dev | 16:36 | |
| mordred | zul: oh - nope. that's right | 16:37 |
| zul | erm? | 16:37 |
| *** sacharya has joined #openstack-dev | 16:37 | |
| mordred | zul: it's mildly weird because we used g-based tags for consistency since we'd used ~g1 and ~g2 before | 16:37 |
| zul | so where is the a38 coming from? | 16:37 |
| mordred | zul: per-commit non-tagged versions are still indicating pre-release | 16:38 |
| mordred | so | 16:38 |
| *** dontalton has joined #openstack-dev | 16:38 | |
| mordred | we take the version that we are working towards from setup.py - which is 2013.1 | 16:38 |
| zul | right | 16:38 |
| mordred | we append an alpha tag (a26) | 16:38 |
| mordred | and then the git sha | 16:38 |
| *** radez is now known as radez_g0n3 | 16:38 | |
| zul | ok | 16:38 |
| mordred | if we kept the g3 in there, it would look like we were working towards g3 | 16:39 |
| mordred | (this is baesd on python version sorting) | 16:39 |
| *** gargya has joined #openstack-dev | 16:39 | |
| mordred | NOW - there is a bug in that | 16:39 |
| mordred | because I'm pretty sure the a count is off, given that there was a g3-based tag | 16:40 |
| *** gargya has quit IRC | 16:40 | |
| *** rushiagr has quit IRC | 16:40 | |
| *** sacharya has quit IRC | 16:40 | |
| *** devoid has quit IRC | 16:41 | |
| *** sacharya has joined #openstack-dev | 16:41 | |
| *** trapni has joined #openstack-dev | 16:42 | |
| *** devoid has joined #openstack-dev | 16:42 | |
| *** sstent_ has quit IRC | 16:42 | |
| *** sstent_ has joined #openstack-dev | 16:42 | |
| zul | mordred: well ill leave it with you then | 16:42 |
| zul | jamespage: unless if you have somethign to add | 16:42 |
| dolphm | ayoung: are you working on not requiring roles? still getting a 400 on that in patchset 34 | 16:42 |
| *** zeriouz has quit IRC | 16:42 | |
| ayoung | dolphm, no, I posted that before we talked | 16:43 |
| *** thouveng has quit IRC | 16:43 | |
| *** mdomsch has quit IRC | 16:43 | |
| *** mohits has quit IRC | 16:43 | |
| dolphm | ayoung: err, you're not working on it? | 16:43 |
| dolphm | ayoung: or are working on it, and i shouldn't expect it in 34 :P | 16:43 |
| ayoung | I was not working on it | 16:44 |
| jamespage | mordred, but the alpha tag is not sequential which is what is causing us a bit of a headache | 16:44 |
| jamespage | mordred, its rev's since last tag | 16:44 |
| ayoung | dolphm, I was cleaning up the fact that I query the global role list twice | 16:44 |
| mordred | jamespage: yes, you are right. we don't have any context to know revs since last tag-we-care-about | 16:44 |
| ayoung | dolphm, I'll remove the check, though. | 16:44 |
| mordred | jamespage: I suppose we could go back to revs-since-beginning-of-time | 16:45 |
| openstackgerrit | A change was merged to openstack/tempest: Implements test_update_all_metadata_field_error. https://review.openstack.org/22663 | 16:45 |
| mordred | but it's still going to cause you problems | 16:45 |
| dolphm | ayoung: cool | 16:45 |
| jamespage | mordred, so long as revs continually increments I'm good :-) | 16:45 |
| mordred | jamespage: well, they're not going to from a debian perspective regardless | 16:45 |
| mordred | without modification from you | 16:46 |
| jamespage | mordred, well for the 'released' versions of the packages we can handle that | 16:46 |
| garyk | arosen: ping | 16:46 |
| jamespage | mordred, this is more for the CI stuff | 16:46 |
| mordred | jamespage: indeed. so, the problem is that the tags will be things like 2013.2.b1 and 2013.2.b2 | 16:46 |
| mordred | jamespage: but per-commit versions will be things like 2013.2.aXXX.gXXX | 16:47 |
| mordred | jamespage: regardless of whether they came before or after a particular milesotne | 16:47 |
| jamespage | mordred, I can handle that in the lab versioning | 16:47 |
| jamespage | the output from the lab never gets to the archive - only to a PPA | 16:47 |
| mordred | cool. so what am I breaking? | 16:47 |
| jamespage | mordred, lemme give you an example | 16:47 |
| mordred | ossum | 16:48 |
| openstackgerrit | A change was merged to openstack/tempest: Simplify xml-json inheritance in identity https://review.openstack.org/22788 | 16:48 |
| jamespage | mordred, context cinder | 16:49 |
| jamespage | 2013.1.a12.gb3aa798+git201302260941~raring-0ubuntu1 <= 2013.1.a201.g66532ca+git201302202031~raring-0ubuntu1 | 16:49 |
| jamespage | 2013.1.a201.g66532ca was uploaded pre g3 | 16:49 |
| jamespage | rev counter resets since the 2013.1.g3 tag | 16:49 |
| jamespage | so 2013.1.a12.gb3aa798 < 2013.1.a201 | 16:50 |
| jamespage | .g66532ca | 16:50 |
| mordred | yes | 16:50 |
| dolphm | ayoung: extra data on the entity seems to be ignored? | 16:51 |
| mordred | jamespage: only thing I can think of there is to replace the revcount portion with one calculated by doing git log --online | wc -l | 16:51 |
| dolphm | ayoung: if i add a description it's being discarded | 16:51 |
| mordred | s/online/oneline/ | 16:51 |
| ayoung | dolphm, there is no description field. There is no extra. | 16:51 |
| mordred | jamespage: biggest problem with that is that for nova it takes 0.3 seconds to run that | 16:52 |
| ayoung | dolphm, It is in the db, but Ignored it elsewhere. | 16:52 |
| mordred | but - perhaps that isn't a terrible problem | 16:52 |
| dolphm | ayoung: ... use it | 16:52 |
| *** zaitcev has joined #openstack-dev | 16:52 | |
| *** hemna has joined #openstack-dev | 16:53 | |
| dolphm | ayoung: every other entity supports extra attributes | 16:53 |
| vishy | zul: care to approve the chain ending here: https://review.openstack.org/#/c/22945/ ? | 16:53 |
| openstackgerrit | A change was merged to openstack/glance: Fix broken JSON schemas in v2 tests https://review.openstack.org/22914 | 16:53 |
| jamespage | mordred, or you keep context to the rev count by including the full tag | 16:54 |
| zul | vishy: done | 16:54 |
| jamespage | 2013.1.g3.aXX.YYYYY | 16:54 |
| *** jsindy has joined #openstack-dev | 16:54 | |
| jamespage | mordred, I see you said 'if we kept the g3 in there, it would look like we were working towards g3' | 16:55 |
| *** mindpixel has quit IRC | 16:55 | |
| mordred | yes | 16:55 |
| jamespage | I guess that is one way of looking at it | 16:55 |
| *** zeriouz has joined #openstack-dev | 16:56 | |
| mordred | because 2013.1.g3.aXXX in python will sort like an alpha of g3 | 16:56 |
| ayoung | dolphm, OK. I'll get that, too. | 16:56 |
| jamespage | gah | 16:56 |
| mordred | which is itself a pre-release of 2013.1 | 16:56 |
| gyee | dolphm, ayoung, henrynash, https://review.openstack.org/#/c/22893/ | 16:56 |
| mordred | yeah | 16:56 |
| mordred | jamespage: I promise, ever gah you feel here I'm right there with you on | 16:56 |
| mordred | jamespage: it turns out there is NO versioning model which works for pre-release versioning in a vendor-neutral manner | 16:57 |
| henrynash | gyee: ok, looking at it | 16:57 |
| ayoung | gyee, after I get done with trusts | 16:57 |
| mordred | the only version schemes which sort properly with everyones tools are post-release versioning | 16:57 |
| ayoung | gyee, It counts as a bug fix, and can go in tonight. I need trusts in by the meeting this evening | 16:57 |
| gyee | ayoung, I had a bunch of comments for trusts impl, did you have a chance to look at them? | 16:57 |
| mordred | jamespage: it makes me want to die | 16:57 |
| jamespage | mordred, OK - I think I will rework the versioning on the CI lab to override the generated version number | 16:57 |
| ayoung | gyee, which review #? I might have | 16:58 |
| *** markmcclain has quit IRC | 16:58 | |
| mordred | jamespage: probably the safest thing to do - esesntially, assume the version I give you on a per-commit basis is meaningless | 16:58 |
| *** nunosantos has joined #openstack-dev | 16:58 | |
| *** trapni has quit IRC | 16:58 | |
| dolphm | gyee: ayoung is busy ;) | 16:58 |
| ayoung | and tired | 16:58 |
| jamespage | mordred, I note that OSLO_PACKAGE_VERSION looks like it will allow us to override+ still use sdist | 16:58 |
| *** rpedde_away is now known as rpedde | 16:58 | |
| mordred | jamespage: yes. we put that there just for you | 16:59 |
| mordred | well, actually I believe for SuSE ... but you know :) | 16:59 |
| jamespage | mordred, right-oh! | 16:59 |
| jamespage | mordred, you almost made me feel loved for a minute then ;-) | 16:59 |
| mordred | jamespage: you are loved ... I promise ... | 16:59 |
| jamespage | mordred, lol | 16:59 |
| mordred | jamespage: everytime it feels like I'm screwing you over, I probably could have made things much worse but held back because I like you :) | 17:00 |
| *** markmcclain has joined #openstack-dev | 17:00 | |
| gyee | ayoung, sorry, do your thing | 17:01 |
| ayoung | gyee, on trust impl, what review numbers? | 17:01 |
| *** jcmartin has joined #openstack-dev | 17:01 | |
| gyee | ayoung, 28 | 17:01 |
| ayoung | looking | 17:02 |
| dolphm | gyee: +2 other than nit | 17:02 |
| gyee | ayoung, I can review it one more time | 17:02 |
| dolphm | gyee: tests are awesome | 17:02 |
| openstackgerrit | A change was merged to openstack/nova: Additional tests for safe parsing with minidom https://review.openstack.org/22442 | 17:02 |
| ayoung | gyee, I saw that you -1 it, but I didn;t see the line comments. Did you post them? | 17:03 |
| gyee | dolphm, thanks, I'll correct the comment msg | 17:03 |
| dolphm | ayoung: commit msg | 17:03 |
| ayoung | gyee, I see your patch 16 messages | 17:04 |
| *** yidclare has joined #openstack-dev | 17:04 | |
