| *** vipul|away is now known as vipul | 00:02 | |
| arata | lifeless: well, maybe OK. but I need check other members' opinions. | 00:03 |
|---|---|---|
| arata | lifeless: We will answer to you soon... | 00:04 |
| vishy | otherwiseguy: wow the output of the test is really interesting | 00:05 |
| vishy | otherwiseguy: it looks like wait isn't actually waiting | 00:05 |
| vishy | otherwiseguy: in the third case the returned value comes back before the error message | 00:06 |
| *** lloydde has quit IRC | 00:06 | |
| otherwiseguy | vishy: indeed. | 00:06 |
| *** torandu has quit IRC | 00:06 | |
| vishy | otherwiseguy: putting a time.sleep(0.1) between the Popen and teh wait fixes it | 00:07 |
| openstackgerrit | A change was merged to openstack/swift: use Host: for location rewrites https://review.openstack.org/15681 | 00:09 |
| *** sniperd has quit IRC | 00:10 | |
| *** gongysh has joined #openstack-dev | 00:10 | |
| otherwiseguy | vishy: and using eventlet.green's subprocess | 00:10 |
| * otherwiseguy glares at eventlet | 00:10 | |
| russellb | monkey_patch() doesn't make it do that already? | 00:11 |
| russellb | (use eventlet's subprocess) | 00:11 |
| *** jimfehlig has joined #openstack-dev | 00:11 | |
| networkstatic | subprocess.Popen | 00:11 |
| *** reed has quit IRC | 00:11 | |
| networkstatic | x.wait() is win | 00:11 |
| lifeless | arata: thanks | 00:11 |
| *** sniperd has joined #openstack-dev | 00:11 | |
| *** jcru|away is now known as jcru | 00:12 | |
| *** samkottler|food is now known as samkottler | 00:13 | |
| *** alexpilotti has quit IRC | 00:13 | |
| lifeless | arata: I will put a patch together that documents how it might all work, and as much other stuff as happens to fit, and make a pull request, for you to assess. | 00:14 |
| *** alexpilotti has joined #openstack-dev | 00:14 | |
| otherwiseguy | russellb: looks like monkey_patch just does specific modules, of which subprocess is not one | 00:15 |
| * otherwiseguy is not an eventlet guru | 00:15 | |
| *** s0mik has joined #openstack-dev | 00:16 | |
| *** Sparks has quit IRC | 00:16 | |
| *** alexpilotti has quit IRC | 00:19 | |
| vishy | otherwiseguy: got a fix | 00:19 |
| vishy | otherwiseguy: need to verify that it is correct | 00:19 |
| *** maurosr has quit IRC | 00:20 | |
| otherwiseguy | vishy: woot | 00:20 |
| *** sunilthaha has joined #openstack-dev | 00:21 | |
| vishy | otherwiseguy: so somehow waitpid is being called before the process has actually been created | 00:22 |
| vishy | otherwiseguy: so I have a hacky workaround looking for the right one | 00:23 |
| lifeless | vishy: whaaaat? Don't you need the pid for that ? | 00:23 |
| lifeless | vishy: or do you mean 'before the child has called exec()' ? | 00:23 |
| dansmith | lifeless: you can waitpid(-1) right? | 00:24 |
| *** lloydde has joined #openstack-dev | 00:25 | |
| lifeless | dansmith: thats a bad idea though, because you find out about any child | 00:26 |
| *** jcru has quit IRC | 00:26 | |
| *** arosen has joined #openstack-dev | 00:26 | |
| *** jakedahn is now known as jakedahn_zz | 00:26 | |
| dansmith | lifeless: well, depends on what you're aiming for :) | 00:26 |
| lifeless | dansmith: unless the code structure is ready to deal with that (e.g. has some way of calling back to the right handler for an arbitrary exited child) | 00:26 |
| dansmith | lifeless: well, you get the pid back of the exited child... | 00:27 |
| lifeless | you do | 00:27 |
| dansmith | I'm just saying, lots of things waitpid(-1) ;) | 00:28 |
| *** markmcclain has joined #openstack-dev | 00:28 | |
| dansmith | I wasn't paying attention to the previous conversation though, so I should shut up :) | 00:28 |
| vishy | otherwiseguy: actually i think my fix is correct | 00:29 |
| otherwiseguy | vishy: then I shall say woot again. | 00:30 |
| vishy | otherwiseguy: fix is easy, I will get a diff | 00:30 |
| otherwiseguy | k | 00:30 |
| *** winston-d has joined #openstack-dev | 00:31 | |
| *** maxiz has joined #openstack-dev | 00:33 | |
| *** s0mik has quit IRC | 00:36 | |
| *** alunduil has joined #openstack-dev | 00:37 | |
| vishy | otherwiseguy: http://paste.openstack.org/show/25138/ | 00:38 |
| *** dims has quit IRC | 00:39 | |
| *** lloydde has quit IRC | 00:40 | |
| vishy | otherwiseguy: here is the issue: https://bitbucket.org/which_linden/eventlet/issue/92/eventletgreen-override-of-oswaitpid | 00:40 |
| *** sniperd has quit IRC | 00:40 | |
| otherwiseguy | vishy: huh. that's been open quite a while. | 00:42 |
| *** samkottler is now known as samkottler|out | 00:43 | |
| *** torandu has joined #openstack-dev | 00:43 | |
| *** tgall_foo has joined #openstack-dev | 00:44 | |
| *** patelna has quit IRC | 00:45 | |
| *** metral has quit IRC | 00:45 | |
| *** metral has joined #openstack-dev | 00:45 | |
| *** s0mik has joined #openstack-dev | 00:45 | |
| vishy | jerdfelt: ping any thoughts on testing https://bitbucket.org/which_linden/eventlet/pull-request/24/fix-waitpid-returning-0-0/diff ? | 00:50 |
| *** dims has joined #openstack-dev | 00:53 | |
| *** eglynn_ has joined #openstack-dev | 00:54 | |
| *** s0mik has quit IRC | 00:54 | |
| *** saurabhsurana_ has left #openstack-dev | 00:55 | |
| *** eglynn has quit IRC | 00:57 | |
| otherwiseguy | vishy: I can verify that the issue with openssl zombies from quantum ( https://bugs.launchpad.net/quantum/+bug/1074257 ) goes away with the patch applied. | 00:57 |
| uvirtbot` | Launchpad bug 1074257 in quantum "quantum-server creates openssl zombies until process limit is reached" [Critical,In progress] | 00:57 |
| vishy | otherwiseguy: nifty | 00:57 |
| otherwiseguy | and now...dinner. | 00:58 |
| *** FlaPer87 has quit IRC | 00:58 | |
| *** metral_ has joined #openstack-dev | 00:59 | |
| vishy | jgriffith: real fix for eventlet issue: tests/greenio_test.py | 00:59 |
| vishy | jgriffith: err… https://bitbucket.org/which_linden/eventlet/pull-request/24/fix-waitpid-returning-0-0/diff | 00:59 |
| *** s0mik has joined #openstack-dev | 01:00 | |
| *** colinmcnamara has joined #openstack-dev | 01:00 | |
| *** metral has quit IRC | 01:02 | |
| *** metral_ is now known as metral | 01:02 | |
| *** otherwiseguy has quit IRC | 01:05 | |
| *** colinmcnamara has quit IRC | 01:05 | |
| *** vipul is now known as vipul|away | 01:06 | |
| *** s0mik has joined #openstack-dev | 01:11 | |
| *** hemna_ has joined #openstack-dev | 01:12 | |
| *** juice has quit IRC | 01:13 | |
| *** tgall_foo has quit IRC | 01:13 | |
| *** hemna has quit IRC | 01:16 | |
| *** PaulM has left #openstack-dev | 01:17 | |
| *** sacharya has joined #openstack-dev | 01:19 | |
| *** epim has quit IRC | 01:22 | |
| *** sniperd has joined #openstack-dev | 01:24 | |
| *** vipul|away is now known as vipul | 01:30 | |
| *** jcmartin has quit IRC | 01:32 | |
| *** danwent has quit IRC | 01:32 | |
| *** danwent has joined #openstack-dev | 01:33 | |
| *** dims has quit IRC | 01:33 | |
| *** esp2 has left #openstack-dev | 01:33 | |
| *** adalbas has quit IRC | 01:35 | |
| openstackgerrit | A change was merged to openstack/keystone: Fix default port for identity.internalURL https://review.openstack.org/15678 | 01:38 |
| openstackgerrit | A change was merged to openstack/horizon: Changed "Delete volume" success message text https://review.openstack.org/15633 | 01:38 |
| openstackgerrit | A change was merged to openstack/horizon: When editing an image, correctly set the 'is_public' checkbox. https://review.openstack.org/15646 | 01:38 |
| *** littleidea has quit IRC | 01:39 | |
| *** littleidea has joined #openstack-dev | 01:40 | |
| *** roge has quit IRC | 01:41 | |
| *** metral_ has joined #openstack-dev | 01:41 | |
| *** zaitcev has quit IRC | 01:42 | |
| *** dims has joined #openstack-dev | 01:42 | |
| *** nati_ueno has quit IRC | 01:44 | |
| *** metral has quit IRC | 01:45 | |
| *** metral_ is now known as metral | 01:45 | |
| *** nati_ueno has joined #openstack-dev | 01:45 | |
| *** maxiz has quit IRC | 01:49 | |
| *** lloydde has joined #openstack-dev | 01:50 | |
| *** thingee has quit IRC | 01:52 | |
| *** sacharya1 has joined #openstack-dev | 01:52 | |
| *** sacharya has quit IRC | 01:54 | |
| *** mrodden has joined #openstack-dev | 01:55 | |
| *** littleidea has quit IRC | 01:55 | |
| *** lloydde has quit IRC | 01:55 | |
| *** ZenFag has joined #openstack-dev | 01:56 | |
| *** metral_ has joined #openstack-dev | 01:56 | |
| *** arata has quit IRC | 01:57 | |
| *** ZenFag has quit IRC | 01:57 | |
| *** ZenFag has joined #openstack-dev | 01:58 | |
| *** metral has quit IRC | 01:59 | |
| *** metral_ is now known as metral | 01:59 | |
| *** ZenFag is now known as umadbro | 02:00 | |
| *** umadbro is now known as umad | 02:01 | |
| *** umad has quit IRC | 02:01 | |
| *** problem has joined #openstack-dev | 02:02 | |
| *** problem has left #openstack-dev | 02:02 | |
| *** Sparks has joined #openstack-dev | 02:06 | |
| *** pixelbeat has quit IRC | 02:06 | |
| *** jog0 has quit IRC | 02:08 | |
| *** littleidea has joined #openstack-dev | 02:09 | |
| *** nati_uen_ has quit IRC | 02:10 | |
| *** littleidea has left #openstack-dev | 02:10 | |
| *** roge has joined #openstack-dev | 02:11 | |
| *** gyee has quit IRC | 02:16 | |
| *** yidclare has joined #openstack-dev | 02:16 | |
| *** sniperd has quit IRC | 02:16 | |
| jgriffith | vishy: yeah, saw your request on bit-bucket | 02:20 |
| *** markmcclain has quit IRC | 02:21 | |
| *** epim has joined #openstack-dev | 02:25 | |
| *** epim_ has joined #openstack-dev | 02:25 | |
| *** Mandell has quit IRC | 02:26 | |
| *** crandquist has joined #openstack-dev | 02:26 | |
| openstackgerrit | A change was merged to openstack/keystone: Improve feedback on test failure https://review.openstack.org/15672 | 02:28 |
| *** kobier has quit IRC | 02:28 | |
| *** epim has quit IRC | 02:29 | |
| *** epim_ is now known as epim | 02:29 | |
| *** Sparks has quit IRC | 02:31 | |
| *** stevebake has joined #openstack-dev | 02:34 | |
| *** danwent has quit IRC | 02:35 | |
| *** nati_ueno has quit IRC | 02:35 | |
| *** davidhadas has quit IRC | 02:40 | |
| *** SlickNik has quit IRC | 02:42 | |
| *** cp16net|away is now known as cp16net | 02:44 | |
| *** salv-orlando has quit IRC | 02:46 | |
| gongysh | Akihiro Motoki? | 02:47 |
| *** zbitter has joined #openstack-dev | 02:48 | |
| *** Gordonz has joined #openstack-dev | 02:49 | |
| *** matiu has joined #openstack-dev | 02:50 | |
| *** davidha has joined #openstack-dev | 02:51 | |
| *** zaneb has quit IRC | 02:52 | |
| openstackgerrit | A change was merged to openstack/nova: make diagnostics workable for libvirt with Xen https://review.openstack.org/15662 | 02:53 |
| *** adjohn has quit IRC | 02:54 | |
| *** markmcclain has joined #openstack-dev | 02:55 | |
| *** s0mik has quit IRC | 02:55 | |
| *** s0mik has joined #openstack-dev | 02:56 | |
| *** mrodden has quit IRC | 02:56 | |
| *** mnewby has quit IRC | 02:58 | |
| *** Ruetobas has quit IRC | 02:58 | |
| gongysh | zhong yue: ping | 02:59 |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: removing repeat attempt at authorization in client https://review.openstack.org/14730 | 02:59 |
| *** sunilthaha has quit IRC | 02:59 | |
| *** novas0x2a|laptop has quit IRC | 03:00 | |
| *** spiffxp has quit IRC | 03:00 | |
| *** Ruetobas has joined #openstack-dev | 03:02 | |
| *** cp16net is now known as cp16net|away | 03:03 | |
| *** yidclare has quit IRC | 03:05 | |
| *** torandu has quit IRC | 03:07 | |
| *** sandywalsh has quit IRC | 03:12 | |
| *** anniec has quit IRC | 03:17 | |
| gongysh | garyk: ping | 03:19 |
| *** metral_ has joined #openstack-dev | 03:20 | |
| *** metral has quit IRC | 03:22 | |
| *** metral_ is now known as metral | 03:22 | |
| *** metral_ has joined #openstack-dev | 03:24 | |
| *** metral has quit IRC | 03:28 | |
| *** metral_ is now known as metral | 03:28 | |
| *** lifeless has quit IRC | 03:29 | |
| *** maxiz has joined #openstack-dev | 03:32 | |
| *** jimfehlig has quit IRC | 03:33 | |
| *** sacharya has joined #openstack-dev | 03:38 | |
| *** zyluo has joined #openstack-dev | 03:39 | |
| *** sacharya1 has quit IRC | 03:39 | |
| *** ndipanov has quit IRC | 03:40 | |
| *** Gordonz has quit IRC | 03:41 | |
| *** mjfork has quit IRC | 03:47 | |
| *** ndipanov has joined #openstack-dev | 03:52 | |
| *** eglynn__ has joined #openstack-dev | 03:53 | |
| *** melwitt has quit IRC | 03:54 | |
| *** eglynn_ has quit IRC | 03:56 | |
| *** cp16net|away is now known as cp16net | 03:56 | |
| *** littleidea has joined #openstack-dev | 03:57 | |
| *** sunilthaha has joined #openstack-dev | 03:59 | |
| *** sacharya has quit IRC | 04:04 | |
| *** sacharya has joined #openstack-dev | 04:05 | |
| *** fungi has quit IRC | 04:07 | |
| *** dhellmann has joined #openstack-dev | 04:13 | |
| *** nati_ueno has joined #openstack-dev | 04:14 | |
| *** lifeless has joined #openstack-dev | 04:14 | |
| *** Mandell has joined #openstack-dev | 04:15 | |
| openstackgerrit | A change was merged to openstack/quantum: Fix openssl zombies https://review.openstack.org/15645 | 04:24 |
| *** thingee has joined #openstack-dev | 04:25 | |
