| *** haomaiwang has quit IRC | 00:01 | |
| *** browne has joined #openstack-nova | 00:01 | |
| *** doug-fish has joined #openstack-nova | 00:01 | |
| *** haomaiwang has joined #openstack-nova | 00:01 | |
| *** takashin has joined #openstack-nova | 00:05 | |
| *** mylu has quit IRC | 00:09 | |
| *** rfolco has quit IRC | 00:17 | |
| *** daemontool has joined #openstack-nova | 00:17 | |
| *** Swami has quit IRC | 00:18 | |
| *** Swami has joined #openstack-nova | 00:18 | |
| *** mylu has joined #openstack-nova | 00:18 | |
| *** rk4n has quit IRC | 00:18 | |
| *** rk4n_ has joined #openstack-nova | 00:19 | |
| *** hblixt has quit IRC | 00:19 | |
| *** rk4n_ has quit IRC | 00:19 | |
| *** armax has quit IRC | 00:19 | |
| *** armax_ has joined #openstack-nova | 00:19 | |
| *** rk4n has joined #openstack-nova | 00:19 | |
| *** rfolco has joined #openstack-nova | 00:20 | |
| *** rfolco has quit IRC | 00:20 | |
| *** mrkz has quit IRC | 00:20 | |
| *** doug-fish has quit IRC | 00:22 | |
| *** doug-fish has joined #openstack-nova | 00:24 | |
| *** doug-fish has quit IRC | 00:25 | |
| *** doug-fish has joined #openstack-nova | 00:26 | |
| *** vladikr has joined #openstack-nova | 00:27 | |
| *** doug-fish has quit IRC | 00:27 | |
| *** eil397 has quit IRC | 00:27 | |
| *** thorst has joined #openstack-nova | 00:29 | |
| *** doug-fish has joined #openstack-nova | 00:30 | |
| *** thorst_ has joined #openstack-nova | 00:31 | |
| *** thorst has quit IRC | 00:34 | |
| *** thorst_ has quit IRC | 00:34 | |
| *** thorst has joined #openstack-nova | 00:35 | |
| *** BobBall_ has joined #openstack-nova | 00:39 | |
| *** jwang has quit IRC | 00:39 | |
| *** jwang_ has joined #openstack-nova | 00:39 | |
| *** BobBall_ is now known as Guest15184 | 00:39 | |
| *** thorst has quit IRC | 00:39 | |
| *** doug-fish has quit IRC | 00:40 | |
| *** diana_clarke has joined #openstack-nova | 00:41 | |
| *** BobBall has quit IRC | 00:42 | |
| *** jwang_ has quit IRC | 00:43 | |
| *** Jeffrey4l has joined #openstack-nova | 00:44 | |
| *** doug-fish has joined #openstack-nova | 00:45 | |
| *** jwang has joined #openstack-nova | 00:49 | |
| *** jwang_ has joined #openstack-nova | 00:53 | |
| *** jwang has quit IRC | 00:53 | |
| Anticimex | i've just upgraded a cluster from stable/kilo to stable/liberty, and run into something i hope is config trivial or similar: https://gist.github.com/Millnert/43ddd7c9402d94e3032a | 00:56 |
|---|---|---|
| Anticimex | can't create vm, nova complains about not finding a security group for the project. though, it does exist | 00:56 |
| *** Jeffrey4l has quit IRC | 00:57 | |
| Anticimex | are there different versions of them, is nova configured for some new and shiny os-extensions security group or what could be off? don't see any log entries on neutron server | 00:57 |
| gus | hemna: pong (sorry for delay - meetings :( ) | 00:57 |
| Anticimex | + when the error is written in nova's log. /me continues checking configs | 00:57 |
| *** rk4n has quit IRC | 00:59 | |
| Anticimex | ah, missing "security_group_api", seems related | 00:59 |
| *** haomaiwang has quit IRC | 01:01 | |
| *** browne has quit IRC | 01:01 | |
| *** haomaiwang has joined #openstack-nova | 01:01 | |
| *** vladikr has quit IRC | 01:02 | |
| *** browne has joined #openstack-nova | 01:03 | |