| dolphm | gyee: np, bug will remain Confirmed forever otherwise, because we'll never notice :P | 17:04 |
| ayoung | gyee, and I agree, but can I do that on a future commit? | 17:05 |
| dolphm | ayoung: gyee: what's the issue on trusts? | 17:05 |
| ayoung | dolphm, sorry, test cleanup | 17:05 |
| *** sandywalsh has quit IRC | 17:06 | |
| gyee | dolphm, trust have expires_at while the spec have expires | 17:06 |
| gyee | just some minor inconsistencies | 17:06 |
| dolphm | ayoung: on the /trusts/{trust_id}/roles stuff, write it in the RoleV3 controller so you get the wrappers for free, and then call back to self.trust_api to get the list of roles for the trust, etc | 17:06 |
| dolphm | gyee: i'd rather fix the spec on that, i prefer _at | 17:07 |
| *** jsindy has quit IRC | 17:07 | |
| dolphm | gyee: heckj blocked my similar patch for tokens | 17:07 |
| ayoung | dolphm, can I call that from the Trust Controlle | 17:07 |
| *** jsindy has joined #openstack-dev | 17:07 | |
| dolphm | ayoung: ? you shouldn't need to -- have the role controller call the trust driver | 17:08 |
| dolphm | ayoung: setup routing in identity_api as well | 17:08 |
| ayoung | dolphm, on create trust? | 17:08 |
| *** ijw1 has joined #openstack-dev | 17:08 | |
| dolphm | ayoung: err, not identity_api... identity.routers | 17:08 |
| ayoung | dolphm, ah...can I punt on that for now? It really is not needed for base functionality | 17:08 |
| dolphm | ayoung: create_trust can call self.identity_api.get_role() | 17:08 |
| ayoung | dolphm, yeah, doing that | 17:09 |
| dolphm | ayoung: i'd disagree, and it's pretty simple to impl | 17:09 |
| *** jgallard has quit IRC | 17:09 | |
| *** sandywalsh has joined #openstack-dev | 17:09 | |
| dolphm | ayoung: not having basic crud isn't a bug, it's a missing feature | 17:09 |
| ayoung | dolphm, what is needed is the ability to create trusts, and to execute them. There is a way to query them, and the roles | 17:09 |
| gyee | ayoung, also, in auth/method/tokens.py, you can just raise Unauthorized if user attempt to authenticate with a trust token | 17:10 |
| ayoung | let me get the extra...in there | 17:10 |
| *** ijw has quit IRC | 17:10 | |
| gyee | only use case for token auth is re-scoping, trust token can't be used for re-scoping | 17:10 |
| *** ijw1 has quit IRC | 17:10 | |
| *** ijw has joined #openstack-dev | 17:11 | |
| *** john5223 has quit IRC | 17:11 | |
| *** mmagr has quit IRC | 17:11 | |
| dolphm | gyee: want me to revise your commit message? | 17:11 |
| dolphm | gyee: i'd like to +2 :) | 17:11 |
| *** lglenden has joined #openstack-dev | 17:12 | |
| gyee | dolphm, I am rebasing | 17:12 |
| dolphm | gyee: ah, cool | 17:13 |
| gyee | will push another review with the correct comment shortly | 17:13 |
| dolphm | gyee: that's why i want to +2 -- i know i'm going to rebase on top of yours | 17:13 |
| *** edmund has quit IRC | 17:13 | |
| dolphm | gyee: but also have to test ayoung's | 17:13 |
| *** NobodyCam has joined #openstack-dev | 17:13 | |
| gyee | dolphm, looks like my rebasing is not happen, time for some manual merge | 17:14 |
| gyee | s/happen/happy | 17:14 |
| *** hemna has quit IRC | 17:15 | |
| *** romcheg1 has joined #openstack-dev | 17:20 | |
| *** romcheg1 has left #openstack-dev | 17:20 | |
| *** heckj has joined #openstack-dev | 17:21 | |
| gyee | dolphm, new patch is ready | 17:22 |
| *** heckj has quit IRC | 17:23 | |
| *** heckj has joined #openstack-dev | 17:23 | |
| *** john5223 has joined #openstack-dev | 17:24 | |
| *** hemna has joined #openstack-dev | 17:25 | |
| *** eharney has quit IRC | 17:27 | |
| *** eharney has joined #openstack-dev | 17:28 | |
| jgriffith | jamespage: ping | 17:28 |
| *** adjohn has quit IRC | 17:28 | |
| jamespage | jgriffith, hey! | 17:28 |
| jgriffith | jamespage: hola | 17:28 |
| jgriffith | jamespage: So I wanted to try and recreate what you're seeing... | 17:29 |
| ayoung | dolphm, updated trusts patch. Does not have the roles controller work, but should store extra. Haven't added a test for that yet. | 17:29 |
| jamespage | jgriffith, please - how can I help? | 17:29 |
| jgriffith | jamespage: anything special I need to know? Your config looks fine | 17:29 |
| dolphm | ayoung: no worries, i've already got the test | 17:29 |
| jgriffith | jamespage: Are you doing devstack or packages or ???? | 17:29 |
| ayoung | figured | 17:29 |
| jamespage | jgriffith, well this is running on Ubuntu raring from packages | 17:29 |
| jgriffith | jamespage: Ok... | 17:29 |
| dolphm | ayoung: still fleshing stuff out, but i want to get this into review on top of yours soon | 17:29 |
| jamespage | jgriffith, I did a bit of debugging and the code that searches for the filters does not seem to find them | 17:29 |
| jamespage | i.e. the array is empty | 17:30 |
| jgriffith | jamespage: odd | 17:30 |
| jgriffith | jamespage: It should just pick that up from the cinder package defaults | 17:30 |
| jgriffith | jamespage: wondering if we have a compat issue on raring | 17:30 |
| jgriffith | jamespage: I'll try the easy approach first and do devstack on Raring and see if I repro that way | 17:30 |
| jamespage | jgriffith, I did wonder; which deps are used for discovery of the filters? | 17:30 |
| jgriffith | Unless you already know that works | 17:30 |
| *** markmc has quit IRC | 17:31 | |
| jgriffith | jamespage: TBH there's been quite a bit of churn in deps lately | 17:31 |
| jgriffith | jamespage: But it should just be picked up by the cinder package install | 17:31 |
| jgriffith | jamespage: I'll have a look and shout when I get something going | 17:31 |
| jamespage | jgriffith, great! | 17:32 |
| jgriffith | jamespage: Probably be a little bit before I get ot it | 17:32 |
| jamespage | jgriffith, sure - I'll check back in later | 17:32 |
| jgriffith | jamespage: BTW, testing on Quantal or Precise? | 17:32 |
| jgriffith | jamespage: In other words have you done the same tests there? | 17:32 |
| dolphm | ayoung: this fixes expires -> expires_at in the api to match your impl, and adds issued_at to tokens https://review.openstack.org/#/c/22039/ | 17:33 |
| *** ociuhandu has quit IRC | 17:33 | |
| dolphm | gyee: ^ | 17:33 |
| jamespage | jgriffith, I have reports of the same issue on precise with grizzly from the cloud archive | 17:33 |
| dolphm | heckj: you downvoted this earlier ^ | 17:33 |
| jgriffith | jamespage: boooo | 17:33 |
| *** esp1 has joined #openstack-dev | 17:33 | |
| jgriffith | jamespage: that means it's packages dependent | 17:33 |
| jamespage | jgriffith, I think so | 17:33 |
| heckj | dolphm: ayoung looking | 17:33 |
| jgriffith | sighh, I hate installing via packages :) | 17:33 |
| dolphm | heckj: not sure if we answered your question out of band? | 17:33 |
| ayoung | dolphm, I wouldn't suggest changing expires to expires_at right now. The amount of churn it will generate in the tokens api is not worth it. Not sure that it should be changed | 17:34 |
| ayoung | I guess for V3 it is ok.... | 17:34 |
| dolphm | heckj: issued_at is intended to avoid token collisions when PKI is encrypting two otherwise identical tokens | 17:34 |
| heckj | gotcha | 17:34 |
| dolphm | ayoung: for self-consistency, i think it's worth fixing | 17:35 |
| ayoung | dolphm, I guess if we vary from the spec it is a bug? | 17:36 |
| ayoung | OK, I can get behind that | 17:36 |
| gyee | dolphm, +2 on expires_at | 17:36 |
| gyee | dolphm, hold off on https://review.openstack.org/#/c/22893/, I need to fix expires_at so I don't have to do two reviews | 17:37 |
| heckj | dolphm: I'm good with it - we are introducing a bit of churn with expired to expired_at, but nothing is deeply using this yet, so it's the best time to make it self consistent | 17:37 |
| *** roampune has joined #openstack-dev | 17:37 | |
| dolphm | heckj: yes, i'm happy to fix | 17:37 |
| ayoung | gyee, no | 17:38 |
| gyee | dolphm, heckj, should I make the code change now or wait | 17:38 |
| ayoung | gyee, I am afraid that if you chage 22893, It will bump the trusts. Lets fix both of them afterwards | 17:38 |
| gyee | ayoung, should I wait? | 17:38 |
| heckj | dolphm: heh, figured if you were pushing it, you'd be ready to dive onto the code and fix it too :-) | 17:38 |
| dolphm | gyee: don't hold off, i'll address seperately | 17:38 |
| gyee | dolphm, so https://review.openstack.org/#/c/22893/ is good as is? | 17:38 |
| *** andrewbogott_afk is now known as andrewbogott | 17:39 | |
| *** amerine has joined #openstack-dev | 17:39 | |
| sstent_ | anyone seen 'Cannot delete or update a parent row: a foreign key constraint fails' on trying the nova-manage db archive_deleted_rows command? | 17:39 |
| dolphm | gyee: you unblocked that test regarding domain scoping? | 17:39 |
| dolphm | gyee: did that get implemented? | 17:39 |
| gyee | dolphm, domain-scoping is implemented | 17:40 |
| dolphm | gyee: awesome | 17:40 |
| dolphm | gyee: +2 | 17:41 |
| gyee | w00t! | 17:41 |
| gyee | heckj, what's the procedure to get it into rc1? | 17:41 |
| *** jpich has quit IRC | 17:41 | |
| *** BobBall_ has quit IRC | 17:42 | |
| *** gael has quit IRC | 17:42 | |
| ayoung | dolphm, I fat fingered the 'extra' thing. Got it fixed now. What are you suggesting I do on roles? Add a member function toe trh roles v3 controller list_roles_for_trust? | 17:43 |
| ayoung | to the ... | 17:43 |
| dolphm | ayoung: exactly | 17:43 |
| ayoung | OK..doing that now. | 17:43 |
| dolphm | list_roles_for_trust, get_role_for_trust, and check_role_for_trust | 17:43 |
| heckj | gyee: sorry, not entirely understanding the question - what do you mean "get it into rc1"? | 17:44 |