| *** sacharya1 has joined #openstack-dev | 04:25 | |
| *** nati_ueno_2 has joined #openstack-dev | 04:27 | |
| *** nati_ueno has quit IRC | 04:27 | |
| *** sacharya has quit IRC | 04:28 | |
| *** sacharya1 has quit IRC | 04:32 | |
| *** sacharya has joined #openstack-dev | 04:33 | |
| *** sacharya has quit IRC | 04:33 | |
| openstackgerrit | A change was merged to openstack/glance: Unit test remaining glance-replicator methods. https://review.openstack.org/15546 | 04:35 |
| *** markwash has quit IRC | 04:49 | |
| *** aeperezt has quit IRC | 04:54 | |
| *** roge has quit IRC | 05:01 | |
| *** asalkeld has quit IRC | 05:13 | |
| *** littleidea has quit IRC | 05:17 | |
| *** markmcclain has quit IRC | 05:18 | |
| *** markwash has joined #openstack-dev | 05:25 | |
| *** adjohn has joined #openstack-dev | 05:30 | |
| *** tgall_foo has joined #openstack-dev | 05:36 | |
| *** spiffxp has joined #openstack-dev | 05:40 | |
| *** s0mik has quit IRC | 06:00 | |
| *** andrewbogott is now known as andrewbogott_afk | 06:01 | |
| *** mnewby has joined #openstack-dev | 06:03 | |
| *** fungi has joined #openstack-dev | 06:08 | |
| openstackgerrit | A change was merged to openstack/cinder: Scheduler API clean up and refactor https://review.openstack.org/14131 | 06:11 |
| *** koolhead17 has joined #openstack-dev | 06:19 | |
| *** zyluo has quit IRC | 06:22 | |
| *** e1mer has quit IRC | 06:25 | |
| *** spiffxp has quit IRC | 06:25 | |
| *** lianhao has quit IRC | 06:27 | |
| *** tgall_foo has quit IRC | 06:28 | |
| *** tgall_foo has joined #openstack-dev | 06:29 | |
| *** e1mer has joined #openstack-dev | 06:31 | |
| *** colinmcnamara has joined #openstack-dev | 06:32 | |
| *** PaulM has joined #openstack-dev | 06:38 | |
| *** armaan has joined #openstack-dev | 06:38 | |
| *** asalkeld has joined #openstack-dev | 06:39 | |
| *** Mandell has quit IRC | 06:49 | |
| *** mrunge has joined #openstack-dev | 06:50 | |
| *** koolhead17 has quit IRC | 06:57 | |
| *** armaan has left #openstack-dev | 06:58 | |
| *** nati_ueno_2 has quit IRC | 06:58 | |
| *** e1mer has quit IRC | 07:01 | |
| *** rushiagr has joined #openstack-dev | 07:03 | |
| *** digantya has joined #openstack-dev | 07:06 | |
| *** afazekas has joined #openstack-dev | 07:06 | |
| *** thovden has joined #openstack-dev | 07:08 | |
| *** adjohn has quit IRC | 07:11 | |
| *** colinmcnamara has quit IRC | 07:14 | |
| *** reed has joined #openstack-dev | 07:16 | |
| *** Vek has quit IRC | 07:17 | |
| openstackgerrit | A change was merged to openstack/quantum: Clean up executable modules https://review.openstack.org/15609 | 07:28 |
| openstackgerrit | A change was merged to openstack/quantum: Ensure that the anyjson version is correct. https://review.openstack.org/15560 | 07:29 |
| *** thovden has quit IRC | 07:30 | |
| *** saurabh has joined #openstack-dev | 07:35 | |
| *** gongysh has quit IRC | 07:35 | |
| *** e1mer has joined #openstack-dev | 07:38 | |
| saurabh | anyones got any idea how i'd run the quantum/tests/unit/test_extensions.py file? I want to use extensions for my plugin and feel that running the test_extensions.py is my only hope of understanding the flow | 07:38 |
| *** markwash has quit IRC | 07:45 | |
| *** cmagina has quit IRC | 07:49 | |
| *** ndipanov has quit IRC | 07:49 | |
| *** e1mer has quit IRC | 07:52 | |
| *** Mandell has joined #openstack-dev | 07:53 | |
| *** reidrac has joined #openstack-dev | 07:55 | |
| *** e1mer has joined #openstack-dev | 07:58 | |
| *** mrunge has quit IRC | 08:00 | |
| *** nunosantos has quit IRC | 08:02 | |
| *** ndipanov has joined #openstack-dev | 08:02 | |
| *** FlaPer87 has joined #openstack-dev | 08:04 | |
| *** mrunge has joined #openstack-dev | 08:05 | |
| rushiagr | just run 'run_tests.sh test_extensions' | 08:05 |
| *** ondergetekende has joined #openstack-dev | 08:09 | |
| *** jdurgin has quit IRC | 08:09 | |
| *** e1mer has quit IRC | 08:09 | |
| *** jdurgin has joined #openstack-dev | 08:11 | |
| *** e1mer has joined #openstack-dev | 08:13 | |
| *** mindpixel has joined #openstack-dev | 08:15 | |
| *** cmagina has joined #openstack-dev | 08:16 | |
| *** armaan has joined #openstack-dev | 08:17 | |
| *** jpich has joined #openstack-dev | 08:24 | |
| *** davidha has quit IRC | 08:28 | |
| digantya | hi rushiagr: | 08:32 |
| *** salv-orlando has joined #openstack-dev | 08:34 | |
| *** Mandell has quit IRC | 08:40 | |
| *** egallen has joined #openstack-dev | 08:41 | |
| *** yjiang5 has quit IRC | 08:46 | |
| *** yjiang5 has joined #openstack-dev | 08:49 | |
| *** PaulM has quit IRC | 08:54 | |
| *** PaulM has joined #openstack-dev | 08:55 | |
| rushiagr | digantya: hi | 08:55 |
| *** winston-d has quit IRC | 08:56 | |
| digantya | now i am using ubuntu , can u tell me the procedure to setup and test the Nova project to fix bugs ? | 08:57 |
| digantya | rushiagr: now i am using ubuntu , can u tell me the procedure to setup and test the Nova project to fix bugs ? | 08:58 |
| *** maxiz has quit IRC | 09:05 | |
| *** aloga has quit IRC | 09:07 | |
| *** aloga has joined #openstack-dev | 09:09 | |
| *** PaulM has quit IRC | 09:09 | |
| *** pixelbeat has joined #openstack-dev | 09:11 | |
| *** Vek has joined #openstack-dev | 09:13 | |
| *** rafaduran has joined #openstack-dev | 09:14 | |
| *** PaulM has joined #openstack-dev | 09:14 | |
| digantya | rushiagr: may be busy now , can anyone tell me , the procedure to setup and test the Nova project on my local system ? | 09:15 |
| *** s0mik has joined #openstack-dev | 09:18 | |
| janisg | digantya, try devstack | 09:18 |
| *** egallen has quit IRC | 09:19 | |
| *** egallen has joined #openstack-dev | 09:24 | |
| *** egallen has quit IRC | 09:25 | |
| *** egallen has joined #openstack-dev | 09:26 | |
| *** egallen has left #openstack-dev | 09:27 | |
| *** derekh has joined #openstack-dev | 09:27 | |
| *** s0mik has quit IRC | 09:31 | |
| *** ociuhandu has quit IRC | 09:31 | |
| *** darraghb has joined #openstack-dev | 09:33 | |
| *** thingee has quit IRC | 09:34 | |
| *** s0mik has joined #openstack-dev | 09:35 | |
| *** rushiagr1 has joined #openstack-dev | 09:36 | |
| *** e1mer has quit IRC | 09:36 | |
| *** thingee has joined #openstack-dev | 09:37 | |
| *** s0mik has quit IRC | 09:37 | |
| rushiagr1 | digantya: try devstack is what i too will say, it is very simple to setup | 09:38 |
| *** mmagr has joined #openstack-dev | 09:38 | |
| *** rushiagr has quit IRC | 09:39 | |
| digantya | actually i builded the nova code successfully and then ? | 09:40 |
| *** Tross has joined #openstack-dev | 09:41 | |
| *** brucer is now known as brucer_out | 09:47 | |
| *** brucer_out has quit IRC | 09:49 | |
| *** gongysh has joined #openstack-dev | 09:52 | |
| *** Dorogs has joined #openstack-dev | 09:54 | |
| *** rha has quit IRC | 09:57 | |
| *** rha has joined #openstack-dev | 10:00 | |
| *** rha has quit IRC | 10:00 | |
| *** rha has joined #openstack-dev | 10:00 | |
| *** FlaPer87 has quit IRC | 10:03 | |
| saurabh | when i try to run test_extensions .py i get an error .. any ideas? | 10:04 |
| saurabh | machine131:~/devstack$ python /opt/stack/quantum/run_tests.py test_extensions nose.config: INFO: Ignoring files matching ['^\\.', '^_', '^setup\\.py$'] Failure Failure: ImportError (No module named webtest) ERROR ====================================================================== ERROR: Failure: ImportError (No module named webtest) ---------------------------------------------------------------------- Tra | 10:05 |
| *** e1mer has joined #openstack-dev | 10:05 | |
| *** Tross has quit IRC | 10:09 | |
| *** arata has joined #openstack-dev | 10:13 | |
| *** alexpilotti has joined #openstack-dev | 10:17 | |
| *** zing has joined #openstack-dev | 10:20 | |
| *** s0mik has joined #openstack-dev | 10:22 | |
| *** armaan has left #openstack-dev | 10:25 | |
| *** aloga has quit IRC | 10:26 | |
| openstackgerrit | A change was merged to openstack/quantum: Clean up quantum.api.v2.validators https://review.openstack.org/15706 | 10:27 |
| *** s0mik has quit IRC | 10:33 | |
| *** aloga has joined #openstack-dev | 10:34 | |
| *** aloga has joined #openstack-dev | 10:34 | |
| *** eglynn__ is now known as eglynn | 10:38 | |
| *** ociuhandu has joined #openstack-dev | 10:48 | |
| *** maurosr has joined #openstack-dev | 10:52 | |
| *** ociuhandu has quit IRC | 10:52 | |
| *** mjfork has joined #openstack-dev | 10:53 | |
| rushiagr1 | digantya: what exactly you did to build the nova code? | 10:54 |
| digantya | python setup.py install | 10:54 |
| rushiagr1 | saurabh: the test files are not supposed to be run that way. go to openstack docs/quantum docs | 10:55 |
| *** brucer has joined #openstack-dev | 10:55 | |
| rushiagr1 | digantya: check out the code from github, then go to the nova folder and then execute './run_tests.sh' as root user. when you see no virtual environment found, ask it to create one | 10:56 |
| digantya | ok after that ? | 10:57 |
| digantya | nosetest command not found | 10:58 |
| rushiagr1 | digantya: i am sorry, i give you instructions for running tests. you want a working openstack environment right? | 10:58 |
| digantya | yeah | 10:58 |
| digantya | rushiagr1: my goal is to fix bug in nova | 11:06 |
| gongysh | salvator: pping | 11:06 |
| gongysh | salv-orlando: ping | 11:07 |
| *** sniperd has joined #openstack-dev | 11:12 | |
| *** ociuhandu has joined #openstack-dev | 11:18 | |
| *** sniperd has quit IRC | 11:19 | |
| *** sniperd has joined #openstack-dev | 11:19 | |
| *** ociuhandu has quit IRC | 11:20 | |
| *** sniperd has quit IRC | 11:24 | |
| *** adalbas has joined #openstack-dev | 11:25 | |
| *** Dorogs has quit IRC | 11:31 | |
| *** rushiagr1 has left #openstack-dev | 11:35 | |
| *** \etc\bin has joined #openstack-dev | 11:48 | |
| *** e1mer has quit IRC | 11:48 | |
| *** zing has quit IRC | 11:52 | |
| *** Dorogs has joined #openstack-dev | 11:52 | |
| gongysh | garyk: ping | 11:54 |
| *** marcus_ has joined #openstack-dev | 11:56 | |
| digantya | hi is there anyone can tell me the procedure to setup and test the nova project on my local machine , that i have downloaded from Github ? | 11:59 |
| *** mnewby_ has joined #openstack-dev | 12:00 | |
| *** mnewby has quit IRC | 12:01 | |
| *** mnewby_ is now known as mnewby | 12:01 | |
| marcus_ | digantya: Have look at the documentation websie, that is at least a starting point http://docs.openstack.org/developer/nova/devref/development.environment.html | 12:11 |
| digantya | thanks marcus_ i am looking at it | 12:12 |
| *** sagar_nikam has joined #openstack-dev | 12:24 | |
| *** mindpixel has quit IRC | 12:29 | |
| *** roge has joined #openstack-dev | 12:32 | |
| *** aloga has quit IRC | 12:38 | |
| *** littleidea has joined #openstack-dev | 12:42 | |
| *** markvoelker has joined #openstack-dev | 12:43 | |
| *** roadTripper has joined #openstack-dev | 12:44 | |
| *** davidha has joined #openstack-dev | 12:47 | |
| *** Sparks has joined #openstack-dev | 12:47 | |
| *** johngarbutt has quit IRC | 12:49 | |
| *** aloga has joined #openstack-dev | 12:51 | |
| *** anniec72 has joined #openstack-dev | 12:54 | |
| *** littleidea has quit IRC | 12:56 | |
| *** afazekas has quit IRC | 12:56 | |
| *** Dorogs has quit IRC | 12:58 | |
| *** mrodden has joined #openstack-dev | 12:58 | |
| *** markmc has joined #openstack-dev | 12:59 | |
| *** ociuhandu has joined #openstack-dev | 12:59 | |
| *** dolphm has joined #openstack-dev | 13:02 | |
| *** dolphm has quit IRC | 13:03 | |
| *** mrodden has quit IRC | 13:05 | |
| *** roge has quit IRC | 13:06 | |
| *** mrodden has joined #openstack-dev | 13:07 | |
| *** dolphm has joined #openstack-dev | 13:08 | |
| *** alunduil has quit IRC | 13:14 | |
| *** sagar_nikam has quit IRC | 13:16 | |
| *** jaypipes has joined #openstack-dev | 13:17 | |
| *** Dorogs has joined #openstack-dev | 13:18 | |
| *** sunilthaha has quit IRC | 13:22 | |
| *** arbrandes has joined #openstack-dev | 13:29 | |
| *** kgriffs has joined #openstack-dev | 13:30 | |
| *** radez_g0n3 is now known as radez | 13:31 | |
| *** Dr_Who has joined #openstack-dev | 13:35 | |
| *** Dr_Who has quit IRC | 13:35 | |
| *** Dr_Who has joined #openstack-dev | 13:35 | |
| *** aloga has quit IRC | 13:36 | |
| *** dolphm has quit IRC | 13:37 | |
| *** samkottler|out is now known as samkottler | 13:38 | |
| *** blamar has quit IRC | 13:41 | |
| *** samkottler is now known as samkottler|brb | 13:41 | |
| *** dolphm has joined #openstack-dev | 13:44 | |
| *** EmilienM has left #openstack-dev | 13:47 | |
| *** arata has quit IRC | 13:48 | |
| *** littleidea has joined #openstack-dev | 13:48 | |
| *** EmilienM has joined #openstack-dev | 13:48 | |
| *** corXi has joined #openstack-dev | 13:50 | |
| *** aloga has joined #openstack-dev | 13:51 | |
| *** dprince has joined #openstack-dev | 13:51 | |
| *** winston-d has joined #openstack-dev | 13:52 | |
| *** networkstatic has quit IRC | 13:52 | |
| *** networkstatic has joined #openstack-dev | 13:53 | |
| *** Dorogs has quit IRC | 13:55 | |
| *** tmishler has joined #openstack-dev | 13:55 | |
| *** Dr_Who has quit IRC | 13:57 | |
| *** m4xmr has joined #openstack-dev | 13:58 | |
| *** boden has joined #openstack-dev | 13:59 | |
| *** dave_mcn has quit IRC | 13:59 | |
| *** samkottler|brb is now known as samkottler | 14:01 | |
| *** alunduil has joined #openstack-dev | 14:05 | |
| *** PaulM has quit IRC | 14:06 | |
| *** jcru has joined #openstack-dev | 14:06 | |
| *** esheffield has joined #openstack-dev | 14:08 | |
| *** kbringard has joined #openstack-dev | 14:09 | |
| *** zbitter is now known as zaneb | 14:10 | |
| *** gongysh has quit IRC | 14:24 | |
| *** anniec72 has quit IRC | 14:24 | |
| *** jcru is now known as jcru|away | 14:24 | |
| *** aloga has quit IRC | 14:25 | |
| *** aloga has joined #openstack-dev | 14:25 | |
| *** gongysh has joined #openstack-dev | 14:25 | |
| *** jcru|away is now known as jcru | 14:25 | |
| gongysh | garyk: ping | 14:27 |
| davidkranz | Could some one on nova core please approve https://review.openstack.org/#/c/15655/ It is the last tempest blocker for now. | 14:37 |
| russellb | davidkranz: done | 14:38 |
| davidkranz | russellb: Thanks! | 14:38 |
| *** nicromancer has joined #openstack-dev | 14:39 | |
| *** \etc\bin has quit IRC | 14:40 | |
| *** rkukura has joined #openstack-dev | 14:40 | |
| *** japage has joined #openstack-dev | 14:40 | |
| *** Dorogs has joined #openstack-dev | 14:41 | |
| *** sagar_nikam has joined #openstack-dev | 14:41 | |
| *** aeperezt has joined #openstack-dev | 14:42 | |
| *** garyk has quit IRC | 14:42 | |
| boden | ayoung or dolphm -- either of you get a sec for a quick question? | 14:43 |
| dolphm | boden: what's up | 14:43 |
| boden | dolphm thanks... question about the service catalog behavior, specifically with the catalog return in a POST /v2.0/tokens response... it appears the catalog can contain at most 1 service of a given type... so for example you have 1 'compute' services registered, each has the type 'compute' but they have different names... it appears you only get back 1. | 14:45 |
| boden | is this by design or maybe its a user error on my part... | 14:45 |
| boden | btw -- I meant to say "2 compute services register" | 14:46 |
| *** rbasak has joined #openstack-dev | 14:47 | |
| dolphm | boden: i think that's correct -- assuming you have multiple compute nodes deployed, they both provide the same service, but will have seperate sets of endpoints | 14:47 |
| boden | this would also appear to be what I see in the the get_catalog() method of the sql catalog driver too | 14:47 |
| *** Dorogs has quit IRC | 14:47 | |
| *** aloga has quit IRC | 14:49 | |
| *** sacharya has joined #openstack-dev | 14:49 | |
| *** aloga has joined #openstack-dev | 14:49 | |
| *** ijw1 has joined #openstack-dev | 14:50 | |
| *** ijw1 is now known as ijw | 14:50 | |
| davidkranz | harlowja: Josh, can you please rebase https://review.openstack.org/#/c/15655/ | 14:51 |
| dolphm | boden: from a specification perspective, i'm pretty sure you can have multiple services with the same type | 14:51 |
| davidkranz | jaypipes: I have everything working locally. Just waiting for https://review.openstack.org/#/c/15655/ to merge and we should be good to go. | 14:52 |
| dolphm | boden: however, i'm not sure the current keystoneclient would behave as expected with such a response (if i remember the service-finding code correctly, it'll return the first matching service type it finds) | 14:53 |
| davidkranz | jaypipes: Don't know what happend with your glance issue but my theory of the world works without it. | 14:53 |
| *** rnirmal has joined #openstack-dev | 14:53 | |
| davidkranz | jaypipes: The next step is to turn on the full gate. | 14:53 |
| *** aloga has quit IRC | 14:54 | |
| davidkranz | jaypipes: I profiled the tests and I think if we just mark a few of them as 'slow' we can take the next step. | 14:54 |
| davidkranz | jaypipes: The evidence from the last week is that we really need the full gate. | 14:54 |
| *** gongysh has quit IRC | 14:54 | |
| *** sagar_nikam has quit IRC | 14:55 | |
| *** jcru is now known as jcru|away | 14:56 | |
| *** gongysh has joined #openstack-dev | 14:56 | |
| *** adepti37 has joined #openstack-dev | 14:57 | |
| *** sacharya has quit IRC | 14:57 | |
| *** brucer is now known as brucer_afk | 14:58 | |
| *** jcru|away is now known as jcru | 15:01 | |
| *** torandu has joined #openstack-dev | 15:01 | |
| *** mrodden has quit IRC | 15:03 | |
| *** Dr_Who has joined #openstack-dev | 15:03 | |
| *** Dr_Who has quit IRC | 15:03 | |
| *** Dr_Who has joined #openstack-dev | 15:03 | |
| boden | dolphm thanks... based on what we are seeing, you only get 1 of the services (and its endpoints) back... we are using the REST APIs directly... based on where I'm sitting, this looks like a bug | 15:05 |
| *** annegentle_ has joined #openstack-dev | 15:05 | |
| dolphm | boden: that's what i would expect -- what kind of metadata are you using to differentiate the two services of the same type? | 15:06 |
| boden | dolhm -- they have the same type but different names and thus i would expect them to differ by name | 15:06 |
| boden | dolphm ^^ | 15:06 |
| dolphm | boden: and you expect clients to look for a particular service name? name *really* isn't intended to be machine-cansumable metadata | 15:08 |
| boden | dolphm -- we had assumed it was... how else can one register multiple services of the same type and differentiate them in a service catalog response? | 15:09 |
| dolphm | boden: that's why i'm asking about your specific use case -- what actually differentiates the two services? | 15:11 |
| *** aloga has joined #openstack-dev | 15:11 | |
| *** markmcclain has joined #openstack-dev | 15:12 | |
| jaypipes | davidkranz: I solved my glance issue but still get masses of errors and failures (in images_client and volume stuff) :( | 15:14 |
| gongysh | garyk: ping | 15:15 |
| openstackgerrit | A change was merged to openstack/nova: Another case of dictionary access https://review.openstack.org/15655 | 15:15 |
| boden | dolphm -- under the covers one is the actual glance service and 1 is a custom provider which implements the glance APIs.. we are looking to differentiate them... we could probably add an 'extra' metadata to the non-glance service but I don't think that 'shows through' on the CLI -- only the REST API | 15:15 |
| *** davidha has quit IRC | 15:16 | |
| *** apevec has joined #openstack-dev | 15:16 | |
| *** apevec has left #openstack-dev | 15:16 | |
| *** kgriffs has quit IRC | 15:16 | |
| *** kgriffs has joined #openstack-dev | 15:16 | |
| *** brucer_afk has quit IRC | 15:17 | |
| dolphm | boden: you probably want two sets of endpoints for the same service -- the appropriate attribute you should use depends on your use case -- here's the spec https://github.com/openstack/identity-api/blob/master/openstack-identity-api/src/docbkx/common/xsd/endpoints.xsd | 15:17 |
| *** jcru has quit IRC | 15:17 | |
| *** gongysh has quit IRC | 15:17 | |
| *** ondergetekende has quit IRC | 15:17 | |
| dolphm | boden: if the client doesn't support filtering endpoints by one of those attributes, it's definitely a bug :) (and honestly i have no idea if it does, beyond region) | 15:17 |
| *** jcru has joined #openstack-dev | 15:17 | |
| *** mrodden has joined #openstack-dev | 15:19 | |
| *** mjfork has quit IRC | 15:24 | |
| *** roge has joined #openstack-dev | 15:25 | |
| *** imsplitbit has joined #openstack-dev | 15:25 | |
| *** aloga has quit IRC | 15:27 | |
| openstackgerrit | A change was merged to openstack/nova: Make utils.mkfs() set label when fs=swap https://review.openstack.org/15714 | 15:27 |
| *** aloga has joined #openstack-dev | 15:29 | |
| *** sacharya has joined #openstack-dev | 15:31 | |
| *** oneiroi has joined #openstack-dev | 15:40 | |
| *** littleidea has quit IRC | 15:43 | |
| jaypipes | dtroyer, bcwaldon: if you get a chance... https://review.openstack.org/15735 | 15:43 |
| *** markwash has joined #openstack-dev | 15:43 | |
| *** nicromancer has quit IRC | 15:44 | |
| *** marcus_ has quit IRC | 15:44 | |
| *** winston-d has quit IRC | 15:44 | |
| *** armaan has joined #openstack-dev | 15:44 | |
| *** armaan has left #openstack-dev | 15:44 | |
| *** jcru is now known as jcru|away | 15:45 | |
| *** davidhadas has joined #openstack-dev | 15:50 | |
| *** apevec has joined #openstack-dev | 15:52 | |
| *** apevec has joined #openstack-dev | 15:52 | |
| *** hemna_ has quit IRC | 15:53 | |
| *** winston-d has joined #openstack-dev | 15:57 | |
| *** anniec72 has joined #openstack-dev | 15:58 | |
| *** jcru|away is now known as jcru | 16:00 | |
| *** Mandell has joined #openstack-dev | 16:00 | |
| *** marcus_ has joined #openstack-dev | 16:02 | |
| *** Dr_Who has quit IRC | 16:03 | |
| *** Dr_Who has joined #openstack-dev | 16:04 | |
| *** Dr_Who has joined #openstack-dev | 16:04 | |
| *** aloga has quit IRC | 16:04 | |
| *** reidrac has quit IRC | 16:06 | |
| *** Dorogs has joined #openstack-dev | 16:08 | |
| *** Mandell has quit IRC | 16:11 | |
| ayoung | dolphm, boden if the two services are fundamentally different, they should probably be registered as two different services. It sounds like boden has found a point where maybe the glance service is too granular, at least for their use case. I wonder if the "service" abstraction is really a useful abstraction, or if instead it should be endpoints and attributes? Maybe I've just been reading the Federation spec for too long. | 16:12 |
| *** colinmcnamara has joined #openstack-dev | 16:14 | |
| boden | ayoung -- I tend to think of a 'sevice' as a definition of capabilities... i.e. I'm an 'image' service and can do the things from the 'image api contract'.. where as endpoints are actual 'locations' of services implementing those capabilities | 16:14 |
| davidkranz | jaypipes: What's going on with the one remaining failure in http://logs.openstack.org/15653/1/gate/gate-tempest-devstack-vm/17878/console.html | 16:15 |
| *** datsun180b has joined #openstack-dev | 16:15 | |
| davidkranz | jaypipes: This test passes for me locally unless this broke in the past day or two. I will resync and see. | 16:15 |
| boden | ayoung -- so I think our use case is really about -- ok so I know these endpoints have the 'image' capabilities, how can I make some distinction between the endpoints | 16:15 |
| ayoung | boden, I realize that. But I don;t like it. I think that it might prov to be too rigid a framework over time | 16:15 |
| ayoung | boden, right. | 16:15 |
| ayoung | so you have to provide attributes to differentiate | 16:15 |
| ijw | boden: specification versus implementation, basically then | 16:16 |
| *** metral has quit IRC | 16:16 | |
| *** metral has joined #openstack-dev | 16:16 | |
| ayoung | and now you have two ways of designating purpose. I'm just muttering here, but I also don't like the confusion of calling them services. In kerberos, a service is what we call and endpoint. | 16:16 |
| boden | ayoung -- yes, but it seems awkward that the catalog only returns 1 serivce and associated endpoint when you have more than 1 service of the same name... maybe its for good reason, but was not as expected | 16:17 |
| jaypipes | davidkranz: I see that locally as well. | 16:18 |
| ayoung | boden, yeah, that is not a good setup. It prechooses for the end user, among other things, which to go to. I'm not sure what the right solution is, though | 16:18 |
| jaypipes | davidkranz: looks to be a problem with the tools/configure_tempest.sh script in devstack. I will investigate further. | 16:19 |
| *** metral_ has joined #openstack-dev | 16:19 | |
| jaypipes | davidkranz: FYI.. the infinite loop glance workaround for devstack is here: https://review.openstack.org/#/c/15735/ | 16:19 |
| *** corXi has quit IRC | 16:20 | |
| *** Mandell has joined #openstack-dev | 16:21 | |
| *** andrewbogott_afk is now known as andrewbogott | 16:22 | |
| *** metral has quit IRC | 16:23 | |
| *** metral_ is now known as metral | 16:23 | |
| *** garyk has joined #openstack-dev | 16:23 | |
| *** colinmcnamara has quit IRC | 16:26 | |
| davidkranz | jaypipes: This db problem happend very recently. After updating I see it too. | 16:27 |
| *** mrunge has quit IRC | 16:27 | |
| *** colinmcnamara has joined #openstack-dev | 16:27 | |
| *** danwent has joined #openstack-dev | 16:29 | |
| jaypipes | davidkranz: got it. pushing a fix to devstack shortly. | 16:30 |
| *** Dr_Who has quit IRC | 16:30 | |
| davidkranz | jaypipes: Thx | 16:30 |
| *** edygarcia has joined #openstack-dev | 16:31 | |
| jaypipes | davidkranz: np | 16:32 |
| *** markwash has quit IRC | 16:33 | |
| *** jtran has joined #openstack-dev | 16:34 | |
| *** maoy has joined #openstack-dev | 16:36 | |
| *** AnilV4 has quit IRC | 16:37 | |
| *** maurosr has quit IRC | 16:38 | |
| *** brucer has joined #openstack-dev | 16:39 | |
| *** maurosr has joined #openstack-dev | 16:41 | |
| *** jcooley has joined #openstack-dev | 16:41 | |
| *** jcru is now known as jcru|away | 16:42 | |
| enikanorov_ | folks, how to fix the comments posted on review? I mean do i need to post another diff with fixes according to comments? or somehow differently? | 16:43 |