| *** rafaduran has left #openstack-dev | 17:44 | |
| gyee | heckj, are all the bug fixes from now on automatically goes into rc1? | 17:45 |
| gyee | or do we need to make explicit requests? | 17:45 |
| heckj | gyee: ah, yeah - ttx has a script or something that takes all resolved bugs at the end of the release period | 17:45 |
| gyee | awesome | 17:45 |
| *** anniec has joined #openstack-dev | 17:46 | |
| *** ilblackd_ has quit IRC | 17:46 | |
| heckj | yeah, it's a nice setup | 17:46 |
| ayoung | dolphm, that should go on the public router? | 17:46 |
| heckj | gyee: we can assign bugs explicitly to RC1, the meaning of which is "hold the RC1 release until this is fixed", but otherwise we leave assignment into RC1 to the close-out cycle | 17:47 |
| gyee | heckj, sounds good, I just want to make sure bug 1131840 is includes | 17:48 |
| uvirtbot | Launchpad bug 1131840 in keystone "v3 auth API untranslatable to XML" [High,In progress] https://launchpad.net/bugs/1131840 | 17:48 |
| gyee | included | 17:48 |
| heckj | gyee: you should be able to tag it to RC1, which will do what you want (i.e. request that we block the release until that's fixed) | 17:48 |
| heckj | if you can't retarget the bug, let me know - that's just a permissions thing on Launchpad | 17:48 |
| heckj | Nvmd, looks like it already is at rc1 | 17:49 |
| dolphm | heckj: when is rc2 being forked? | 17:49 |
| gyee | heckj, looks like we are OK | 17:49 |
| dolphm | heckj: err rc1 | 17:49 |
| gyee | ayoung, will the trust APIs be available in v2.0? | 17:50 |
| ayoung | gyee, no | 17:51 |
| heckj | dolphm: there's no set time, ttx I'm sure will be telling us in the RC meeting later today. https://wiki.openstack.org/wiki/GrizzlyReleaseSchedule is all we have for formal schedule right now | 17:51 |
| *** iartarisi has quit IRC | 17:52 | |
| heckj | ttx: around? | 17:53 |
| *** cp16net|away is now known as cp16net | 17:55 | |
| *** jog0 has joined #openstack-dev | 17:55 | |
| *** zeriouz has quit IRC | 17:55 | |
| ayoung | dolphm, got it. | 17:56 |
| ayoung | posing new patch | 17:56 |
| dolphm | ayoung: awesome | 17:56 |
| *** ondergetekende has joined #openstack-dev | 17:57 | |
| *** derekh has quit IRC | 17:58 | |
| *** markwash has joined #openstack-dev | 17:58 | |
| ayoung | dolphm, can you post your tests? | 17:59 |
| *** ondergetekende has quit IRC | 17:59 | |
| *** darraghb has quit IRC | 17:59 | |
| *** aeperezt has quit IRC | 18:00 | |
| dolphm | ayoung: erm, you're returning 'extra' in the response now | 18:00 |
| dolphm | ayoung: and it's empty | 18:01 |
| dolphm | ayoung: ah, remove 'extra' from TrustModel.attributes | 18:01 |
| *** kagan has joined #openstack-dev | 18:02 | |
| *** radez_g0n3 is now known as radez | 18:03 | |
| ayoung | dolphm, I thought it should be there. No? | 18:03 |
| *** aeperezt has joined #openstack-dev | 18:03 | |
| ayoung | dolphm, If I remote it from attributes, I thought it never gets returned. Isn't that what populates the dictionary? | 18:03 |
| dolphm | ayoung: that's basically a list of contract attributes, which extra is not | 18:04 |
| ayoung | OK...lets see if it breaks... | 18:04 |
| *** nachi has joined #openstack-dev | 18:05 | |
| dolphm | ayoung: you don't want 'extra' to be returned verbatim -- sql.DictBase.to_dict() will unpack extra for you | 18:06 |
| dolphm | ayoung: TrustRole doesn't need an extra column, doesn't need to extend sql.DictBase, nor have a list of attributes | 18:06 |
| ayoung | dolphm, SO just sql.ModelBase? | 18:07 |
| *** flepied1 has joined #openstack-dev | 18:08 | |
| dolphm | ayoung: yeah | 18:09 |
| dolphm | ayoung: you don't need to call to_dict on it, do you? it's an internal model | 18:09 |
| *** lorin11 has quit IRC | 18:10 | |
| dolphm | ayoung: internal to the driver | 18:10 |
| ayoung | dolphm, you mean just return trust_ref? | 18:11 |
| *** lorin1 has joined #openstack-dev | 18:11 | |
| *** torandu has quit IRC | 18:12 | |
| *** flepied has quit IRC | 18:12 | |
| *** torandu has joined #openstack-dev | 18:12 | |
| dolphm | ayoung: maybe we're talking about two different issues | 18:12 |
| *** bryansd has quit IRC | 18:12 | |
| *** mrodden1 has joined #openstack-dev | 18:12 | |
| ayoung | dolphm, I think I'm good. | 18:12 |
| dolphm | ayoung: TrustModel should extend DictBase because it's exposed to the controller | 18:12 |
| dolphm | ayoung: TrustRole should not | 18:12 |
| openstackgerrit | A change was merged to openstack/tempest: test_live_block_migration cleanup https://review.openstack.org/22790 | 18:12 |
| openstackgerrit | A change was merged to openstack-dev/devstack: move sysstat earlier (but not too early) https://review.openstack.org/22984 | 18:12 |
| ayoung | I am goingto remove that field from the database. | 18:12 |
| openstackgerrit | A change was merged to openstack/tempest: Update stress tests to properly use tempest.config https://review.openstack.org/22902 | 18:12 |
| *** Ryan_Lane has joined #openstack-dev | 18:13 | |
| *** Yada has quit IRC | 18:14 | |
| *** mrodden has quit IRC | 18:15 | |
| *** lorin1 has quit IRC | 18:16 | |
| dolphm | ayoung: trust_roles.extra? (or whatever that table is called) | 18:16 |
| ayoung | dolphm, yep, its agone! | 18:16 |
| *** ilblackdragon has joined #openstack-dev | 18:16 | |
| *** avishay has quit IRC | 18:17 | |
| ayoung | dolphm, #37 is posted | 18:18 |
| dolphm | ayoung: /salute | 18:18 |
| *** dhellmann has quit IRC | 18:19 | |
| *** sacharya has quit IRC | 18:20 | |
| *** woodspa has joined #openstack-dev | 18:21 | |
| *** morganfainberg has joined #openstack-dev | 18:24 | |
| *** sacharya has joined #openstack-dev | 18:24 | |
| *** diogogmt has joined #openstack-dev | 18:27 | |
| *** armaan has joined #openstack-dev | 18:29 | |
| dolphm | ayoung: extra attributes still being discarded, posted a partial diff | 18:29 |
| ayoung | dolphm, I have a test in place that shows description getting returned | 18:30 |
| ayoung | dolphm, also, I am rebasing on top of gyee's recent | 18:30 |
| *** nati_ueno has joined #openstack-dev | 18:30 | |
| dolphm | ayoung: cool | 18:30 |
| ayoung | dolphm, but I like the code in that diff. Let me get mine working again, and then I'll merge that in | 18:31 |
| *** danpb has quit IRC | 18:31 | |
| *** adjohn has joined #openstack-dev | 18:31 | |
| *** epim_ has joined #openstack-dev | 18:32 | |
| *** epim has quit IRC | 18:33 | |
| *** epim_ is now known as epim | 18:33 | |
| *** anniec has quit IRC | 18:34 | |
| *** anniec_ has joined #openstack-dev | 18:34 | |
| *** jbr_ is now known as jbr_afk | 18:36 | |
| *** vipul|away is now known as vipul | 18:36 | |
| YorikSar | ayoung: Hi. Looking at Trusts diff. Which parts are mostly done so that all my code beauty comments are appropriate already? | 18:39 |
| YorikSar | ayoung: Found myself adding comment to almost every line of keystone/token/controllers.py | 18:39 |
| ayoung | YorikSar, heh, not surprising | 18:39 |
| ayoung | YorikSar, that is ripe for review, | 18:40 |
| *** Ryan_Lane has quit IRC | 18:40 | |
| ayoung | the only thing changingthat I know of are the sql backend and the v3 auth code | 18:40 |
| *** Ryan_Lane has joined #openstack-dev | 18:40 | |
| *** jbr_afk has quit IRC | 18:40 | |
| YorikSar | ayoung: So I'll post that comments altogether. | 18:41 |
| ayoung | YorikSar, yes, please do. | 18:41 |
| *** kagan_ has joined #openstack-dev | 18:41 | |
| *** zzs has joined #openstack-dev | 18:42 | |
| *** flepied1 has quit IRC | 18:43 | |
| *** zoresvit has quit IRC | 18:44 | |
| *** kagan has quit IRC | 18:44 | |
| *** kagan_ is now known as kagan | 18:44 | |
| *** tomoe_ has quit IRC | 18:46 | |
| *** tomoe_ has joined #openstack-dev | 18:46 | |
| *** garyk has quit IRC | 18:48 | |
| *** rushiagr has joined #openstack-dev | 18:48 | |
| *** jkordish_ is now known as jkordish | 18:48 | |
| *** unix has joined #openstack-dev | 18:49 | |
| *** rushiagr1 has joined #openstack-dev | 18:49 | |
| *** tomoe_ has quit IRC | 18:51 | |
| YorikSar | ayoung: done | 18:51 |
| *** roampune1 has joined #openstack-dev | 18:52 | |
| *** rushiagr has quit IRC | 18:53 | |
| *** rushiagr1 has quit IRC | 18:54 | |
| *** roampune has quit IRC | 18:54 | |
| *** dev_sa has quit IRC | 18:56 | |
| *** arbrandes has quit IRC | 18:56 | |
| sdague | jgriffith: so my local devstack can't run cinder volume tests, which I think I've narrowed down to this - http://paste.openstack.org/show/32536/ | 18:57 |
| *** rushiagr has joined #openstack-dev | 18:57 | |
| jgriffith | sdague: yup, this just started showing up | 18:57 |
| sdague | realistically I don't understand how we pass in gate, as I don't have any local changes that should be an issue | 18:57 |
| jgriffith | sdague: Looks like missing egg-info in rtslib, but I'm not sure how it's impacting here | 18:58 |
| sdague | jgriffith: hmmm.. ok, my volumes fails have been really long lasting :) | 18:58 |
| jgriffith | oh? | 18:58 |
| jgriffith | I did a fresh setup last night when I was working on that no-screen option and all worked fine | 18:58 |
| sdague | yeh, some time in January volumes tests on nova stopped passing for me locally, though they passed in gate. Just hadn't gotten around to looking at it until now | 18:59 |
| jgriffith | sdague: well that disturbing | 19:00 |
| *** rnirmal_ has joined #openstack-dev | 19:00 | |
| *** rnirmal_ has quit IRC | 19:00 | |
| jgriffith | are you talking tempest? | 19:00 |
| sdague | yes | 19:00 |
| *** digitalsanctum has quit IRC | 19:00 | |
| jgriffith | k... I haven't tried tempest in a while. I'll have a look | 19:01 |