| *** jcru|away is now known as jcru | 16:44 | |
| *** maurosr has quit IRC | 16:44 | |
| *** maurosr has joined #openstack-dev | 16:44 | |
| devananda | vishy: since the nova volume code is getting cut, is the snapshot code also going away? | 16:45 |
| *** colinmcnamara has quit IRC | 16:46 | |
| *** brucer has quit IRC | 16:46 | |
| *** kmartin has quit IRC | 16:47 | |
| clarkb | enikanorov_: typically you will amend the existing commit then push the changes back to gerrit | 16:48 |
| openstackgerrit | A change was merged to openstack/nova: clean up dnsdomain_unregister https://review.openstack.org/15580 | 16:48 |
| clarkb | enikanorov_: http://wiki.openstack.org/GerritWorkflow describes the process | 16:49 |
| openstackgerrit | A change was merged to openstack/nova: add missing opts to test_db_api.py https://review.openstack.org/15661 | 16:49 |
| *** saurabhsurana has joined #openstack-dev | 16:49 | |
| *** saurabhsurana has left #openstack-dev | 16:50 | |
| *** danpb has joined #openstack-dev | 16:50 | |
| *** egallen has joined #openstack-dev | 16:51 | |
| openstackgerrit | A change was merged to openstack/quantum: Explicitly include versioninfo in tarball. https://review.openstack.org/15736 | 16:52 |
| *** marcus_ has quit IRC | 16:54 | |
| *** gyee has joined #openstack-dev | 16:55 | |
| *** maurosr has quit IRC | 16:56 | |
| *** arbrandes has quit IRC | 16:58 | |
| *** maurosr has joined #openstack-dev | 16:59 | |
| *** egallen has quit IRC | 16:59 | |
| *** maurosr has quit IRC | 16:59 | |
| *** markwash has joined #openstack-dev | 17:00 | |
| *** maurosr has joined #openstack-dev | 17:01 | |
| *** maurosr has quit IRC | 17:01 | |
| *** maurosr has joined #openstack-dev | 17:02 | |
| *** Dr_Who has joined #openstack-dev | 17:03 | |
| *** Dr_Who has quit IRC | 17:03 | |
| *** Dr_Who has joined #openstack-dev | 17:03 | |
| *** saurabhs has joined #openstack-dev | 17:03 | |
| *** littleidea has joined #openstack-dev | 17:04 | |
| *** mmagr has quit IRC | 17:09 | |
| *** jcmartin has joined #openstack-dev | 17:09 | |
| *** saurabhs has left #openstack-dev | 17:10 | |
| *** Mandell has quit IRC | 17:10 | |
| *** daddyjoseph97 has joined #openstack-dev | 17:12 | |
| *** daddyjoseph97 has quit IRC | 17:12 | |
| *** maurosr has quit IRC | 17:13 | |
| *** maurosr has joined #openstack-dev | 17:13 | |
| *** eharney has joined #openstack-dev | 17:19 | |
| *** maurosr has quit IRC | 17:19 | |
| *** blamar has joined #openstack-dev | 17:19 | |
| *** EmilienM has left #openstack-dev | 17:19 | |
| *** maurosr has joined #openstack-dev | 17:20 | |
| openstackgerrit | A change was merged to openstack/quantum: Add indication when quantum server started. https://review.openstack.org/15695 | 17:20 |
| *** darraghb has quit IRC | 17:20 | |
| *** blamar has quit IRC | 17:21 | |
| *** anniec has joined #openstack-dev | 17:23 | |
| jaypipes | davidkranz: https://review.openstack.org/15744 | 17:25 |
| jaypipes | dtroyer: ^^ | 17:25 |
| *** s0mik has joined #openstack-dev | 17:25 | |
| *** crandquist has quit IRC | 17:26 | |
| *** hemna has joined #openstack-dev | 17:26 | |
| alexpilotti | russellb: hi | 17:26 |
| alexpilotti | russellb: just saw your comment | 17:26 |
| alexpilotti | russellb: on https://review.openstack.org/#/c/15743/ | 17:27 |
| *** crandquist has joined #openstack-dev | 17:27 | |
| alexpilotti | russellb: since the bug fixes listed are required to pass the tests in the rest of the code | 17:28 |
| openstackgerrit | A change was merged to openstack/nova: Remove nova-volume DB https://review.openstack.org/15602 | 17:28 |
| alexpilotti | russellb: should I commit the bug fixes, wait for them to get approved and only afterwards commit the rest? | 17:28 |
| *** jcooley has quit IRC | 17:28 | |
| *** anniec has quit IRC | 17:29 | |
| *** s0mik has quit IRC | 17:30 | |
| russellb | alexpilotti: you can submit them all at once. if the commits are all in the same branch and you submit them all at the same time, gerrit knows the patch dependencies | 17:32 |
| russellb | so it won't let one merge before the other if it shouldn't | 17:32 |
| *** marcus__ has joined #openstack-dev | 17:33 | |
| russellb | and patches will only get tested with all of the patches they depend on, if that makes sense | 17:33 |
| *** maurosr has quit IRC | 17:33 | |
| *** maurosr has joined #openstack-dev | 17:33 | |
| *** maurosr has quit IRC | 17:33 | |
| alexpilotti | russellb: so separate commits in the same branch and a single "git review" at the end. It makes definitely sense! | 17:33 |
| *** esp2 has joined #openstack-dev | 17:33 | |
| russellb | alexpilotti: yep, that's it! thanks | 17:34 |
| russellb | it will make it a lot easier to review each bug fix | 17:34 |
| russellb | and hopefully get the code in faster overall | 17:34 |
| *** esp2 has left #openstack-dev | 17:34 | |
| alexpilotti | russellb: any idea about how to solve the "Please rebase your change and upload a new patchset." error that Jenkins gives? | 17:34 |
| *** anniec72 has left #openstack-dev | 17:35 | |
| alexpilotti | russellb: I did a git pull in master and a git rebase in the new branch | 17:35 |
| dansmith | alexpilotti: did you get actual changes in your master branch? | 17:35 |
| alexpilotti | russellb: no, otherwise git would ask to stash them I guess | 17:36 |
| *** maurosr has joined #openstack-dev | 17:36 | |
| devananda | russellb: what would it take to convince you to review baremetal patch #1 today? (13920) | 17:36 |
| dansmith | and, you did "git rebase master" in your topic branch, right? | 17:36 |
| dansmith | eh? | 17:36 |
| dansmith | alexpilotti: what I meant was, when you did "git checkout master && git pull origin master" did you see it pull in a bunch of changes? | 17:36 |
| alexpilotti | dansmith: yep | 17:37 |
| dansmith | alexpilotti: is that a 'yep' to the rebase question above as well? | 17:37 |
| alexpilotti | dansmith: to the latter, I'm going to redo a "git rebase master". My doubt is that I did a "git rebase" instead | 17:38 |
| dansmith | I think you definitely want the former :) | 17:38 |
| *** imsplitbit has quit IRC | 17:38 | |
| openstackgerrit | A change was merged to openstack/nova: Remove generic topic support from filter scheduler https://review.openstack.org/15716 | 17:40 |
| *** jcooley has joined #openstack-dev | 17:40 | |
| dtroyer | jaypipes: will Tempest work with postgresql? | 17:41 |
| *** datsun180b has quit IRC | 17:41 | |
| alexpilotti | dansmith: before the next git review, based on russellb suggestion I need to split the commits in 3 separate ones | 17:41 |
| alexpilotti | dansmith: I'm quite new to git, so thanks for helping :-) | 17:41 |
| *** timello has quit IRC | 17:42 | |
| dansmith | alexpilotti: split commits, or pull three commits into three separate branches? | 17:42 |
| *** timello has joined #openstack-dev | 17:42 | |
| alexpilotti | dansmith: in the same branch | 17:42 |
| dansmith | alexpilotti: is this 15743? | 17:42 |
| alexpilotti | dansmith: yes: https://review.openstack.org/#/c/15743/ | 17:43 |
| *** maurosr has quit IRC | 17:43 | |
| dansmith | alexpilotti: so, if you want to split based on file changes, this is pretty easy | 17:43 |
| dansmith | is that the case? | 17:43 |
| alexpilotti | dansmith: yes, being simple fixes is just one file per commit | 17:44 |
| *** maurosr has joined #openstack-dev | 17:44 | |
| dansmith | alexpilotti: no, what I mean is, you touch a bunch of files.. are your three commits going to be purely groupings of your existing files? | 17:45 |
| dansmith | for example: | 17:45 |
| dansmith | commit1: nova/api/* | 17:45 |
| dansmith | commit 2: nova/virt/* | 17:45 |
| dansmith | commit 3: nova/tests/* | 17:45 |
| dansmith | ? | 17:45 |
| alexpilotti | dansmith: commit 1: nova/api/metadata/base.py | 17:45 |
| *** aeperezt has quit IRC | 17:45 | |
| alexpilotti | dansmith: commit 2: nova/utils.py | 17:45 |
| *** jpich has left #openstack-dev | 17:45 | |
| *** derekh has quit IRC | 17:45 | |
| alexpilotti | dansmith: commit 3: all the rest | 17:46 |
| dansmith | okay, perfect | 17:46 |
| dansmith | so here's what I'd do: | 17:46 |
| *** koolhead17 has joined #openstack-dev | 17:46 | |
| *** timello has quit IRC | 17:46 | |
| dansmith | pull out the changes that will be the first commit, and then pull out those that will be the second, | 17:46 |
| *** roz has quit IRC | 17:46 | |
| dansmith | then ammend your current patch to be everything else without those | 17:46 |
| *** timello has joined #openstack-dev | 17:46 | |
| *** marcus__ has quit IRC | 17:46 | |
| dansmith | then create two commits with each of those changes and then get them all lined up, sound good? | 17:47 |
| dansmith | alexpilotti: if so, this is what I'd do: | 17:47 |
| *** maurosr has quit IRC | 17:47 | |
| dansmith | git show nova/api > api_changes.patch | 17:47 |
| dansmith | git show nova/utils.py > utils_changes.patch | 17:47 |
| dansmith | patch -p1 -R < api_changes.patch | 17:47 |
| dansmith | patch -p1 -R < utils_changes.patch | 17:47 |
| dansmith | git commit -a --amend | 17:47 |
| *** maurosr has joined #openstack-dev | 17:48 | |
| dansmith | patch -p1 < api_changes.patch | 17:48 |
| dansmith | git commit -a | 17:48 |
| dansmith | patch -p1 < utils_changes.patch | 17:48 |
| *** SlickNik has joined #openstack-dev | 17:48 | |
| dansmith | git commit -a | 17:48 |
| dansmith | git rebase -i HEAD ^^^ | 17:48 |
| *** jcooley has quit IRC | 17:48 | |
| dansmith | and change the order around so they look like you want | 17:48 |
| dansmith | 3. Profit! | 17:48 |
| alexpilotti | dansmith: thanks! :-) | 17:48 |
| dansmith | np, let me know how it works :) | 17:48 |
| alexpilotti | dansmith: going to do it right away! | 17:49 |
| dansmith | oops, above "HEAD ^^^" should be "HEAD^^^" | 17:49 |
| openstackgerrit | A change was merged to openstack/keystone: Fix transient test failures (bug 1077065, bug 1045962) https://review.openstack.org/15742 | 17:49 |
| uvirtbot` | Launchpad bug 1077065 in keystone "Transient test failures: test_service.RemoteUserTest" [Medium,In progress] https://launchpad.net/bugs/1077065 | 17:49 |
| uvirtbot` | Launchpad bug 1045962 in keystone "Transient test failure: test_token_expiry_maintained" [Low,In progress] https://launchpad.net/bugs/1045962 | 17:49 |
| *** timello has quit IRC | 17:51 | |
| bcwaldon | jaypipes: thoughts on https://review.openstack.org/#/c/15735/ ? | 17:51 |
| *** saurabhs has joined #openstack-dev | 17:51 | |
| *** danpb has quit IRC | 17:52 | |
| alexpilotti | dansmith: so "git commit -a --amend" is executed first but contains the last commit, is this correct? | 17:52 |
| alexpilotti | dansmith: I mean, the nova/api and nova/utils bug fixes are required by the rest of the code | 17:54 |
| alexpilotti | dansmith: should I invert it? | 17:54 |
| russellb | devananda: HM. How about ... an IOU for a beer. Nah, nothing ... I'll try to review it today. | 17:57 |
| *** timello has joined #openstack-dev | 17:57 | |
| devananda | russellb: i'll gladly buy you a beer at the next summit :) | 17:57 |
| russellb | i'm sure i'll bug you for a review one day and we'll be even | 17:57 |
| *** Sparks has quit IRC | 17:57 | |
| devananda | but really, once that patch is merged, we (hp folks) should be able to work a lot quicker with the rest of the baremetal driver | 17:57 |
| vishy | devananda: snapshot db code? yes | 17:57 |
| dansmith | alexpilotti: yeah | 17:57 |
| dansmith | alexpilotti: when you do the last step, | 17:57 |
| dansmith | you'll reorder the commits so they're right (and so the current one with the tests is last) | 17:57 |
| devananda | vishy: thanks | 17:57 |
| dansmith | alexpilotti: invert what? maybe I'm confused. what step are you on above? | 17:58 |
| *** colinmcnamara has joined #openstack-dev | 17:58 | |
| *** timello has quit IRC | 17:58 | |
| vishy | smoser: ping | 17:58 |
| *** timello has joined #openstack-dev | 17:58 | |
| dansmith | alexpilotti: you're on "commit -a --amend" above, right? when you do that, you should see the old commit message, and the file list should _just_ show the files that will eventually be commit #3 | 17:58 |
| vishy | devananda: in fact i think jog0 has a patch in review that removes it. | 17:58 |
| alexpilotti | dansmith: the commits should be in order: 1,2, 3 as 3 depends on 1 and 2 | 17:58 |
| dansmith | alexpilotti: I know.. we'll fix the order at the end | 17:59 |
| *** maurosr has quit IRC | 17:59 | |
| *** andrewbogott is now known as andrewbogott_afk | 17:59 | |
| *** otherwiseguy has joined #openstack-dev | 17:59 | |
| alexpilotti | dansmith: cool! I start liking git! ;-) | 17:59 |
| dansmith | alexpilotti: I promise :D | 17:59 |
| *** maurosr has joined #openstack-dev | 17:59 | |
| dansmith | at the end, the "rebase -i" part will let you arbitrarily re-order what you've done | 18:00 |
| *** aeperezt has joined #openstack-dev | 18:00 | |
| devananda | vishy: ah, i didn't see that review linked onthe no-volume bp | 18:00 |
| *** jcmartin_ has joined #openstack-dev | 18:01 | |
| *** colinmcnamara has quit IRC | 18:02 | |
| devananda | vishy: any chance i can also convince you to review the first baremetal patch (13920) again today? getting that into trunk should unblock us enough to work on the other ones | 18:02 |
| *** jcmartin has quit IRC | 18:02 | |
| *** jcmartin_ is now known as jcmartin | 18:02 | |
| vishy | devananda: sure, will do | 18:02 |
| *** annegentle_ has quit IRC | 18:03 | |
| devananda | awesome, ty | 18:03 |
| alexpilotti | dansmith: in the commit message of the last 2 "git commit -a", I don't have to put the change-id from the 15743 review, correct? | 18:04 |
| openstackgerrit | A change was merged to openstack/nova: Fix import ordering in /bin scripts. https://review.openstack.org/15748 | 18:05 |
| dansmith | alexpilotti: right, don't do that so that they'll get new ones automatically | 18:06 |
| *** jcru is now known as jcru|away | 18:06 | |
| *** apevec has quit IRC | 18:07 | |
| *** armaan has joined #openstack-dev | 18:07 | |
| alexpilotti | dansmith: git rebase -i HEAD^^^ gives: fatal: Needed a single revision | 18:08 |
| alexpilotti | dansmith: and "invalid upstream HEAD^ " | 18:08 |
| *** timello has quit IRC | 18:08 | |
| *** timello has joined #openstack-dev | 18:08 | |
| dansmith | alexpilotti: ...works for me | 18:09 |
| dansmith | alexpilotti: what shell and what version of git? | 18:09 |
| alexpilotti | git 1.711 | 18:09 |
| alexpilotti | dansmith: 1.7.11, sorry | 18:09 |
| alexpilotti | dansmith: shell: cmd.exe ;-) | 18:09 |
| dansmith | alexpilotti: is your shell interpreting the ^ chars? some unix shells will, and mess up that command | 18:10 |
| russellb | can do HEAD~3 syntax instead | 18:10 |
| russellb | if so | 18:10 |
| *** eglynn has quit IRC | 18:10 | |
| *** zaitcev has joined #openstack-dev | 18:10 | |
| *** gatuus has joined #openstack-dev | 18:10 | |
| *** timello has quit IRC | 18:11 | |
| alexpilotti | dansmith: good point! it worked with: git rebase -i "HEAD^^^" | 18:11 |
| *** timello has joined #openstack-dev | 18:11 | |
| dansmith | alexpilotti: cool | 18:12 |
| alexpilotti | dansmith: now I'm in vi, with one line per debug starting with "pick" | 18:12 |
| alexpilotti | dansmith: do I have to reorder the lines? | 18:12 |
| dansmith | alexpilotti: right, reorder the lines | 18:12 |
| alexpilotti | dansmith: and then simply :wq? wow | 18:12 |
| * dansmith boggles at git and vi in cmd.exe | 18:12 | |
| dansmith | alexpilotti: yep | 18:12 |
| dansmith | then look at git log | 18:12 |
| russellb | *magic* | 18:13 |
| *** eglynn has joined #openstack-dev | 18:13 | |
| *** torandu has quit IRC | 18:14 | |
| *** dspano has joined #openstack-dev | 18:14 | |
| openstackgerrit | A change was merged to openstack/nova: Make tox.ini run pep8/hacking checks on bin. https://review.openstack.org/15749 | 18:14 |
| *** annegentle_ has joined #openstack-dev | 18:15 | |
| jaypipes | bcwaldon: I don't think that's easier actually :) I think it's easier to just use what's already there :) | 18:15 |
| alexpilotti | dansmith: the git log should be in .git? | 18:16 |
| dansmith | no, run "git log" | 18:17 |
| alexpilotti | dansmith: got it :-) | 18:17 |
| alexpilotti | dansmith: the patches are in correct order (last one on top) | 18:17 |
| dansmith | alexpilotti: excellent! | 18:18 |
| alexpilotti | dansmith: should I go with "git review"? | 18:18 |
| dansmith | alexpilotti: yeah, so when you do git review, | 18:18 |
| dansmith | it's going to ask if you really want to submit three things | 18:18 |
| dansmith | just say yes, and you'll have three new reviews up there | 18:18 |
| *** colinmcnamara has joined #openstack-dev | 18:20 | |
| *** colinmcnamara has quit IRC | 18:21 | |
| alexpilotti | dansmith: you earned 3 beers (one per commit) :-) | 18:22 |
| bcwaldon | jaypipes: ok, then I'm fine with it | 18:22 |
| dansmith | alexpilotti: sweet! :D | 18:22 |
| alexpilotti | dansmith: seriously, tx for the accelerated git training!! | 18:22 |
| alexpilotti | dansmith: normally I'd have gone with a git clone reapplying all the steps manually | 18:22 |
| dansmith | alexpilotti: no problem :) | 18:22 |
| dansmith | alexpilotti: hah :) | 18:22 |
| *** colinmcnamara has joined #openstack-dev | 18:23 | |
| *** annegentle_ has quit IRC | 18:23 | |
| davidkranz | jaypipes: Did you see Dean's comment about postgres in https://review.openstack.org/#/c/15744/1 | 18:24 |
| *** colinmcnamara has quit IRC | 18:24 | |
| *** jcooley has joined #openstack-dev | 18:24 | |
| *** daddyjoseph97 has joined #openstack-dev | 18:24 | |
| jaypipes | davidkranz: yup, already pushed a fix. | 18:24 |
| *** jog0 has joined #openstack-dev | 18:24 | |
| alexpilotti | dansmith: In case some reviewers should ask to change something in commit 1 or 2 | 18:24 |
| davidkranz | jaypipes: Ah. Stale browser. Sorry. | 18:25 |
| alexpilotti | dansmith: that would bo beyond my knowledge of "git commit -a --amend" | 18:25 |
| dansmith | alexpilotti: do the rebase -i command again, but change the "pick" to "edit" | 18:25 |
| dansmith | then it will stop at that commit, | 18:25 |
| dansmith | let you commit -a --amend, and then "git rebase --continue" again | 18:25 |
| alexpilotti | dansmith: wow | 18:25 |
| dansmith | so you can kinda "replay" the commits, but stop on the desired one to amend some changes | 18:25 |
| alexpilotti | dansmith: tx again! Now I have to look for some reviewers ;-) | 18:26 |
| dansmith | alexpilotti: that's the hard part :) | 18:26 |
| *** Mandell has joined #openstack-dev | 18:27 | |
| *** SlickNik has quit IRC | 18:28 | |
| *** koolhead17 has quit IRC | 18:30 | |
| dripton | devananda: https://review.openstack.org/#/c/15651/ is the patch I mentioned yesterday at the nova-db meeting. PTAL. | 18:30 |
| devananda | dripton: great! Yea, i have poked at nearly all of those functions and wondered if they were in use anywhere | 18:31 |
| devananda | glad to see them getting cut | 18:31 |
| *** rafaduran has left #openstack-dev | 18:31 | |
| davidkranz | dtroyer: Can you take a look at Jay's fix on https://review.openstack.org/#/c/15744/2 ? This is the last issue breaking the tempest gate. | 18:32 |
| devananda | dripton: though i probably should go abandon a change or two that i just submitted now ... heh | 18:32 |
| *** saurabhs has left #openstack-dev | 18:33 | |
| dripton | devananda: yes, I had the same problem since Joe was removing things at the same time I was. | 18:33 |
| ociuhandu | smoser: ping | 18:33 |
| devananda | dripton: what are you using to verify that a function isn't in use any more? | 18:33 |
| *** saurabhs has joined #openstack-dev | 18:33 | |
| dtroyer | davidkranz: +2 | 18:33 |
| dripton | devananda: git grep, then ack from the top of my git trees | 18:34 |
| davidkranz | dtroyer: Thx. Who can approve it? | 18:34 |
| devananda | dripton: ftr, https://review.openstack.org/#/c/15738/ | 18:34 |
| *** saurabhs has left #openstack-dev | 18:34 | |
| dripton | devananda: the only way to get a false positive is if someone is being tricky with getattr and building function names from partial strings. I hit one of those but the unit tests caught it. | 18:34 |
| *** saurabhs has joined #openstack-dev | 18:34 | |
| devananda | dripton: ok, so i'm still wrapping my head around some of the API / RPC stuff. can something outside of openstack/nova be calling these? or are you searching multiple projects? | 18:35 |
| *** ewindisch has quit IRC | 18:35 | |
| devananda | if so which ones :) | 18:35 |
| *** nati_ueno has joined #openstack-dev | 18:35 | |
| *** ewindisch has joined #openstack-dev | 18:36 | |
| *** annegentle_ has joined #openstack-dev | 18:36 | |
| alexpilotti | dansmith: about your comment request for a bug opened. Does it mean that I have to commit it in a separate branch? | 18:36 |
| dripton | devananda: I believe they should only be called from inside nova, and cross-project usage is supposed to go through the http API. | 18:36 |
| dansmith | alexpilotti: no, just put the bug in the commit message, like "Fixes bug #1234" | 18:36 |
| uvirtbot` | Launchpad bug 1234 in launchpad "Gina is an unmaintainable mess of command line options, environment variables and shell scripts" [Medium,Fix released] https://launchpad.net/bugs/1234 | 18:36 |
| *** stevebake has quit IRC | 18:37 | |
| alexpilotti | dansmith: sure. | 18:37 |
| dripton | devananda: some storage functions were moved to Cinder, so there were hits on the same name there. | 18:37 |
| dtroyer | davidkranz: bcwaldon got it | 18:37 |
| *** gatuus has quit IRC | 18:38 | |
| boden | ayoung or dolphm -- do you guys consider it a bad practice to use the 'admin_token' (shared secret) for non openstack services integrating with Keystone that need to perform admin type operations? | 18:38 |
| dripton | devananda: the backstop test is to clone every OpenStack git repo under one directory, then use ack from there. But I don't think it's actually needed since these functions are not supposed to be called across projects. (If we're doing that then we're in deep trouble.) | 18:38 |
| ayoung | boden, one moment | 18:38 |
| *** sniperd has joined #openstack-dev | 18:40 | |
| *** littleidea has quit IRC | 18:40 | |
| *** echohead has joined #openstack-dev | 18:40 | |
| *** littleidea has joined #openstack-dev | 18:40 | |
| *** armaan has left #openstack-dev | 18:40 | |
| dolphm | boden: yes 110% | 18:41 |
| *** saurabhs has left #openstack-dev | 18:42 | |
| *** eglynn_ has joined #openstack-dev | 18:42 | |
| *** eglynn has quit IRC | 18:42 | |
| dolphm | boden: admin_token is not associated with a user, it therefore has no roles, etc. using it for anything other than bootstrapping an admin in keystone is bad practice | 18:42 |
| boden | dolphm ayoung -- ok thanks... I kind of assumed that but the .conf comments and docs dont necessarly paint the full picture | 18:43 |
| *** tmishler has quit IRC | 18:43 | |
| *** littleidea has quit IRC | 18:43 | |
| *** annashen has quit IRC | 18:45 | |
| *** littleidea has joined #openstack-dev | 18:45 | |
| russellb | arosen: hey ... i see that you added me to some reviews, but they are for stable/folsom. i can't review those. For stable branches it's the people on the openstack-stable-maint team. | 18:46 |
| russellb | http://launchpad.net/~openstack-stable-maint | 18:47 |
| russellb | should be able to see the member list there | 18:47 |
| *** pabelanger has joined #openstack-dev | 18:47 | |
| *** adalbas has quit IRC | 18:47 | |
| arosen | russellb: yup I can see the list. Thanks | 18:49 |
| *** jakedahn_zz is now known as jakedahn | 18:51 | |
| *** s0mik has joined #openstack-dev | 18:52 | |
| *** gatuus has joined #openstack-dev | 18:54 | |
| *** annegentle_ has quit IRC | 18:55 | |
| devananda | dripton: gotcha, thanks | 18:55 |
| dansmith | alexpilotti: nice, eh? | 18:55 |
| alexpilotti | dansmith: you mean the git rebasing trick? I love it! :-) | 18:56 |
| dansmith | alexpilotti: I mean the rebasing trick to address feedback, turn around changes before the reviewer gets distracted, resulting in merges-in-minutes like you just got :) | 18:56 |
| alexpilotti | dansmith: about the bugs and descriptions, if you look at my other commits I always put proper descriptions, it this case I just took teh lines out of the initial "macro" commit, my bad | 18:57 |
| *** jeblair is now known as one_at_9985 | 18:57 | |
| dansmith | alexpilotti: no problem, just gotta make sure the commits stand on their own | 18:57 |
| *** one_at_9985 is now known as jeblair | 18:57 | |
| alexpilotti | dansmith: opened a bug for the normpath() one as well: https://review.openstack.org/#/c/15763/ | 18:58 |
| dansmith | alexpilotti: yep, cool | 18:58 |
| *** thingee has quit IRC | 18:58 | |
| *** saurabhs has joined #openstack-dev | 18:59 | |
| *** jcru|away is now known as jcru | 19:00 | |
| *** jcmartin has quit IRC | 19:00 | |
| alexpilotti | dansmith: I lost your posix normpath alternative. Can you please retype it? | 19:00 |
| openstackgerrit | A change was merged to openstack/glance: Pep8 fixes to bin/glance* scripts. https://review.openstack.org/15753 | 19:00 |
| alexpilotti | dansmith: got it, it was actually jerdfelt's comment | 19:01 |
| alexpilotti | jerdfelt: Hi | 19:02 |
| *** datsun180b has joined #openstack-dev | 19:02 | |
| alexpilotti | jerdfelt: posixpath.normpath() works fine! | 19:02 |
| alexpilotti | jerdfelt: do you see potential issues w replacing all url normpath's in Nova and other projects? | 19:03 |
| *** adalbas has joined #openstack-dev | 19:04 | |
| *** dolphm has quit IRC | 19:05 | |
| alexpilotti | smoser: hi! | 19:10 |
| *** s0mik has quit IRC | 19:11 | |
| openstackgerrit | A change was merged to openstack/quantum: Adds test scripts for _validate_values https://review.openstack.org/15696 | 19:12 |
| *** jimfehlig has joined #openstack-dev | 19:12 | |
| garyk | nati_ueno: ping | 19:13 |
| nati_ueno | garyk: pong | 19:14 |
| garyk | nati_ueno: hi, how are you? can you please update https://review.openstack.org/#/c/15156/ - minor issue | 19:14 |
| nati_ueno | I got it! | 19:15 |