| *** morganfainberg has quit IRC | 19:01 | |
| *** bdpayne has quit IRC | 19:01 | |
| sdague | I guess the question is, is the scheduler issue going to block any other debug | 19:01 |
| *** morganfainberg has joined #openstack-dev | 19:02 | |
| jgriffith | sdague: not sure I follow the question? | 19:02 |
| *** rnirmal has quit IRC | 19:02 | |
| openstackgerrit | A change was merged to openstack/glance: Redact location from notifications https://review.openstack.org/22298 | 19:04 |
| sdague | jgriffith: I guess I was wondering if the scheduler issues were causing the fail | 19:04 |
| sdague | but I actually have my answer | 19:04 |
| sdague | and it's no (at least I think not) | 19:04 |
| *** rushiagr1 has joined #openstack-dev | 19:05 | |
| jgriffith | sdague: ahh... so yet another issue | 19:05 |
| *** lorin1 has joined #openstack-dev | 19:05 | |
| *** lorin1 has left #openstack-dev | 19:05 | |
| jgriffith | sdague: oh crud.... | 19:05 |
| openstackgerrit | A change was merged to openstack/python-glanceclient: Replace SchemaNotFound with HTTPNotFound https://review.openstack.org/22722 | 19:05 |
| jgriffith | sdague: another issue of the service not starting | 19:06 |
| *** edmund has joined #openstack-dev | 19:06 | |
| gyee | dolphm, come to think of it, I am fine with having the auth plugins to return a valid user_id and do the orchestration at the controller level | 19:06 |
| sdague | jgriffith: http://paste.openstack.org/show/32538/ so my issue is actually a permissions problem | 19:06 |
| sdague | but given that I'm just using default devstack + tempest I'm curious why | 19:06 |
| jgriffith | sdague: I ran into this | 19:08 |
| sdague | jgriffith: oh, cool, that means you probably have a solution? :) | 19:08 |
| *** rushiagr has quit IRC | 19:08 | |
| jgriffith | No... | 19:08 |
| sdague | ummm.... ok | 19:08 |
| jgriffith | But it means now it's the next thing on my list | 19:09 |
| sdague | heh, ok | 19:09 |
| jgriffith | I'm doing a fresh devstack now | 19:09 |
| sdague | mtreinish: can you check so see if that's your tempest issue as well? | 19:09 |
| *** nachi has quit IRC | 19:09 | |
| sdague | it's entirely possible it's actually a nova bug, as it's in the volumes extension | 19:09 |
| *** johnthetubaguy has quit IRC | 19:10 | |
| jgriffith | sdague: perhaps, or just an issue in the tenant/token's being used | 19:10 |
| jgriffith | sdague: ie the gates pick up a step that we miss | 19:11 |
| mtreinish | sdague: sure I'll check | 19:11 |
| *** dhellmann has joined #openstack-dev | 19:11 | |
| sdague | jgriffith: yeh, that could be | 19:12 |
| *** annegentle has joined #openstack-dev | 19:12 | |
| *** olaph has quit IRC | 19:13 | |
| mtreinish | sdague: no mine looks like an iscsi config issue. | 19:14 |
| mtreinish | sdague: it throws create error iscsi errors on create | 19:14 |
| sdague | hmmm ok | 19:14 |
| *** rushiagr1 has quit IRC | 19:14 | |
| *** olaph has joined #openstack-dev | 19:14 | |
| *** imsplitbit has quit IRC | 19:15 | |
| *** zing has quit IRC | 19:18 | |
| jgriffith | mtreinish: logs? | 19:21 |
| *** rnirmal has joined #openstack-dev | 19:21 | |
| *** digitalsanctum has joined #openstack-dev | 19:22 | |
| *** afazekas has joined #openstack-dev | 19:22 | |
| *** turul_ has joined #openstack-dev | 19:22 | |
| mtreinish | jgriffith: I'll pastebin them one sec, but I suspect that I just broke the config by doing manual changes to iscsi when I was debugging the tempest volume stuff. | 19:23 |
| *** Mandell has joined #openstack-dev | 19:23 | |
| *** anniec_ has quit IRC | 19:24 | |
| *** anniec has joined #openstack-dev | 19:24 | |
| *** zbitter has joined #openstack-dev | 19:25 | |
| jgriffith | mtreinish: Well that sounds fine to me :) | 19:25 |
| jgriffith | sdague: running tempest now... I'll let ya know | 19:25 |
| *** anniec has quit IRC | 19:26 | |
| *** anniec_ has joined #openstack-dev | 19:26 | |
| jgriffith | sdague: The problem I was seeing seems to be fixed somewhere, I've made it through ec2_volumes tests ok | 19:26 |
| *** zaneb has quit IRC | 19:26 | |
| sdague | jgriffith: I think I fail later | 19:27 |
| sdague | ec2 takes a different path | 19:27 |
| mtreinish | jgriffith: http://paste.openstack.org/show/32540/ I can paste more, but it's really just more of the same | 19:29 |
| sc68cal | Anyone from the nova-conductor team that is available to review a change - https://review.openstack.org/23012 | 19:29 |
| *** annegentle has left #openstack-dev | 19:31 | |
| *** digitalsanctum has quit IRC | 19:31 | |
| vkmc | Is there anybody with some experience using Mox? I just can't see where are this tests failing -.- http://paste.openstack.org/show/32541/ | 19:32 |
| *** zaneb has joined #openstack-dev | 19:32 | |
| *** almaisan-away is now known as al-maisan | 19:32 | |
| *** zbitter has quit IRC | 19:33 | |
| jgriffith | sdague: hmmm.... awful lot of failures in here | 19:34 |
| *** digitalsanctum has joined #openstack-dev | 19:34 | |
| *** flepied has joined #openstack-dev | 19:37 | |
| *** markmcclain has quit IRC | 19:37 | |
| sc68cal | vkmc: Do you have a diff of your changes to horizon? | 19:37 |
| *** gyee has quit IRC | 19:38 | |
| sc68cal | vkmc: looks like you added a call to a new method somewhere in there. You need to tell Mox to stub it out | 19:38 |
| *** jsindy has quit IRC | 19:38 | |
| vkmc | sc68cal, Thanks for looking at it! Here is the diff http://paste.openstack.org/show/32542/ | 19:39 |
| *** ondergetekende has joined #openstack-dev | 19:39 | |
| vkmc | sc68cal, You may be right... yeah... my environment is kinda messy though | 19:40 |
| *** flepied has quit IRC | 19:40 | |
| topol | YorikSar, easy one if you have a minute. https://review.openstack.org/#/c/22818/ | 19:40 |
| zykes- | :) vkmc | 19:41 |
| vkmc | zykes- o/ | 19:41 |
| *** olaph_ has joined #openstack-dev | 19:43 | |
| sc68cal | vkmc: I think you need to add api.nova.flavor_list(http.HttpRequest).AndReturn( <some mocked out values > ) near https://github.com/openstack/horizon/blob/master/openstack_dashboard/dashboards/admin/flavors/tests.py#L69 | 19:43 |
| vkmc | Oh thanks sc68cal, I'll give that a try! | 19:45 |
| *** otherwiseguy has quit IRC | 19:45 | |
| sc68cal | either that or try api.nova_flavor_get | 19:45 |
| *** olaph_ has quit IRC | 19:46 | |
| *** novas0x2a|laptop has joined #openstack-dev | 19:46 | |
| sc68cal | Mox is a bit tricky to learn - but essentially it gets pissed off the code you're testing calls a method that you haven't mocked out | 19:46 |
| *** olaph has quit IRC | 19:46 | |
| sc68cal | which is usually done a few lines above a mox.ReplayAll() call | 19:46 |
| jgriffith | mtreinish: thanks | 19:47 |
| *** olaph has joined #openstack-dev | 19:48 | |
| *** crandquist has quit IRC | 19:49 | |
| *** markmcclain has joined #openstack-dev | 19:50 | |
| *** brianr-gone is now known as brianr_ | 19:51 | |
| ayoung | YorikSar, all good suggestions. | 19:51 |
| *** corXi has quit IRC | 19:51 | |
| *** monst has joined #openstack-dev | 19:53 | |
| monst | has anyone gotten Folsom Keystone integrated with PAM? | 19:53 |
| *** lorin1 has joined #openstack-dev | 19:56 | |
| *** alexpilotti has joined #openstack-dev | 19:57 | |
| jgriffith | sdague: I think I know what the issue you're seeing in the tempest tests is | 19:58 |
| jbresnah_ | bcwaldon: wrt https://review.openstack.org/#/c/22753/, would it be better if I just submitted the test as a separate review? | 19:58 |
| jgriffith | sdague: are you setting secure_delete=False? | 19:58 |
| sdague | jgriffith: no | 19:58 |
| sdague | that's going to change the permissions? | 19:59 |
| *** lucid- has quit IRC | 19:59 | |
| *** markmc has joined #openstack-dev | 19:59 | |
| jgriffith | No, but the permission errors I *believe* are due to trying to delete error state volumes | 19:59 |
| sdague | oh, interesting | 19:59 |
| jgriffith | and the error states come up I *think* because we blow quota | 20:00 |
| bcwaldon | jbresnah_: feel free | 20:00 |
| jgriffith | although I added waits for secure delete.... | 20:00 |
| jgriffith | hmmm... lemme dig in a bit more | 20:00 |
| ayoung | dolphm, henrynash heckj https://review.openstack.org/#/c/23016/ trivial review that I chipped off the trusts one. | 20:01 |
| jbresnah_ | bcwaldon: i was just looking for your preference. I can leave it alone if that is not the better route. | 20:01 |
| bcwaldon | jbresnah_: go ahead and push it up for review | 20:01 |
| ayoung | actually, let me submit that standalone. doesn't need gyee;s | 20:02 |
| jbresnah_ | bcwaldon: also I am feeling guilty for holding up the other patch a bit | 20:02 |
| *** gabrielhurley has joined #openstack-dev | 20:02 | |
| jbresnah_ | bcwaldon: ok thanks | 20:02 |
| *** riskable has quit IRC | 20:02 | |
| bcwaldon | jbresnah_: don't worry! We're notoriously slow in reviews | 20:03 |
| *** hub_cap has joined #openstack-dev | 20:03 | |
| jbresnah_ | bcwaldon: i was writing it just to help me actually effectively review the submission and thought it might be useful | 20:03 |
| jbresnah_ | bcwaldon: cool enough, thanks! | 20:03 |
| hub_cap | hey, im interested in porting some of the instance-actions BP (https://wiki.openstack.org/wiki/NovaInstanceActions) into oslo so that reddwarf can consume it. anyone think of opposition or any reason _not_ to do that? | 20:04 |
| ayoung | https://review.openstack.org/#/c/23017/ dolphm heckj henrynash YorikSar | 20:04 |
| zykes- | anyone here know when the talks will be announced ? | 20:05 |