| *** timello has quit IRC | 19:15 | |
| nati_ueno | But it looks "I would prefer that you didn't merge this; Approved" ... I'm little bit confused | 19:15 |
| nati_ueno | It looks starting gate jobs | 19:15 |
| *** timello has joined #openstack-dev | 19:15 | |
| garyk | thanks. i may have approved by mistake | 19:15 |
| garyk | my bad. | 19:16 |
| nati_ueno | I got it | 19:16 |
| *** arosen has quit IRC | 19:16 | |
| *** s0mik has joined #openstack-dev | 19:16 | |
| *** arosen has joined #openstack-dev | 19:16 | |
| nati_ueno | I'll fix this order issue on next patch about security groups | 19:16 |
| garyk | ok. sorry for the mess | 19:17 |
| nati_ueno | Sorry for this bug. I mistook order of alphabetic order | 19:18 |
| *** eglynn_ has quit IRC | 19:18 | |
| alexpilotti | jerdfelt: I resubmitted using posipath.normpath(): https://review.openstack.org/#/c/15763/ | 19:19 |
| openstackgerrit | A change was merged to openstack/nova: Fixes a bug in nova.utils, due to Windows compatibility issues. https://review.openstack.org/15762 | 19:21 |
| openstackgerrit | A change was merged to openstack/swift: give swob default content-type https://review.openstack.org/15765 | 19:22 |
| openstackgerrit | A change was merged to openstack/quantum: Import lockutils and fileutils from openstack-common https://review.openstack.org/15156 | 19:22 |
| *** kmansel has joined #openstack-dev | 19:23 | |
| *** relateable has quit IRC | 19:24 | |
| *** Mandell has quit IRC | 19:24 | |
| *** colinmcnamara has joined #openstack-dev | 19:24 | |
| *** colinmcnamara has quit IRC | 19:24 | |
| *** Shrews has quit IRC | 19:24 | |
| *** Shrews has joined #openstack-dev | 19:25 | |
| *** eglynn_ has joined #openstack-dev | 19:26 | |
| *** otherwiseguy has quit IRC | 19:26 | |
| *** m4xmr has left #openstack-dev | 19:27 | |
| *** jcru is now known as jcru|away | 19:29 | |
| *** vipul is now known as vipul|away | 19:31 | |
| *** jcru|away is now known as jcru | 19:33 | |
| *** ndipanov has quit IRC | 19:33 | |
| *** novas0x2a|laptop has joined #openstack-dev | 19:35 | |
| *** daddyjoseph97 has quit IRC | 19:35 | |
| *** jog0_ has joined #openstack-dev | 19:35 | |
| *** jog0_ has quit IRC | 19:35 | |
| *** Mandell_ has joined #openstack-dev | 19:36 | |
| *** jog0_ has joined #openstack-dev | 19:39 | |
| *** jog0 has quit IRC | 19:39 | |
| *** jog0_ is now known as jog0 | 19:39 | |
| *** eglynn_ has quit IRC | 19:39 | |
| *** Aaton has quit IRC | 19:42 | |
| *** Aaton has joined #openstack-dev | 19:43 | |
| *** Aaton is now known as EntropyWorks | 19:43 | |
| *** melwitt has joined #openstack-dev | 19:45 | |
| *** eglynn_ has joined #openstack-dev | 19:45 | |
| *** andrewbogott_afk is now known as andrewbogott | 19:47 | |
| *** colinmcnamara has joined #openstack-dev | 19:47 | |
| alexpilotti | mikal: Hi! | 19:47 |
| *** colinmcnamara has quit IRC | 19:48 | |
| openstackgerrit | A change was merged to openstack/quantum: Make tox.ini run pep8 checks on bin. https://review.openstack.org/15764 | 19:48 |
| *** annegentle_ has joined #openstack-dev | 19:48 | |
| *** eglynn__ has joined #openstack-dev | 19:51 | |
| *** eglynn_ has quit IRC | 19:52 | |
| *** dolphm has joined #openstack-dev | 19:54 | |
| *** eglynn has joined #openstack-dev | 19:56 | |
| *** eglynn__ has quit IRC | 19:56 | |
| davidkranz | jaypipes: We need to do something about https://review.openstack.org/#/c/15554/. notmyname didn't like the fact that the test is checking specifically for the documented return code. | 19:57 |
| davidkranz | jaypipes: And it looks like the gate is back up! | 19:57 |
| *** nunosantos has joined #openstack-dev | 20:00 | |
| *** edygarcia has quit IRC | 20:01 | |
| *** maurosr has quit IRC | 20:04 | |
| *** aeperezt has quit IRC | 20:05 | |
| *** jcooley has quit IRC | 20:05 | |
| *** s0mik has quit IRC | 20:05 | |
| *** esp2 has joined #openstack-dev | 20:06 | |
| *** Kaiwei has joined #openstack-dev | 20:06 | |
| *** Kaiwei has left #openstack-dev | 20:06 | |
| *** s0mik has joined #openstack-dev | 20:07 | |
| *** annashen has joined #openstack-dev | 20:07 | |
| *** aeperezt has joined #openstack-dev | 20:08 | |
| *** s0mik has quit IRC | 20:09 | |
| jaypipes | davidkranz: for the swift one, what are you suggesting we do exactly? | 20:12 |
| *** markwash has quit IRC | 20:12 | |
| *** littleidea has quit IRC | 20:13 | |
| *** otherwiseguy has joined #openstack-dev | 20:13 | |
| jaypipes | notmyname: any comment on davidkranz' question re: the documented return codes? | 20:13 |
| *** littleidea has joined #openstack-dev | 20:13 | |
| jaypipes | notmyname: just want to make sure we understand what your expectations are. | 20:13 |
| davidkranz | Well, tempest in general checks for document return codes. John says don't do that, check for 2xx. I'm not sure what the reasoning behind that is. | 20:14 |
| *** jog0 has quit IRC | 20:15 | |
| jaypipes | davidkranz: I suspect because that's what the swift unit tests do... but let's check with him first :) | 20:16 |
| openstackgerrit | A change was merged to openstack/swift: Make DELETE requests to expired objects return 404. https://review.openstack.org/15687 | 20:17 |
| *** vipul|away is now known as vipul | 20:18 | |
| notmyname | davidkranz: jaypipes: I think I marked that on a DELETE, which isn't as big of a deal. but for PUTs, eg, you may not always get a 201. sometimes you may get a 202. the class of 2xx responses is "the request was good and the server did the right thing". (3xx=redirect,4xx=client error,5xx=server error) | 20:19 |
| notmyname | davidkranz: jaypipes: and so the general idea that clients should check for a class of errors is recommended over checking for a particular response code | 20:20 |
| davidkranz | notmyname: Understood. But in tempest we just check for what the spec says. If the spec says, some 2xx code, that is what we check for. | 20:20 |
| davidkranz | notmyname: If the spec says 202, we check for that. | 20:20 |
| davidkranz | notmyname: The point is that a user will see the spec and write code that may depend on that value. | 20:21 |
| davidkranz | notmyname: Tempest is specifically trying to catch changes in specified behavoir. | 20:21 |
| *** jcooley has joined #openstack-dev | 20:22 | |
| notmyname | davidkranz: yes, but I'd actually prefer that the docs mention 2xx rather than 204, for example. (and to some extent, that depends on if you consider those to be a "spec" or a description) | 20:23 |
| davidhadas | hey, anyone from the keystone core team around? | 20:23 |
| dolphm | davidkranz: o/ | 20:24 |
| dolphm | davidhadas: * | 20:24 |
| *** PaulM has joined #openstack-dev | 20:24 | |
| *** PaulM has left #openstack-dev | 20:24 | |
| davidhadas | dolphm: Hi | 20:24 |
| davidkranz | dolphm: Looking for me, or davidhadas ? | 20:24 |
| *** markwash has joined #openstack-dev | 20:24 | |
| *** colinmcnamara has joined #openstack-dev | 20:24 | |
| jaypipes | notmyname: I don't think either davidkranz or I have any objection to making the docs say 2xx, but typically what we do is code the tests in tempest to match the docs... and mark a @skip if there is some temporary change in either the docs/spec and the behaviour of the implementation | 20:24 |
| dolphm | davidkranz: davidhadas | 20:25 |
| jaypipes | hehe | 20:25 |
| *** PaulM has joined #openstack-dev | 20:25 | |
| davidhadas | dolphm: reg tenant anmes and tenant ids - and I know it is all changing in v3 | 20:25 |
| davidhadas | dolphm: but I am trying to understand the past so I can figure out what is changing... :) | 20:25 |
| jaypipes | notmyname: so what we'd do is submit a doc-impact bug to openstack-manuals to s/201/2XX where appropriate and then mod the test case method in question to correspond to the changed spec. | 20:25 |
| dolphm | davidhadas: id/names will certainly stick around | 20:26 |
| vishy | jgriffith: I don't know if you saw but i spent some time tracking down the dd isssue | 20:26 |
| davidhadas | dolphm: when creating a tenat one gives a tenant_Name - than after all other actions are done by tenant_id | 20:26 |
| notmyname | jaypipes: :-) that's where we disagree. I don't consider the docs to be a spec | 20:26 |
| jaypipes | notmyname: so, short story, would you go back on your resistance on that particular point in your comment if we submitted that doc bug? | 20:26 |
| vishy | jgriffith: it only appears to happen with loopback devices so i think we are ok on real hardware | 20:26 |
| davidhadas | dolphm: yet, tenant_name is unique | 20:26 |
| *** john5223 has joined #openstack-dev | 20:26 | |
| davidkranz | notmyname: So where is the spec? | 20:26 |
| jaypipes | notmyname: what do you consider to be a spec? we'll use what that is. | 20:26 |
| davidkranz | notmyname: Please don't say it's the code. | 20:27 |
| *** pabelanger has quit IRC | 20:27 | |
| davidhadas | dolphm: whats troubling me is 1. why use tenant_id at all (when facing the client) - why not use tenant_name - which was teh anme that the tenant had chosen | 20:27 |
| notmyname | jaypipes: I think that if you are eventually going to use 2xx then it doesn't make sense to use 201 right now | 20:27 |
| davidhadas | dolphm: this also introduces a similat problem in swift... but lets start with pure keystone | 20:28 |
| jaypipes | notmyname: the tempest tests are a representation of the spec as it currently exists, not what it will be... | 20:28 |
| jaypipes | notmyname: which is why we're interested in what you consider to be the spec. | 20:29 |
| davidkranz | notmyname: FWIW I think there is a big disconnect between your view about this and http://wiki.openstack.org/APIChangeGuidelines | 20:29 |
| jaypipes | notmyname: and we'll use whatever that is to determine these things if it's not the dfocs | 20:29 |
| *** Dr_Who has quit IRC | 20:29 | |
| notmyname | davidkranz: the docs can't be the spec because the docs follow the code. I don't think there is a "spec document" for swift. in part, this is because "swift" isn't and API that happens to have an openstack implementation. swift is the implementation+the API | 20:29 |
| annegentle_ | jaypipes: just a gentle reminder that all API doc bugs now get logged at https://bugs.launchpad.net/openstack-api-site | 20:30 |
| *** jog0 has joined #openstack-dev | 20:30 | |
| dolphm | davidhadas: valid question... the API (& ID's) are intended to be machine-consumed. we support queries by name, but the CLI got stuck in a rut of requiring an ID as a positional argument.. hoping to change that with openstackclient | 20:30 |
| notmyname | davidkranz: the swift API hasn't changed in about 4 years (things have been added, but existing behavior hasn't changed) | 20:31 |
| jaypipes | annegentle_: ah, cheers. thx. | 20:31 |
| davidhadas | dolphm: Is it agreeable by all that tenant_name is unique? | 20:31 |
| dolphm | davidhadas: that's a popular debate :) | 20:31 |
| davidhadas | dolphm: (It is currently enforced) | 20:31 |
| jaypipes | notmyname: so what is Cloud files then? Swift? the Swift API? the Swift API on specific hardware? | 20:32 |
| dolphm | davidhadas: unique, yes -- but in what context is debatable | 20:32 |
| jaypipes | notmyname: nm, we're getting off track here :) | 20:32 |
| davidkranz | notmyname: I know that and am not expecting it to change. But if not, what's wrong with checking for 201? | 20:32 |
| davidhadas | dolphm: In v2 it is enforced, will domain name be eforced in v3? | 20:32 |
| jaypipes | notmyname: david and I just need to know how to determine what the official API specification for Swift is... | 20:32 |
| davidhadas | dolphm: I mean domain name uniqeness | 20:32 |
| *** esp2 has quit IRC | 20:32 | |
| dolphm | davidhadas: for example, we use UUID's as ID's in part to ensure that two isolated keystone deployments won't collide... names could collide there today | 20:32 |
| notmyname | jaypipes: cloud files is a product built on swift that includes vale add feature and integration with a company's larger product suite ;-) | 20:32 |
| jaypipes | notmyname: :) | 20:32 |
| dolphm | davidhadas: yes, domain names are definitely unique | 20:32 |
| davidhadas | dolphm: at least we have a starting point :) | 20:33 |
| dolphm | davidhadas: in v3, we've talked about tenant/project names only being unique within their owning domain, although that probably won't happen in v3, probably more like v3.1 | 20:33 |
| notmyname | /phone call | 20:33 |
| jaypipes | k | 20:33 |
| davidhadas | dolphm: I certenly do not think project names or user names or any name should be unique between domains! | 20:34 |
| davidhadas | dolphm: this is what a domain is all about -it is a namespace (and a secutiry zone etc) | 20:34 |
| *** pabelanger has joined #openstack-dev | 20:35 | |
| *** esp2 has joined #openstack-dev | 20:35 | |
| davidhadas | dolphm: What I means is that domain.project should be unique but you could have domain1.,yproject and domain2.myproject, and you could have domain1.myuser and domain2.myuser etc. | 20:36 |