| *** melwitt has joined #openstack-dev | 20:06 | |
| heckj | zykes-: internal or external talks? | 20:08 |
| heckj | for external, I'd pester ttx later today (or maybe stephano) - for internal (dev summit topics specifically), they're still getting set up | 20:09 |
| *** al-maisan is now known as almaisan-away | 20:09 | |
| *** olaph_ has joined #openstack-dev | 20:12 | |
| ayoung | dolphm, how are we looking | 20:13 |
| dolphm | ayoung: i'm not sure i've tried 38 -- are /roles implemented? | 20:14 |
| ayoung | dolphm, yep | 20:14 |
| *** olaph_ has quit IRC | 20:14 | |
| *** olaph has quit IRC | 20:14 | |
| dolphm | ayoung: are roles removed from the POST /trusts response? | 20:14 |
| ayoung | dolphm, made the change to the test that confirms description is persisted | 20:14 |
| zykes- | heckj: summit | 20:15 |
| ayoung | dolphm, you mean GET? | 20:15 |
| dolphm | ayoung: no, on POST -- roles are consumed and then not returned in the response | 20:15 |
| dolphm | ayoung: i put a paragraph in the spec about that | 20:15 |
| dolphm | ayoung: or two, maybe | 20:15 |
| dolphm | ayoung: roles are provided by the client in the POST, but never provided in a response.. and instead are available as a read-only collection <-- or something like that | 20:16 |
| ayoung | dolphm, that is pretty easy to do, I didn't realize... | 20:16 |
| ayoung | OK...on toward 40 | 20:16 |
| dolphm | ayoung: no worries | 20:17 |
| dolphm | i'll need to rebase on 40 if you rebased ontop of gyee's patch | 20:17 |
| *** olaph has joined #openstack-dev | 20:18 | |
| *** unix has quit IRC | 20:18 | |
| heckj | zykes-: summit has just basically opened up this past week - blueprints and talks being accepted. Last cycle we coordinated and scheduled them a week or to prior to summit, so it's all still ingest right now | 20:18 |
| *** Tross has quit IRC | 20:20 | |
| *** lglenden has quit IRC | 20:20 | |
| *** john5223 has quit IRC | 20:21 | |
| *** andreaf has quit IRC | 20:22 | |
| *** olaph_ has joined #openstack-dev | 20:24 | |
| *** dprince has quit IRC | 20:24 | |
| *** andreaf has joined #openstack-dev | 20:24 | |
| dolphm | ayoung: expired trust should raise a 404, right? | 20:26 |
| *** olaph has quit IRC | 20:26 | |
| dolphm | ayoung: i'm getting a 403... i assume that's why | 20:26 |
| ayoung | dolphm, depends on the activity | 20:26 |
| ayoung | 403 is correct for getting a token but not for listing | 20:26 |
| dolphm | ayoung: GET /trusts/{expired_trust} | 20:27 |
| dolphm | ayoung: i was assuming 404 like tokens or a 200 | 20:27 |
| *** termie_ is now known as termie | 20:27 | |
| ayoung | dolphm, yeah, should be a 404 | 20:27 |
| *** termie has quit IRC | 20:27 | |
| *** termie has joined #openstack-dev | 20:27 | |
| sdague | any horizon folks around? looks like something went haywire and the gate's rejecting lots of stuff because of a django config issue | 20:27 |
| dolphm | ayoung: maybe a bad test... if i set expires to null i still get a 403 :P | 20:28 |
| ayoung | dolphm, exception.TrustNotFound | 20:28 |
| YorikSar | topol: It looks fine, but I'm not sure if businessCategory is the right type for this. | 20:28 |
| dolphm | ayoung: but it doesn't say why -- {"error": {"message": "You are not authorized to perform the requested action: %(action)s", "code": 403, "title": "Not Authorized"}} | 20:28 |
| dolphm | ayoung: something is raising ForbiddenAction() without specifying an action | 20:28 |
| YorikSar | topol: So I can neither +1 nor -1 this | 20:28 |
| ayoung | dolphm, policy? | 20:28 |
| zykes- | heckj: I mean the talks, where the voting just closed today | 20:29 |
| *** Tross has joined #openstack-dev | 20:29 | |
| dolphm | ayoung: nope- http://paste.openstack.org/raw/32546/ | 20:29 |
| *** olaph_ has quit IRC | 20:29 | |
| heckj | zykes-: ah - that I don't know - that's where I was suggesting to pester ttx :-) | 20:29 |
| dolphm | ayoung: those should all have an action='do something' | 20:29 |
| ayoung | dolphm, ah, that was me being paranoid, pre-policy | 20:31 |
| sdague | gabrielhurley: https://review.openstack.org/23021 | 20:31 |
| *** olaph has joined #openstack-dev | 20:31 | |
| sdague | I think that will get things right again | 20:31 |
| * zykes- pesters ttx | 20:31 | |
| gabrielhurley | sdague: thanks. the march of progress makes fools of us all. ;-) | 20:32 |
| sdague | heh | 20:32 |
| *** sandywalsh has quit IRC | 20:32 | |
| sdague | gabrielhurley: the sooner you can +A that through, the better. The merge gate is currently blocked on that - http://logs.openstack.org/22909/6/gate/gate-tempest-devstack-vm-full/5573/logs/screen-horizon.txt.gz | 20:33 |
| sdague | as an example | 20:33 |
| ayoung | dolphm, OK, I was swallowing the 404. I have a rebased patch...but now I am waiting on | 20:33 |
| *** davidha has quit IRC | 20:33 | |
| *** Tross has quit IRC | 20:33 | |
| *** john5223 has joined #openstack-dev | 20:34 | |
| ayoung | https://review.openstack.org/#/c/22893/ and https://review.openstack.org/#/c/23017/1 | 20:34 |
| gabrielhurley | sdague: Given that the gate will decide whether it's good or not, I guess I might as well just approve it too. | 20:34 |
| ayoung | dolphm, can I push https://review.openstack.org/#/c/23017 | 20:34 |
| ayoung | I removed the import you commented on | 20:34 |
| *** sandywalsh has joined #openstack-dev | 20:35 | |
| *** Tross has joined #openstack-dev | 20:36 | |
| *** olaph_ has joined #openstack-dev | 20:38 | |
| *** olaph has quit IRC | 20:38 | |
| *** jaypipes has quit IRC | 20:40 | |
| *** lglenden has joined #openstack-dev | 20:41 | |
| *** Ryan_Lane has quit IRC | 20:41 | |
| dolphm | ayoung: approved | 20:42 |
| *** jaypipes has joined #openstack-dev | 20:43 | |
| *** alexpilotti has quit IRC | 20:43 | |
| ayoung | dolphm, so, I am not sure how long those gate jobs will take, but I'm guessing a while. In order to post a new version of my patch without resubmitting those, I can hack the commit messages on those two patches to have a different change id | 20:45 |
| ayoung | actually, I am ok with depending on guang's patch, I just don't want to make the api-controller patch dependant on it | 20:45 |
| dolphm | ayoung: rebase on one and duplicate the diff of the other into your patch -- after they gate, it won't be seen as a change | 20:45 |
| dolphm | ayoung: cleaner hack :) | 20:46 |
| ayoung | looks like mine just gated.... | 20:46 |
| dolphm | ayoung: that was fast | 20:46 |
| dolphm | ayoung: gating or merged? | 20:46 |
| ayoung | ah, just smokestack | 20:46 |
| dolphm | ayoung: rebase onto gyee's patch, and include your own diff into your own change | 20:47 |
| ayoung | dolphm, will do | 20:47 |
| dolphm | ayoung: thanks for pushing that stuff out faster though | 20:47 |
| *** lcheng has joined #openstack-dev | 20:48 | |
| *** kagan has quit IRC | 20:49 | |
| *** vkmc has quit IRC | 20:49 | |
| ayoung | dolphm, #39 is up | 20:51 |
| *** digitalsanctum has quit IRC | 20:51 | |
| *** Ryan_Lane has joined #openstack-dev | 20:52 | |
| *** arbrandes has joined #openstack-dev | 20:52 | |
| *** cp16net is now known as cp16net|away | 20:53 | |
| dolphm | ayoung: thanks | 20:53 |
| *** vkmc has joined #openstack-dev | 20:54 | |
| zykes- | 7win 38 | 20:54 |
| dolphm | ayoung: roles still included in the GET response | 20:58 |
| dolphm | ayoung: GET /trust/{trust_id} | 20:58 |
| ayoung | dolphm, yes | 20:58 |
| dolphm | ayoung: when i said post earlier i just meant that's where i was seeing them | 20:58 |
| dolphm | ayoung: remove them there too :) | 20:58 |
| *** ondergetekende has quit IRC | 20:58 | |
| ayoung | dolphm, I am too tired to argue with you about that, but I think that one is a mistake | 20:58 |
| dolphm | ayoung: and from GET /trusts if they're in there too | 20:58 |
| ayoung | they are gone from list | 20:59 |
| bknudson | anyone seen this doing ./run_tests.sh in keystone: test_admin_requires_adminness (test_keystoneclient.Kc11TestCase) ... error: gnutls_handshake() failed: A TLS packet with unexpected length was received. while accessing https://review.openstack.org/p/openstack/python-keystoneclient.git/info/refs | 20:59 |
| bknudson | fatal: HTTP request failed | 20:59 |
| ayoung | but I think they should be there on GET | 20:59 |
| dolphm | ayoung: it's a collection that should be returned independently | 20:59 |
| ayoung | dolphm, no. They are tightly related. They belong together at that level. Usually, I am all for normalization, but making two web calls to get the rols separate from the body is going to mess people up | 21:00 |
| ayoung | you will alwys end up doing both calls | 21:00 |
| *** cp16net|away is now known as cp16net | 21:01 | |
| ayoung | list, even create, I can understand, although I would tend to say roles should be visible post create. THat is going to mess people up, too. | 21:01 |
| dolphm | ayoung: i almost think there should be a one time create-collection after creating a trust, e.g. POST /trust/{trust_id}/roles | 21:02 |
| zykes- | is v3 support being added to horizon ? | 21:02 |
| ayoung | dolphm, no. It is a single transaction. It feels like you are overthinking this | 21:02 |
| dolphm | "single transaction" not "single request" :) | 21:02 |
| ayoung | dolphm, you don't split up transactions across multiple requests if you can help it. It ties up the database. | 21:04 |
| dolphm | ayoung: it's the same db load either way | 21:04 |
| dolphm | ayoung: technically it's just a relation, right? so PUT /trusts/{trust_id}/roles/{role_id} would be the most appropriate -- but it's difficult to say when you would stop accepting PUT requests (unless you acknowledge that trusts are mutable, but that you can only add to them) | 21:06 |
| ayoung | dolphm, well, not if you start the transaction on one remote request and then complete it on another. You lock the DB for a non-trivial amount of time. But that is ireelevant. Creating a trrust is a singe action, and looking at what roles are assocaited with the trust should also be a single action. List is different, in that you want an abbreviated veiw | 21:06 |
| dolphm | ayoung: i would have requested PUT instead of including roles on POST /trusts except for the immutability gotcha | 21:06 |
| *** esp1 has left #openstack-dev | 21:07 | |
| ayoung | dolphm, I wouldn't code it that way. You want to package up everything for a single RPC that you can, and have it succeed or fail as a unit | 21:07 |
| ayoung | no need for a remote "builder" object | 21:07 |
| *** dprince has joined #openstack-dev | 21:08 | |
| dolphm | ayoung: gate failure on https://review.openstack.org/#/c/23017/ | 21:10 |
| ayoung | dolphm, FAILED horizon | 21:10 |
| dprince | dims: Thanks for the quantum XML DOS fix. Would prefer if you grabbed the ticket first next time though... | 21:11 |
| *** davidh_ has joined #openstack-dev | 21:12 | |
| *** pvo has quit IRC | 21:12 | |
| ayoung | dolphm, looks like there are issues with merge/gats | 21:13 |
| ayoung | gates | 21:13 |
| *** digitalsanctum has joined #openstack-dev | 21:14 | |
| *** pvo has joined #openstack-dev | 21:14 | |
| ayoung | dolphm, so...are we at an impasse? I don't really think we should be removing the roles from things that show the individual trusts. | 21:14 |
| *** lucid- has joined #openstack-dev | 21:14 | |
| dolphm | ayoung: well, you get the client support basically for free if you put them on /roles | 21:15 |
| ayoung | ah... | 21:15 |
| ayoung | not really a good enough reason. | 21:15 |
| dolphm | ayoung: you get the server-side support basically for nearly free on /roles | 21:15 |
| ayoung | dolphm, I'm ok with roles getting their own URL as well | 21:15 |
| dolphm | ayoung: then why implement it twice? | 21:16 |
| dolphm | ayoung: and double the maintenance cost | 21:16 |
| ayoung | dolphm, if we were going to cut one, it would be the /roles. | 21:16 |
| YorikSar | dolphm, ayoung: as I understand, you argue about too detailed response to GET /trust/{id}. I can suggest a compromise - GET /trust/{id}/details | 21:17 |
| ayoung | YorikSar, nah | 21:18 |
| dolphm | ayoung: i don't see any reason to make an exception in the API for anything but create, which provides for the immutability requirement | 21:18 |
| ayoung | YorikSar, the question is what should be on the good name. | 21:18 |
| YorikSar | It's pretty common in REST APIs | 21:18 |
| ayoung | dolphm, there is a difference between listing all of the roles in a system and listing the roles assigned to a trust. FOr a user, the role assignements are not integral, but for a trust they are. | 21:18 |
| *** devoid has quit IRC | 21:19 | |
| YorikSar | ayoung: It's like list of roles in a token, right? | 21:19 |
| *** terry7 has joined #openstack-dev | 21:19 | |
| ayoung | YorikSar, yes, very much so | 21:19 |
| ayoung | YorikSar, in fact, it leads directly to what you would get in a token | 21:20 |
| ladquin | thingee, around? | 21:20 |
| thingee | ladquin: hello | 21:20 |
| YorikSar | ayoung: Then I'm totally on your side. We don't split GET /token/id in pieces, here the reasons are the same. | 21:21 |
| ladquin | thingee, hey! (that was fast) | 21:21 |
| dolphm | ayoung: you can apply the same argument to any other request in the API | 21:21 |
| dolphm | ayoung: i think the correct compromise would be an additional query parameter to have the client opt-in to nesting relational resources into the parent element | 21:22 |
| ayoung | dolphm, no, you can't. It is a question of referential integrity. Tokens and trusts are single, immutable objects. Other collections are malleable. | 21:22 |
| dolphm | ayoung: GET /trusts/{trust_id}?roles for example | 21:22 |
| ayoung | dolphm, no. | 21:22 |
| ayoung | Make the normal path the simple path | 21:22 |
| ladquin | thingee, question: I'm helping review the doc patches for volume-api v2 and got a doubt regarding your pagination comment | 21:22 |
| dolphm | ayoung: GET /users/{user_id}&projects&domains | 21:22 |
| dolphm | ayoung: etc | 21:22 |
| *** jpich has joined #openstack-dev | 21:23 | |
| ayoung | dolphm, the good name is GET /trusts/{trust_id} | 21:23 |
| dolphm | ayoung: they're not a single object at all, they're a relation | 21:23 |
| thingee | ladquin: I believe the permissions make it so it's not viewable. | 21:23 |
| YorikSar | dolphm: Please, don't assume that every backend is relational. | 21:23 |
| thingee | ladquin: it was moved over from v1 and removed from v1 since pagination never existed in that version | 21:24 |
| dolphm | YorikSar: it's relational from an http api perspective -- there's a relation from one api resource to a subset of another collection | 21:24 |
| ayoung | YorikSar, even if the backend is relational, the objects that you fetch out of them do not have to be. | 21:24 |
| ayoung | dolphm, REST does not mean 3rd normal form | 21:24 |
| thingee | ladquin: I can remove it entirely though until it's completed though and doesn't need to be hidden | 21:24 |
| ladquin | thingee, oh, that was my thought | 21:24 |
| YorikSar | dolphm: If we store trusts in memcached (just for example), there will be one object that contains a list of roles among other things. | 21:25 |
| dolphm | ayoung: what's the problem with just having the client opt-in? | 21:25 |
| thingee | ladquin: sounds good | 21:25 |
| ayoung | dolphm, OK, I think I want to revert the roles removal from the create. I will leave the /roles URL in there. | 21:25 |
| ladquin | thingee, but wasn't sure since didn't see it in v1, thanks! | 21:25 |
| thingee | ladquin: do you want me to push that up now or are you still going through it? | 21:25 |
| YorikSar | dolphm: Well, one resource would include a list of roles bydef. | 21:25 |
| ayoung | dolphm, it is how the thing is meant to be used | 21:25 |
| *** topol has quit IRC | 21:25 | |
| dolphm | ayoung: then default the client-side option to true | 21:26 |
| YorikSar | dolphm: Which is not the case for users, for example. | 21:26 |
| vishy | davidkranz: are you still seing this? https://bugs.launchpad.net/nova/+bug/1101504 | 21:26 |
| ayoung | dolphm, the primary consumer is not the python-keystoneclient | 21:26 |
| uvirtbot | Launchpad bug 1101504 in nova "Horrible vm create performance problem with latest devstack" [Undecided,New] | 21:26 |
| dolphm | ayoung: c.trusts.get(trust_id, roles=True) | 21:26 |
| ladquin | thingee, I'm going through it, just helping annegentle with reviews, but there is no rush :) | 21:26 |
| davidkranz | vishy: Checking... | 21:26 |
| *** flaper87 has quit IRC | 21:27 | |
| dolphm | ayoung: so you're saying this shouldn't be a core api then? | 21:27 |
| ayoung | dolphm, this is REST. The primary useage of GET trust should be getting a trust. | 21:27 |
| ayoung | dolphm, once we have endpoints, those should be in there as well. | 21:28 |
| thingee | ladquin: got it. if that's the case I think I'll push the completed section later today then. | 21:28 |
| ayoung | dolphm, the thing is, most people are not going to need to query trusts. People are going to need to create them, and then use them to make tokens. If people are looking at trusts, it is either audit or debugging | 21:28 |
| davidkranz | vishy: I will try it again tomorrow morning. | 21:29 |
| *** esp1 has joined #openstack-dev | 21:29 | |
| ayoung | THere is not need to query the set of roles....and these are really "potentila role assignemnts" but the role entries don;'t have the project ID on them | 21:29 |
| vishy | davidkranz: thanks. I haven't seen the issue before on mine | 21:30 |
| vishy | davidkranz: keep in mind the first boot of a large image will be slow | 21:30 |
| ayoung | henrynash, heckj would appreciate some external input to this discussion. | 21:30 |
| davidkranz | vishy: I know. This was repeatable when I saw it. We'll see. | 21:30 |
| dolphm | ayoung: i don't really follow the notion that the primary consumer of openstack api wouldn't be an openstack client | 21:31 |
| *** clayg has quit IRC | 21:31 | |
| *** ndipanov has quit IRC | 21:31 | |
| ayoung | dolphm, just ask davidkranz ^^ | 21:31 |
| ayoung | Web API is the primary consumer. Not the Python Client | 21:31 |
| dolphm | ayoung: that either means it's not openstack or ... ? i'm lost | 21:31 |
| davidkranz | ayoung: ??? | 21:31 |
| ayoung | davidkranz, question on an API design approach. I am pointing out that the Python client is not the primary api into the system, it is direct web calls | 21:32 |
| ladquin | thingee, cool, I pointed really minor stuff so if you're pushing today you can edit that too, thanks again! | 21:32 |
| davidkranz | ayoung: Well, it depends on what you mean. | 21:33 |
| thingee | ladquin: sounds good. thanks for the help | 21:33 |
| ayoung | then a mix between horizon, with the Python/CLI | 21:33 |
| *** koolhead17 has quit IRC | 21:33 | |
| davidkranz | ayoung: From the user perspective it could be either. | 21:33 |
| ayoung | davidkranz, of course. | 21:33 |
| dolphm | ayoung: so what's wrong with non-openstack client defaulting to the api behavior it wants, i.e., including a nested collection in the response? GET /trusts/{trust_id}&roles | 21:33 |
| davidkranz | ayoung: Why does it matter? | 21:33 |