| dolphm | davidhadas: the complication is that you'll also need to specify your domain name to authenticate, which has impact across openstack... so we're not making that jump today | 20:36 |
| davidhadas | dolphm: You need to specify you doamin name in every API - right - this what makes it a domain :) | 20:37 |
| davidhadas | dolphm: but if this is where you guys are heading, it is good news to me at least :) | 20:37 |
| dolphm | davidhadas: the impact of introducing domains in v3 is entirely limited to keystone, and it will actually be optional out of the box (we'll create a default domain for you with a migration, and you can stick with that if you'd like) | 20:38 |
| dolphm | (after all, not everyone has a use case for domains) | 20:38 |
| davidhadas | dolphm: ok, so this is exactly where I am - trying to figure out how this affects swift :) | 20:38 |
| davidhadas | dolphm: but I would like to go back to names vs. IDs | 20:39 |
| davidhadas | dolphm: if a domain name is unique, doe sit really make sense to have a domain id? | 20:39 |
| dolphm | davidhadas: we'll start making other projects aware of domains via middleware (e.g. X_DOMAIN_ID X_DOMAIN_NAME) but there's no reason why swift would suddenly *have* to use that data suddenly | 20:39 |
| davidhadas | dolphm: well, v3 eliminates tenants right? | 20:40 |
| dolphm | davidhadas: i'd rather domain names be unique, freeform unicode, and keep system-generated ID's as consistent, reliable, guaranteed-to-be-URL-friendly identifiers | 20:40 |
| dolphm | davidhadas: our impl doesn't allow you to change ID's either, whereas names are mutable | 20:41 |
| dolphm | davidhadas: v3 does *not* eliminate tenants | 20:41 |
| dolphm | davidhadas: just renamed them to "projects" | 20:41 |
| davidhadas | dolphm: ok, so you bring up a new argument - stating that one may choose a name which is not url freindly - is this the reason why we have ids? | 20:41 |
| dolphm | davidhadas: it's one, yes | 20:42 |
| davidhadas | dolphm: any other? | 20:42 |
| *** colinmcnamara has quit IRC | 20:42 | |
| openstackgerrit | A change was merged to openstack/nova: Simplify topic handling in network rpcapi. https://review.openstack.org/15636 | 20:43 |
| dolphm | davidhadas: id's are currently url-friendly, system-generated, immutable, and unique even among isolated deployments | 20:43 |
| openstackgerrit | A change was merged to openstack/nova: Remove unused functions. https://review.openstack.org/15651 | 20:43 |
| dolphm | davidhadas: names are the exact opposite- freeform, user-supplied & mutable | 20:44 |
| openstackgerrit | A change was merged to openstack/nova: Validate is-public parameter to flavor creation https://review.openstack.org/15055 | 20:44 |
| davidhadas | dolphm: but at the same time you force the client to use a system gerenated ID instead of his own | 20:44 |
| dolphm | davidhadas: no- the current CLI implementation forces that | 20:45 |
| davidhadas | dolphm: So whats the purpose of providing a tenantd_name to begin with? | 20:45 |
| dolphm | davidhadas: but you're otherwise correct... that pattern on the CLI is completely broken | 20:46 |
| davidhadas | dolphm: I see - so a tenant name is unique - within a keystone but not between keystones and a concern is two keystones may serve the same openstack cloud and choose allow the nsame name to be created | 20:46 |
| dolphm | davidhadas: ID's *should* ideally be completely hidden from users | 20:46 |
| davidhadas | dolphm: I would vote for hidden IDs | 20:46 |
| davidhadas | dolphm: now for swift in v2 | 20:47 |
| dolphm | davidhadas: on your last point, federation is the primary concern | 20:47 |
| davidhadas | dolphm: ok | 20:47 |
| davidhadas | dolphm: swith - the tenant_id is today being used by swift as a "reseller prefix" | 20:48 |
| dolphm | davidhadas: ah, you're wondering if that should be the domain_id moving forward? | 20:48 |
| davidhadas | dolphm: yes - I am wondering 1. why tenant_id and not tenant)name in v2 and 2. what will it be in the upcoming release :) | 20:49 |
| *** juice has joined #openstack-dev | 20:49 | |
| *** crandquist has quit IRC | 20:49 | |
| *** crandquist has joined #openstack-dev | 20:49 | |
| *** gyee has quit IRC | 20:49 | |
| davidhadas | dolphm: back to v2 - why tenant_id and not tenant_name | 20:49 |
| dolphm | davidhadas: tenant_id will map directly to project_id ... i'm sure we'll provide the same value via middleware under both keys | 20:50 |
| *** apevec has joined #openstack-dev | 20:50 | |
| *** apevec has joined #openstack-dev | 20:50 | |
| dolphm | davidhadas: i think swift used tenant_name at some point, and switched to tenant_id? | 20:50 |
| dolphm | davidhadas: ~diablo | 20:50 |
| davidhadas | dolphm: but why id - this is horrific | 20:50 |
| dolphm | davidhadas: why is it a prefix? | 20:51 |
| *** e1mer has joined #openstack-dev | 20:51 | |
| ayoung | dolphm, I was just looking at https://review.openstack.org/#/c/12111/4 . I think that is going to depend on https://review.openstack.org/#/c/15770/ | 20:51 |
| davidhadas | dolphm: you force a 74398723863285809483298493 id as part of any object url instead of the human readble tenant_name | 20:51 |
| dolphm | davidhadas: there's a conversation on the mailing list about that right now | 20:51 |
| dolphm | ayoung: ha, that review is abandoned, and i just did the same work in your other review | 20:52 |
| ayoung | dolphm, did you submit that ? | 20:52 |
| dolphm | ayoung: https://review.openstack.org/#/c/15724/ <-- the dependencies there are exactly as you'd expect | 20:52 |
| dolphm | ayoung: ripe for your approval :P | 20:53 |
| ayoung | dolphm, great. Still want to put the extra check in first. | 20:53 |
| davidhadas | dolphm: whats the subject used for the discussion in the mailing list? | 20:53 |
| ayoung | dolphm, https://review.openstack.org/#/c/15769/ | 20:54 |
| ayoung | that makes sure that we can insert data into the table. A minor check, but I also use the logic later in the upgrade tests | 20:54 |
| dolphm | ayoung: sold +2 | 20:54 |
| dolphm | davidhadas: [openstack-dev] Specifying Tenant-ID in Openstack REST API URLs and Quatum 2.0 APIs | 20:55 |
| ayoung | dolphm, Let me give yours a little close look. There were some details around pkeys and the such that I want to make sure we get right | 20:55 |
| dolphm | ayoung: i'm also making intra-driver foreignkey's proper foreignkeys in the domain review, as you asked | 20:57 |
| ayoung | Not sure if I still stand by that. It does mean that those drivers all need to be in sync, I think. If you do that, and use LDAP, will it break the other domains. | 20:57 |
| ayoung | ? | 20:57 |
| dolphm | ayoung: *intra*-driver foreign keys? | 20:58 |
| ayoung | ah, ok | 20:58 |
| dolphm | ayoung: not catalog to policy, for example | 20:58 |
| ayoung | I just was thinking like catalog onto tenants type stuff. Intra is correct. | 20:58 |
| dolphm | ayoung: cool | 20:59 |
| *** dprince has quit IRC | 21:04 | |
| *** alexpilotti has quit IRC | 21:04 | |
| davidhadas | dolphm: why do you consider the swift reseller prefix as a project under v3 - isnt it a domain? | 21:04 |
| dolphm | davidhadas: i'm just saying existing tenant_id's are going to literally be project_id's | 21:05 |
| *** thingee has joined #openstack-dev | 21:05 | |
| dolphm | davidhadas: however, the "billable entity" in v2 is the tenant (hence the name), whereas in v3 (i imagine) that will become the domain for most people | 21:06 |
| *** s0mik has joined #openstack-dev | 21:06 | |
| davidhadas | dolphm: yes - I think v3 domains is closer to what I call a tennat :) - we would need domains to be unique and hve their own seperate namespaces in orde rto be a real tenant (under my terminology) | 21:08 |
| *** jimfehlig has quit IRC | 21:09 | |
| gatuus | One more on windows passwords provisioning... How to do that??? With linux images a Got a SSH key.. but in windows?? what to do??? | 21:09 |
| dansmith | gatuus: install linux!!! | 21:09 |
| gatuus | I need to... we are going to offer windows!! on our OScloud | 21:10 |
| dansmith | :) | 21:10 |
| gatuus | HP does it.. how? I don't know! | 21:10 |
| davidhadas | dolphm: I guess under v3 - pepole who seek to offer sharing between groups of users would use 'projects' and those who require isolation between groups work with 'domains' | 21:10 |
| *** s0mik has quit IRC | 21:11 | |
| *** danwent has quit IRC | 21:11 | |
| *** dolphm has quit IRC | 21:11 | |
| *** s0mik has joined #openstack-dev | 21:12 | |
| *** lloydde has joined #openstack-dev | 21:13 | |
| *** jog0 has quit IRC | 21:13 | |
| *** jog0 has joined #openstack-dev | 21:14 | |
| *** jog0 has quit IRC | 21:14 | |
| *** jog0 has joined #openstack-dev | 21:14 | |
| *** s0mik has quit IRC | 21:16 | |
| vishy | ayoung: ping | 21:17 |
| vishy | jerdfelt: ping | 21:17 |
| ayoung | vishy, yes | 21:18 |
| vishy | ayoung: so using eventlet.green.subprocess is a necessary workaround for this bug: | 21:18 |
| *** markvoelker has quit IRC | 21:19 | |
| vishy | https://bitbucket.org/which_linden/eventlet/pull-request/24/fix-waitpid-returning-0-0-and-add-test/diff | 21:19 |
| ayoung | vishy, yes. Which is what I posted for review. | 21:19 |
| vishy | ayoung: oh i haven't seen it yet | 21:19 |
| vishy | (i'm not subscribed to keystone reviews) link? | 21:20 |
| ayoung | vishy, it was a while ago...let me post | 21:20 |
| ayoung | villep, https://review.openstack.org/#/c/15429/ | 21:20 |
| ayoung | vishy ^^ | 21:20 |
| vishy | oh eventlet merged my upstream fix :) | 21:20 |
| vishy | ayoung: eww, that is nasty | 21:21 |
| ayoung | vishy, as it should be. We are so dependent on Eventlet, we need to be good community members | 21:21 |
| vishy | ayoung: we have to do that in all workers | 21:21 |
| ayoung | vishy, So is eventle. | 21:22 |
| *** japage has quit IRC | 21:22 | |
| vishy | ayoung: you didn't change it in auth_middleware | 21:22 |
| ayoung | vishy, no. I want to be able to run auth_token middleware in Apache. I think Apache should be the default web server over time, and I don't want to build in additional eventlet dependencies | 21:23 |
| vishy | ayoung: but you are catching the exception.. | 21:23 |
| vishy | in auth_token | 21:23 |
| ayoung | THat is a a subprocess excpetion from standard, not green | 21:24 |
| vishy | the subprocess exception is not the same ass the eventlet.green subprocess exception | 21:24 |
| ayoung | vishy, I think they are. One sec | 21:24 |
| ayoung | vishy, https://bitbucket.org/which_linden/eventlet/src/112a73783baf6027cd727fad33b4b4928571b173/eventlet/green/subprocess.py?at=default | 21:25 |
| ayoung | It just reuses the standard impls for everything, but does its own thing with Popen | 21:25 |
| ayoung | vishy, I think the magic is in line 10 or so | 21:26 |
| vishy | ayoung: hmm, just tried this | 21:27 |
| vishy | ayoung let me try again | 21:27 |
| *** e1mer has quit IRC | 21:28 | |
| *** utlemming has quit IRC | 21:28 | |
| *** steveleon has joined #openstack-dev | 21:28 | |
| *** maoy has quit IRC | 21:29 | |
| *** blamar has joined #openstack-dev | 21:29 | |
| notmyname | ttx: ping (re milestone-proposed branch) | 21:30 |
| vishy | ayoung: yeah, no worky | 21:31 |
| *** imsplitbit has joined #openstack-dev | 21:32 | |
| *** mnewby_ has joined #openstack-dev | 21:32 | |
| *** maurosr has joined #openstack-dev | 21:33 | |
| notmyname | ttx: nm | 21:33 |
| ayoung | vishy, hrm. | 21:33 |
| openstackgerrit | A change was merged to openstack/swift: Make swift-bench use less memory with large object sizes. https://review.openstack.org/15677 | 21:33 |
| *** mnewby has quit IRC | 21:34 | |
| *** mnewby_ is now known as mnewby | 21:34 | |
| vishy | ayoung: ok i see it works with your fix because you are raising subprocess.CalledProcessError | 21:34 |
| *** sniperd has quit IRC | 21:35 | |
| vishy | ayoung: but with your fix we have to modify every api executable to monkeypatch subprocess | 21:36 |
| ayoung | vishy, which is good. | 21:36 |
| vishy | which is really nasty | 21:36 |
| ayoung | THat way we audit all of the eventlet usage and get it into one place. | 21:36 |
| ayoung | Then we light that place on fire. | 21:36 |
| ayoung | Seriously, anything we do that is eventlet specific needs to be in one part of the code. | 21:37 |
| vishy | ayoung: i think we can make auth_token_middleware detect if anything is monkeypatched | 21:37 |
| ayoung | I tried | 21:37 |
| *** maurosr has quit IRC | 21:37 | |
| ayoung | it has to happen after load time;\ | 21:37 |
| *** blamar has quit IRC | 21:37 | |
| ayoung | vishy, if you try to do it when the code is initially loaded, it fails as it has not been monkeypatched yet | 21:37 |
| *** maurosr has joined #openstack-dev | 21:38 | |
| ayoung | We can't just check for the presence of eventlet on the system | 21:38 |
| ayoung | so, while I am open to a better solution, this was the best I found. | 21:38 |
| *** pabelanger has quit IRC | 21:39 | |
| russellb | devananda: first baremetal patch approved | 21:39 |
| *** saurabhs has left #openstack-dev | 21:39 | |
| * devananda dances | 21:39 | |
| devananda | gah. rebase ... | 21:40 |
| russellb | heh | 21:40 |