| ayoung | davidkranz, we are discussing trusts, and whether, when you do | 21:34 |
| ayoung | GET /trusts/{trust_id} | 21:34 |
| *** Ryan_Lane1 has joined #openstack-dev | 21:34 | |
| ayoung | you should get back the roles with it, or whether you should have to make an additional call for the roles collection | 21:34 |
| *** mlavalle has joined #openstack-dev | 21:34 | |
| *** Ryan_Lane has quit IRC | 21:34 | |
| dolphm | (or just have the client opt-in to get them in a single request..) | 21:34 |
| ayoung | personally, I am pretty set that one call should get you the whole trust. I know that you guys are primarily consuming the web APIs as opposed to using the Python client or Horizon, | 21:35 |
| davidkranz | ayoung: I'm not sure but I don't think it matters if the client is using a python API or coding http calls. | 21:35 |
| davidkranz | ayoung: It is a question of how many round trips are needed. | 21:35 |
| dolphm | ayoung: why can't that call be GET /trusts/{trust_id}&roles ? | 21:35 |
| ayoung | dolphm, discoverablility | 21:35 |
| *** johnthetubaguy has joined #openstack-dev | 21:35 | |
| davidkranz | ayoung: vs volume of response which might or might not be needed by the client. | 21:35 |
| heckj | dolphm, ayoung: maybe I'm missing something - why wouldn't you want to get the entire trust at a shot? | 21:36 |
| ayoung | dolphm, you have to know to do that. | 21:36 |
| ayoung | heckj, that is my point | 21:36 |
| *** Ryan_Lane1 is now known as Ryan_Lane | 21:36 | |
| *** Ryan_Lane has joined #openstack-dev | 21:36 | |
| davidkranz | ayoung: A lot of Openstack list calls have a "details" flavor to deal with that. | 21:36 |
| heckj | that's basically what we've done with token (it encapsulates a lot of data), and the call to get a service catalog | 21:36 |
| YorikSar | dolphm: Do we provide the same option for tokens? Since trusts are like unfinished tokens, the analogy is inevidable. The default option is to return all information that is bound to token since this information defines token. Why should we not do the same with trusts? | 21:36 |
| dolphm | heckj: the rest of the api returns flat api objects when you ask for a single resource, and makes the collections available seperately (hopefully through links real soon) | 21:36 |
| dolphm | heckj: (except for the glaring exception of authentication, which we heavily document otherwise) | 21:37 |
| heckj | dolphm: tokens aren't | 21:37 |
| heckj | yeah - | 21:37 |
| heckj | I think trusts rather fits into that same model for what'll be using them | 21:37 |
| ayoung | heckj, agreed | 21:38 |
| dolphm | heckj: we haven't had a use case to nest collections in the parent entity, but now that we have one -- i'm happy to introduce it as an api-wide convention somehow ... my suggestion is simply to include the name of the collection as a query string | 21:38 |
| heckj | they want to get a lot of detail in one shot, and not have to make 2+ requests to get it all to be able to process the usage of the trust | 21:38 |
| heckj | dolphm: ah - I see where you're going now | 21:38 |
| dolphm | heckj: GET /trusts/{trust_id}&roles would return GET /trusts/{trust_id} plus GET /trusts/{trust_id}/roles (which ayoung is asking for) | 21:38 |
| ayoung | dolphm, I would be more prone to put a query string in to get it to shut off the details | 21:38 |
| heckj | dolphm: do we have that pattern elsewhere in OpenStack? | 21:38 |
| ayoung | dolphm, I actually don;t want any thing more than list and get. | 21:39 |
| heckj | I know gabrielhurley was pushing on some of the nested objects and aggregated collections | 21:39 |
| dolphm | heckj: i'm not aware of one, but i'm not intimately familiar with other API's :( | 21:39 |
| YorikSar | ayoung, dolphm: There's not a list of role ids, but a list of whole role refs in trust? | 21:39 |
| dolphm | heckj: yeah, i know i've seen the request | 21:39 |
| ayoung | So /trusts list all of them and /trsuts/{id} gets the indivdual | 21:39 |
| *** numero8 has joined #openstack-dev | 21:39 | |
| dolphm | heckj: i think it'd be awesome to do GET /domains/{domain_id}&projects for example | 21:40 |
| gabrielhurley | yep, all of the above would be lovely | 21:40 |
| ayoung | dolphm, that looks wierd | 21:40 |
| gabrielhurley | and no, that pattern doesn't exist in openstack and it's the bane of my existence | 21:40 |
| ayoung | projects is not a filter on domains | 21:40 |
| heckj | gabrielhurley: more specificity there please :-) | 21:40 |
| dolphm | heckj: although (hrm) we don't do GET /domains/{domain_id}/projects anymore, in favor of GET /projects?domain_id={domain_id} | 21:40 |
| ayoung | it should be | 21:40 |
| heckj | "all of the above" is a bit to vague | 21:40 |
| gabrielhurley | my previous suggestion for what you're discussing now is a "?fetch_related=<depth of JOINs>" | 21:41 |
| ayoung | dolphm, tell you what. Let me check it in as is. We will open a ticket on it, and we can always remove the details if they end up messing people up. My gut says that it won't | 21:41 |
| dolphm | gabrielhurley: eek... what if there are 10 relations, and you only want to get one or two? | 21:42 |
| gabrielhurley | so if you said /domains/?fetch_related=1 it would do a JOIn on the related foreign keys in the DB and return those projects, etc. whatever | 21:42 |
| gabrielhurley | dolphm: we could make it more descriptive | 21:42 |
| gabrielhurley | most ORMs have this feature | 21:42 |
| gabrielhurley | it's just a matter of describing it adequately in a query string | 21:42 |
| gabrielhurley | it's not an easy feature, but it's sorely needed to stop doing stupid numbers of queries | 21:43 |
| YorikSar | gabrielhurley: Not every storage is relational. | 21:43 |
| heckj | ayoung: I'm fine with that | 21:43 |
| *** esp1 has left #openstack-dev | 21:43 | |
| dolphm | heckj: i'm not | 21:43 |
| gabrielhurley | YorikSar: obviously not, but a lot of non-rel mimics JOINs and a lot of ORMs paper over that difference. | 21:43 |
| heckj | clearly :-) | 21:43 |
| gabrielhurley | YorikSar: and if they don't, then maybe the feeature doesn't work. | 21:44 |
| ayoung | We are not building for automated systems to consume. This is not an RPC. It is a REST API. The difference is that a REST API is designed to be consumed by Humans. THere is a lot of work we need to do to get there. | 21:44 |
| gabrielhurley | ayoung: so very very true | 21:44 |
| YorikSar | gabrielhurley: I'm not talking about NoSQL stuff. I'm talking about LDAP, for example. | 21:44 |
| dolphm | heckj: that just introduces an undocumented api in core that we then have to go back and document because it's suddenly being consumed client-side | 21:44 |
| gabrielhurley | YorikSar: ::shrug:: my point is that differing backends shouldn't stop a much-needed feature. The differences can be worked out in the abstraction layer. | 21:44 |
| dolphm | heckj: and then we have to support it | 21:44 |
| dolphm | heckj: so all that does is bypass the discussion | 21:45 |
| heckj | dolphm: yep, I understand your concern there | 21:45 |
| *** koolhead17 has joined #openstack-dev | 21:45 | |
| heckj | ayoung: leaving the API discussion to so late in the game wasn't… the best move here | 21:45 |
| ayoung | dolphm, Lest not be foolishly consistent here. Trusts and tokens are small trees of small objects that all belong and get consumed together. The are not meant to be managed like collections, but rather as discrete units. It is a very important pattern to support. | 21:45 |
| dolphm | nope | 21:45 |
| ayoung | heckj, it was never designed like that | 21:45 |
| ayoung | heckj, I wrote up the API back in October. | 21:45 |
| YorikSar | gabrielhurley: It's not so easy to make that JOIN on backend level. And it's not so good to force backend to implement complex logic for really rare queries. | 21:46 |
| ayoung | It has evolved since then, but this is all very recent stuff to this API | 21:46 |
| dolphm | ayoung: what's the problem with hardcoding your secret api client to always request the response you want? | 21:46 |
| ayoung | the word secret | 21:46 |
| *** yolanda has quit IRC | 21:46 | |
| ayoung | dolphm, its like needing to know up up down down left right left right BA Start | 21:46 |
| dims | dprince, sorry. will ask/grab ticket next time | 21:46 |
| gabrielhurley | YorikSar: I have no interest in arguing LDAP's particularities. As someone building human-readable interfaces this is a feature I need across every OpenStack API. We can argue this particular one some day in the future. | 21:47 |
| ayoung | gabrielhurley, since YorikSar has bascially agreed with me, I think he is now just being technical | 21:47 |
| *** annegentle has joined #openstack-dev | 21:48 | |
| ayoung | dolphm, why do you feel the need to split up the trusts into multiple remote calls? What does that gain you? | 21:48 |
| dolphm | ayoung: i don't even know what that question means | 21:48 |
| dolphm | ayoung: remote between what and what? | 21:49 |
| ayoung | dolphm, I am turning the discussion around. To me it makes sense to grab a whole trust at once. Why would I want to do otherwise? | 21:49 |
| ayoung | between the client, of whatever sort, and the web server | 21:49 |
| dolphm | ayoung: because you're definition of "trust" defies the conventions set forth by the entire rest of the api in that relations are exposed seperately | 21:50 |
| dolphm | your* | 21:50 |
| ayoung | dolphm, no it does not | 21:50 |
| *** rkukura has quit IRC | 21:50 | |
| ayoung | trusts and tokens are different from users, roles, and projects | 21:50 |
| dolphm | ayoung: if you have an example, i'm happy to file a bug and correct it | 21:50 |
| ayoung | dolphm, when you upload a virtual machine to glance, you upload it all at once. | 21:50 |
| dolphm | ayoung: (side note -- i have to log in about 5 minutes, will be back on in an hour or two) | 21:50 |
| ayoung | dolphm, I am tired. I am probably not going to be awake much longer. I have to disappear as well. | 21:51 |
| dolphm | ayoung: sleep! | 21:51 |
| YorikSar | dolphm: In current patchset Trust.get_trust returns only a list of role_ids, not the roles themselves. So it's exactly what trust consists of. GET /trusts/{is}/roles cat return a list of roles' refs. It looks like pure REST. | 21:52 |
| ayoung | dolphm, A have been given a deadline of today. I am trying to meet my obligations. I have been postponing a lot of other work to knock this out, and I want it closed. We should be able to resolve this. | 21:52 |
| dolphm | ayoung: you don't like my compromise and can't justify why :( | 21:53 |
| ayoung | dolphm, I can justfy why the normalized approach is wrong. When you do a web get, you get adocument, this is a tree, a hierarchy. To force web calls to be strictly single collections is limiting the power of the medium. It is an artificial constraint, and we should not choses to do that to ourselves. | 21:54 |
| ayoung | dolphm, if you look at the other openstack components, I doubt that the APIs are designed that way | 21:54 |
| ayoung | even the ones that are not poorly designed. | 21:55 |
| ayoung | dolphm, It I wanted to get a definition of a virtual machine from Nova, I would want a document that describes the whole vm | 21:55 |
| ayoung | libvirt does just that, actually | 21:55 |
| ayoung | I don;t want to get the cpus seperate from the memory from the nics etc | 21:56 |
| ayoung | that way leads to madness | 21:56 |
| ayoung | human sacrifice, dogs and cats, living together, mass hysteria | 21:56 |
| *** esp1 has joined #openstack-dev | 21:56 | |
| dolphm | ayoung: i'd like to support things like paging on collections -- how do we do that on a nested collection? | 21:56 |
| BLZbubba | what can cause a majority of my openstack guests to go into a "paused" state? | 21:56 |
| ayoung | dolphm, trust will not need that | 21:56 |
| ayoung | listing trusts will | 21:56 |
| *** esp1 has left #openstack-dev | 21:56 | |
| dolphm | ayoung: if a trust grants 10,000 roles, that might suck to get in a single response | 21:56 |
| ayoung | but the individual trusts will not\ | 21:57 |
| ayoung | dolphm, and that person deserves what they get | 21:57 |
| ayoung | a trust like that is a blight | 21:57 |
| dolphm | ayoung: what if all they want to know is the expiration of the trust? | 21:57 |
| ayoung | premature optimization | 21:57 |
| dolphm | ayoung: or whether impersonation is enabled? | 21:57 |
| dolphm | ayoung: i'm not asking to implement paging -- just to have an idea about how to support it | 21:58 |
| *** kagan has joined #openstack-dev | 21:58 | |
| ayoung | dolphm, should we make separate API calls just for the individual attributes? | 21:58 |
| READ10 | BLZbubba: you out of disk space maybe? | 21:58 |
| ayoung | Pretty close we will havereimplemented SQL | 21:58 |
| dolphm | ayoung: i've got to run -- will plan to be back on in < 2 hours | 21:58 |
| dolphm | ayoung: lol | 21:58 |
| dolphm | ayoung: that's not funny because there were a couple calls like that in diablo :P | 21:59 |
| *** dolphm has quit IRC | 21:59 | |
| *** cloudchimp has quit IRC | 22:00 | |
| *** numero8 has quit IRC | 22:00 | |
| *** henrynash_ has joined #openstack-dev | 22:00 | |
| *** martine has quit IRC | 22:01 | |
| BLZbubba | READ10: nope | 22:01 |
| *** turul_ has quit IRC | 22:02 | |
| *** afazekas has quit IRC | 22:02 | |
| *** henrynash has quit IRC | 22:02 | |
| *** henrynash_ is now known as henrynash | 22:02 | |
| YorikSar | ayoung: Why are list_* like this in SQL backend? | 22:03 |
| READ10 | BLZbubba: hmm, I'm not sure then; whenever anybody says 'lots of VMs paused' I think i/o issues, which often means out of disk space | 22:03 |
| YorikSar | ayoung: Is it critical to call session.flush before to_dict? | 22:03 |
| ayoung | YorikSar, I don't think so. | 22:03 |
| *** clayg has joined #openstack-dev | 22:03 | |
| *** lorin1 has left #openstack-dev | 22:03 | |
| ayoung | But it probably is more correct to do so | 22:03 |
| jab416171 | READ10: 133 out of 298 VMs are paused | 22:04 |
| jab416171 | spread across 9 compute nodes | 22:04 |
| *** arbrandes has quit IRC | 22:04 | |
| *** heckj has quit IRC | 22:06 | |
| *** kazoo has joined #openstack-dev | 22:06 | |
| *** ayoung is now known as ayoung-afk | 22:07 | |
| READ10 | jab416171: I couldn't say what's happening there, I just thought I'd throw the I/O idea out there to make sure somebody mentioned it | 22:07 |
| *** markmc has quit IRC | 22:08 | |
| READ10 | jab416171: actually, though, KVM VMs? | 22:08 |
| jab416171 | yeah | 22:08 |
| READ10 | ok, then I stand by the comment | 22:08 |
| READ10 | I'm not sure what other HVs do | 22:09 |
| BLZbubba | READ10: so virsh will actually mark vm's paused if the disk is full? | 22:09 |
| BLZbubba | not a bad way to handle it | 22:09 |
| READ10 | generally qemu will pause the VM on io failure, although that can be overridden | 22:09 |
| READ10 | BLZbubba: yeah, it's very good in the disk full case | 22:10 |
| *** dprince has quit IRC | 22:10 | |
| *** kazoo has quit IRC | 22:10 | |
| *** lcheng has quit IRC | 22:10 | |
| *** pabelanger has quit IRC | 22:11 | |
| BLZbubba | is there a way to have it clean up the instance data for deleted instances? | 22:11 |
| *** dhellmann has quit IRC | 22:12 | |
| *** johnthetubaguy has quit IRC | 22:13 | |
| *** pnavarro has quit IRC | 22:14 | |
| READ10 | BLZbubba: I have to admit I've been using KVM for a few years, but I'm just getting started with openstack, so I don't know the answer to that | 22:14 |
| *** zodiak has quit IRC | 22:14 | |
| BLZbubba | same here | 22:14 |
| *** pixelbeat has quit IRC | 22:14 | |
| BLZbubba | it won't take more than a few minutes to script it up though | 22:15 |
| BLZbubba | the most painful part is having to move between uuid and hex(id) all the time | 22:16 |
| *** pcm_ has quit IRC | 22:18 | |
| *** gasbakid_ has quit IRC | 22:18 | |
| *** zodiak has joined #openstack-dev | 22:18 | |
| *** gabrielhurley has quit IRC | 22:19 | |
| *** annegentle has quit IRC | 22:20 | |
| *** digitals_ has joined #openstack-dev | 22:22 | |
| *** kgriffs has left #openstack-dev | 22:24 | |
| *** digitalsanctum has quit IRC | 22:26 | |
| *** mrunge has quit IRC | 22:26 | |
| *** bswartz has quit IRC | 22:29 | |
| *** danwent has quit IRC | 22:29 | |
| *** annegentle has joined #openstack-dev | 22:29 | |
| *** edmund has quit IRC | 22:31 | |
| *** anniec has joined #openstack-dev | 22:32 | |
| *** anniec_ has quit IRC | 22:32 | |
| *** digitals_ has quit IRC | 22:37 | |
| *** alunduil has quit IRC | 22:37 | |
| *** mtreinish has quit IRC | 22:37 | |
| *** markvoelker has quit IRC | 22:38 | |
| *** diogogmt has quit IRC | 22:40 | |
| *** nunosantos has quit IRC | 22:41 | |
| *** mohits has joined #openstack-dev | 22:44 | |
| *** radez is now known as radez_g0n3 | 22:45 | |
| *** mohits has quit IRC | 22:48 | |
| *** digitalsanctum has joined #openstack-dev | 22:50 | |
| *** dosaboy has quit IRC | 22:54 | |
| *** sacharya has quit IRC | 22:55 | |
| *** armaan has left #openstack-dev | 22:56 | |
| *** yamahata has quit IRC | 22:56 | |
| iccha | hey anyone having problem with glance-api restart and oslo.config import error? | 22:56 |
| *** dubsquared1 has joined #openstack-dev | 22:59 | |
| *** stevemar has quit IRC | 23:02 | |
| *** john5223 has quit IRC | 23:03 | |
| *** digitalsanctum has quit IRC | 23:03 | |
| *** eglynn has quit IRC | 23:04 | |
| *** tomoe_ has joined #openstack-dev | 23:05 | |
| *** alexpilotti has joined #openstack-dev | 23:05 | |
| *** kbringard has quit IRC | 23:05 | |
| *** datsun180b has quit IRC | 23:05 | |
| *** kbringard has joined #openstack-dev | 23:06 | |
| *** dolphm has joined #openstack-dev | 23:10 | |
| *** bknudson has quit IRC | 23:10 | |
| *** shardy is now known as shardy_afk | 23:11 | |
| *** eharney has quit IRC | 23:13 | |
| *** devoid has joined #openstack-dev | 23:13 | |
| *** dolphm has quit IRC | 23:14 | |
| *** jpich has quit IRC | 23:18 | |
| *** digitalsanctum has joined #openstack-dev | 23:20 | |
| *** annegentle has quit IRC | 23:20 | |
| *** esheffield has quit IRC | 23:22 | |
| *** hub_cap has left #openstack-dev | 23:23 | |
| *** Gordonz has quit IRC | 23:24 | |
| *** brianr_ is now known as brianr-away | 23:26 | |
| *** epim has quit IRC | 23:31 | |
| *** AlanClark has quit IRC | 23:31 | |
| *** gordc has quit IRC | 23:37 | |
| *** alunduil has joined #openstack-dev | 23:39 | |
| *** danwent has joined #openstack-dev | 23:39 | |
| *** gongysh has joined #openstack-dev | 23:42 | |
| *** soody has joined #openstack-dev | 23:42 | |
| *** olaph_ has quit IRC | 23:42 | |
| *** annegentle has joined #openstack-dev | 23:45 | |
| *** ritzcarltn has joined #openstack-dev | 23:46 | |
| *** ijw has quit IRC | 23:48 | |
| *** jimfehlig has quit IRC | 23:49 | |
| *** devoid has quit IRC | 23:50 | |
| *** ijw has joined #openstack-dev | 23:50 | |
| *** NobodyCam has quit IRC | 23:51 | |
| *** rpedde is now known as rpedde_away | 23:52 | |
| *** andrewbogott is now known as andrewbogott_afk | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!