| russellb | devananda: you up for doing it now? | 21:40 |
| devananda | yep | 21:40 |
| *** Dorogs has quit IRC | 21:41 | |
| russellb | if it's a trivial rebase (no conflicts require code change) it'll automatically re-add the +2s | 21:41 |
| devananda | taking a look | 21:41 |
| russellb | k | 21:41 |
| devananda | i haven't edited someone else's review before so may take a bit extra to figure that bit out | 21:41 |
| russellb | it doens't seem to automatically re-hit approve, so ping me | 21:41 |
| russellb | nothing special needed really ... use the git command it gives you to pull down the patch for that review | 21:41 |
| russellb | rebase and run "git review" | 21:42 |
| devananda | k | 21:42 |
| russellb | author of the commit is retained unless you tell git to change it | 21:42 |
| russellb | it'll show that you uploaded it, but it's fine, we do it all the time in gerrit | 21:42 |
| *** eharney has quit IRC | 21:42 | |
| ayoung | vishy, so I think the "right" fix is to make Popen separately monkey_patchable, which is, again, an upstream eventlet fix | 21:43 |
| *** maurosr has quit IRC | 21:43 | |
| *** gatuus has quit IRC | 21:43 | |
| *** Tross has joined #openstack-dev | 21:44 | |
| ayoung | I thik the problem is that the Popen code is written right now to require the origianal Popen code to correctly resolve. When I tried monkeypatching it away, I got an endless recursion scenario | 21:44 |
| devananda | russellb: non-trivial changes in scheduler/host_manager | 21:45 |
| *** maurosr has joined #openstack-dev | 21:45 | |
| *** jaypipes has quit IRC | 21:46 | |
| russellb | devananda: ah, k. well i'm about to be away for a bit, but i'll look at what changed once i get back and +2 again if everything looks sane | 21:46 |
| devananda | k | 21:46 |
| ayoung | I think, If I under stand this comment correctly, the problem was that they don't want to rewrite subprocess.call and subprocess.check_call https://bitbucket.org/which_linden/eventlet/src/112a73783baf6027cd727fad33b4b4928571b173/eventlet/green/subprocess.py?at=default#cl-67 | 21:46 |
| *** radez is now known as radez_g0n3 | 21:46 | |
| * mordred apologizes for the flood of patches he's submitting - he just got off a plane | 21:48 | |
| *** anniec72 has joined #openstack-dev | 21:48 | |
| *** echohead has quit IRC | 21:50 | |
| *** yidclare has joined #openstack-dev | 21:51 | |
| vishy | ayoung: the following works: http://paste.openstack.org/show/25341/ | 21:51 |
| vishy | fyi | 21:51 |
| *** anniec72 has quit IRC | 21:52 | |
| vishy | ayoung: i don't think the late load is even necessary if monkeypatch is called before other impoarts | 21:53 |
| vishy | ayoung: testing now | 21:53 |
| ayoung | vishy why except cms.subprocess.CalledProcessError ? | 21:54 |
| vishy | so i don't have to do the same logic in auth_token | 21:54 |
| ayoung | vishy, but by the time this code is executed, won;t the right subprocess be in this .pyc's globalnamespace? | 21:55 |
| *** dani4571 has joined #openstack-dev | 21:55 | |
| ayoung | ah, OK I see, different file | 21:55 |
| vishy | yeah | 21:55 |
| ayoung | OK, that is cool | 21:55 |
| vishy | fyi it works without the special method | 21:55 |
| *** gatuus has joined #openstack-dev | 21:56 | |
| ayoung | vishy, I'm game for a better solution. | 21:56 |
| vishy | ayoung: this one ok? http://paste.openstack.org/show/25342/ | 21:56 |
| ayoung | vishy, I don't think so | 21:57 |
| vishy | i just verified that it works fyi | 21:57 |
| ayoung | IIUC that is going to fail, as that code might get evaluated before the monkeypatch, depending on how it is imported | 21:57 |
| ayoung | I tried something like that in Keystone and it didn't work, as the cms file was in the import chain before monkeypatch was called | 21:58 |
| *** roge has quit IRC | 21:58 | |
| *** lloydde has quit IRC | 21:59 | |
| *** lloydde has joined #openstack-dev | 21:59 | |
| *** lloydde has quit IRC | 22:00 | |
| *** lloydde has joined #openstack-dev | 22:00 | |
| vishy | ayoung monkeypatch should be called first in all binaries | 22:01 |
| vishy | any other usage is likely to fail | 22:01 |
| *** roge has joined #openstack-dev | 22:01 | |
| vishy | all of our binaries do: | 22:01 |
| vishy | import eventlet | 22:01 |
| vishy | eventlet.monkey_patch() | 22:02 |
| vishy | as the first and second lines | 22:02 |
| ayoung | vishy, in the case of Keystone (the only one I checked) we import all of the files starting at keystone-all. Then we parse the command line parameters, because of patching thread messing up the debugger | 22:02 |
| vishy | but if you want the safer ensure_subprocess that is ok | 22:02 |
| vishy | ayoung: sigh that sucks well we can use my late import version if you prefer | 22:03 |
| ayoung | yeah, I think that is the best of a bad situation | 22:03 |
| vishy | another way to do it is to only import subprocess if it isn't in sys.modules so if one of the callers has imported subprocess it will use that version | 22:03 |
| *** echohead has joined #openstack-dev | 22:03 | |
| *** lloydde_ has joined #openstack-dev | 22:04 | |
| *** lloydde has quit IRC | 22:05 | |
| ayoung | lets go with your late import version. I think it is less 'magic' | 22:05 |
| annegentle_ | anyone going on a word journey with me? What are calculs? http://docs.openstack.org/trunk/openstack-compute/admin/content/ceph-rados.html | 22:05 |
| ayoung | vishy, are greenthreads and Eventlet really the right way to develop this kind of application? | 22:06 |
| ayoung | annegentle_, they are people elected to do Calculus, of course | 22:07 |
| *** danwent has joined #openstack-dev | 22:07 | |
| * annegentle_ boggles some more | 22:07 | |
| annegentle_ | I'm really boggling not googling | 22:08 |
| vishy | ayoung: hmm actually that doesn't seem to work i suppose some other dependency is importing subprocess somewhere | 22:09 |
| *** boden has quit IRC | 22:10 | |
| *** rnirmal has quit IRC | 22:10 | |
| *** samkottler is now known as samkottler|bbl | 22:11 | |
| *** Dr_Who has joined #openstack-dev | 22:13 | |
| *** Dr_Who has quit IRC | 22:13 | |
| *** Dr_Who has joined #openstack-dev | 22:13 | |
| vishy | ayoung: that is an interesting question but we don't have an alternative yet. | 22:13 |
| vishy | ayoung: https://review.openstack.org/15791 | 22:13 |
| *** jcooley has quit IRC | 22:14 | |
| *** edygarcia has joined #openstack-dev | 22:15 | |
| *** dolphm has joined #openstack-dev | 22:18 | |
| *** markwash has quit IRC | 22:18 | |
| ayoung | vishy, Actually, we do. We can run in Apache HTTPD. It gets us a lot (as I have ranted about on more than one occasion) but I want to make sure I am not missing anything that drives us toward Eventlet. Eventlet seems to be a bad match for our usage patterns. | 22:19 |
| vishy | for the api worker you could be correct | 22:19 |
| *** gyee has joined #openstack-dev | 22:19 | |
| *** esheffield has quit IRC | 22:20 | |
| vishy | but using eventlet has the advantage of being easily runnable and matching the other workers... | 22:20 |
| ayoung | vishy, I guess it kind of gets into that whole "Threading in Python is ugly" discussion. | 22:21 |
| *** gyee has quit IRC | 22:22 | |
| *** dolphm has quit IRC | 22:23 | |
| *** gyee has joined #openstack-dev | 22:26 | |
| ayoung | vishy, I'm guessing that you change failed due to what happens in keystone-all. I think you need to disable the monkeypatch of Popen in there, too | 22:28 |
| vishy | ayoung: ah ok | 22:28 |
| ayoung | vishy, and pep8 | 22:28 |
| vishy | ayoung: are you ok using regular popen in the rest of keystone? | 22:28 |
| ayoung | run_test.sh --pep8 | 22:28 |
| ayoung | vishy, let me check | 22:29 |
| ayoung | vishy, yeah, none of the rest is called at runtime. Just from utilites and packing and docs etc | 22:30 |
| devananda | russellb: still working on the rebase. lots of conflicts in tests/ due to https://review.openstack.org/#/c/15716/ | 22:33 |
| *** EntropyWorks has quit IRC | 22:36 | |
| *** steveleon has quit IRC | 22:36 | |
| openstackgerrit | A change was merged to openstack/nova: Removed two unused imports. https://review.openstack.org/14934 | 22:39 |
| *** markmcclain has quit IRC | 22:43 | |
| *** spiffxp has joined #openstack-dev | 22:44 | |
| *** jcru is now known as jcru|away | 22:45 | |
| ayoung | vishy, also, unit tests are failing with that change....hmm | 22:45 |
| vishy | ayoung: will look at unit tests next | 22:46 |
| ayoung | vishy, I am getting a 401 from the client test. Are you seeing that? | 22:47 |
| *** EntropyWorks has joined #openstack-dev | 22:47 | |
| *** aravindp has joined #openstack-dev | 22:48 | |
| *** alunduil has quit IRC | 22:48 | |
| *** john5223 has quit IRC | 22:50 | |
| ayoung | vishy, if patcher.already_patched.get('os'): seems to be failing somehow. | 22:50 |
| ayoung | vishy, ah, ok, so that if statement is evaluating to false for me | 22:51 |
| ayoung | and since there is no corresponding else, no import | 22:51 |
| vishy | ayoung: ah yeah i only catch instead of dealing with false | 22:52 |
| ayoung | vishy, I think that, with a small addition I can get your change to work, it might not be getting exercised by the unit test here. | 22:52 |
| ayoung | vishy, I'm going to post an updated patch. I fixed pep 8 as well | 22:52 |
| devananda | russellb: just about done with the rebase | 22:53 |
| vishy | ayoung: | 22:53 |
| vishy | i just reposted | 22:53 |
| vishy | ayoung: see if you did the same thing | 22:53 |
| vishy | ayoung dangit you clobbered me :) | 22:54 |
| ayoung | vishy, yours is wrong anyway | 22:54 |
| ayoung | you left in the call to monkeypatch_cms | 22:55 |
| ayoung | I also left in an import | 22:55 |
| ayoung | I'll take out import keystone common cms | 22:55 |
| vishy | ayoung: yeah we both had an extra line :) | 22:55 |
| ayoung | yeah, but mine ran...heh | 22:56 |
| vishy | good point | 22:56 |
| vishy | :o | 22:56 |
| ayoung | vishy, I'll let you post the updated version. Just, before you do, run bin/keystone-all from the command line, too | 22:57 |
| vishy | ayoung: no its fine you can post | 22:58 |
| ayoung | OK | 22:58 |
| vishy | yours was one line away :) | 22:58 |
| *** dspano has quit IRC | 22:59 | |
| ayoung | vishy, just made sure keystone still worked. I'm going to fire up devstack and see what happens, too. | 23:01 |
| *** nati_ueno_2 has joined #openstack-dev | 23:01 | |
| *** alexpilotti has joined #openstack-dev | 23:03 | |
| *** Dr_Who has quit IRC | 23:03 | |
| *** littleidea has quit IRC | 23:05 | |
| *** steveleon has joined #openstack-dev | 23:05 | |
| *** otherwiseguy has quit IRC | 23:06 | |
| *** annegentle_ has quit IRC | 23:08 | |
| *** jcru|away is now known as jcru | 23:08 | |
| *** edygarcia has quit IRC | 23:09 | |
| *** sunilthaha has joined #openstack-dev | 23:10 | |
| vishy | jgriffith: fyi, i tested the dm-zero strategy and the performance is terrible | 23:11 |
| vishy | jgriffith: so unless I'm doing something really silly, that will not work as a strategy | 23:11 |
| *** kobier has joined #openstack-dev | 23:14 | |
| *** crandquist has quit IRC | 23:15 | |
| *** arata has joined #openstack-dev | 23:16 | |
| *** lloydde_ has quit IRC | 23:18 | |
| *** lloydde has joined #openstack-dev | 23:19 | |
| *** littleidea has joined #openstack-dev | 23:19 | |
| *** EntropyWorks_ has joined #openstack-dev | 23:21 | |
| *** EntropyWorks_ has quit IRC | 23:21 | |
| *** lloydde has quit IRC | 23:24 | |
| *** jcru has quit IRC | 23:27 | |
| *** roge has quit IRC | 23:28 | |
| *** reed has quit IRC | 23:31 | |
| *** jcooley has joined #openstack-dev | 23:31 | |
| *** s0mik has joined #openstack-dev | 23:32 | |
| *** jcru has joined #openstack-dev | 23:33 | |
| *** andrewbogott is now known as andrewbogott_afk | 23:34 | |
| zodiak | hrm. anyone good with tempest around here ? wondering why I can't jst run; nosetests -v tempest.tests.compute.servers.test_server_actions | 23:36 |
| openstackgerrit | A change was merged to openstack/swift: fix error with setting max age and incorrect tests https://review.openstack.org/15794 | 23:36 |
| zodiak | well, I mean, I know why, because flavor is never set anywhere in the setup, but, I mean rather, what is the 'correct way' to run single tests in tempest then | 23:37 |
| *** gatuus has quit IRC | 23:37 | |
| *** kmansel has quit IRC | 23:38 | |
| *** gongysh has joined #openstack-dev | 23:39 | |
| *** ayoung has quit IRC | 23:42 | |
| *** nati_uen_ has joined #openstack-dev | 23:42 | |
| *** apevec has quit IRC | 23:44 | |
| *** nati_ueno_2 has quit IRC | 23:44 | |
| *** jcru has quit IRC | 23:47 | |
| *** sacharya has quit IRC | 23:48 | |
| *** adjohn has joined #openstack-dev | 23:48 | |
| *** Dr_Who has joined #openstack-dev | 23:50 | |
| *** Dr_Who has quit IRC | 23:50 | |
| *** Dr_Who has joined #openstack-dev | 23:50 | |
| *** alunduil has joined #openstack-dev | 23:50 | |
| *** markwash has joined #openstack-dev | 23:52 | |
| *** nati_uen_ has quit IRC | 23:54 | |
| *** nati_ueno_2 has joined #openstack-dev | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!