| gus | hemna: replied via gerrit. Unfortunately you need the latest oslo.privsep from git - or start your servers with some safe non-empty --config-dir value (I screwed up my handling of CONF.config_dir==None). | 01:06 |
| doffm | jaypipes: Stupid question about resource pools... How do they relate to resource providers? Is a resource pool a resource provider? | 01:09 |
| Anticimex | (security_group_api=neutron was it, excuse my noise) | 01:09 |
| *** doug-fish has quit IRC | 01:13 | |
| *** thorst has joined #openstack-nova | 01:15 | |
| *** Daisy has joined #openstack-nova | 01:16 | |
| *** zenoway has joined #openstack-nova | 01:17 | |
| *** funeutron has joined #openstack-nova | 01:18 | |
| *** Daisy has quit IRC | 01:20 | |
| *** Sukhdev has quit IRC | 01:20 | |
| *** zenoway has quit IRC | 01:21 | |
| jaypipes | doffm: all resource pools are resource providers, yes. not all resource providers are resource pools, though. compute nodes are resource providers but not resource pools. resource pools provide a single type of shared resource to multiple consumers. compute nodes provider multiple types of resources to a multiple consumers. | 01:21 |
| *** su_zhang has quit IRC | 01:25 | |
| doffm | jaypipes: Thanks for the patient explanation. So will the resource_pool table have a foreign key link to an entry in the resource_provider table? | 01:27 |
| jaypipes | doffm: yup. | 01:28 |
| *** mtanino has quit IRC | 01:34 | |
| *** Sukhdev has joined #openstack-nova | 01:35 | |
| *** nic has quit IRC | 01:39 | |
| *** Sukhdev has quit IRC | 01:51 | |
| *** zenoway has joined #openstack-nova | 01:52 | |
| *** ducttape_ has joined #openstack-nova | 01:53 | |
| *** apoorvad has quit IRC | 01:54 | |
| *** zenoway has quit IRC | 01:57 | |
| *** sc68cal has quit IRC | 01:57 | |
| *** haomaiwang has quit IRC | 02:01 | |
| *** haomaiwang has joined #openstack-nova | 02:01 | |
| *** ducttape_ has quit IRC | 02:05 | |
| *** ducttape_ has joined #openstack-nova | 02:05 | |
| *** _ducttape_ has joined #openstack-nova | 02:08 | |
| *** ducttape_ has quit IRC | 02:09 | |
| *** doug-fish has joined #openstack-nova | 02:14 | |
| *** Swami has quit IRC | 02:16 | |
| *** doug-fish has quit IRC | 02:19 | |
| *** doug-fish has joined #openstack-nova | 02:25 | |
| *** xyang1 has quit IRC | 02:26 | |
| *** erlon has quit IRC | 02:26 | |
| *** browne has quit IRC | 02:27 | |
| *** zenoway has joined #openstack-nova | 02:28 | |
| *** Marga_ has quit IRC | 02:30 | |
| *** mylu has quit IRC | 02:32 | |
| *** zenoway has quit IRC | 02:32 | |
| *** Marga_ has joined #openstack-nova | 02:34 | |
| *** mylu has joined #openstack-nova | 02:35 | |
| *** Marga_ has quit IRC | 02:40 | |
| *** edmondsw has quit IRC | 02:41 | |
| *** doug-fish has quit IRC | 02:45 | |
| *** doug-fish has joined #openstack-nova | 02:46 | |
| *** fawadkhaliq has joined #openstack-nova | 02:47 | |
| *** doug-fish has quit IRC | 02:50 | |
| *** unicell1 has joined #openstack-nova | 02:52 | |
| *** vilobhmm11 has quit IRC | 02:52 | |
| *** unicell has quit IRC | 02:53 | |
| *** armax_ is now known as armax | 02:54 | |
| *** ijw has quit IRC | 02:55 | |
| *** haomaiwang has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-nova | 03:01 | |
| *** _ducttape_ has quit IRC | 03:02 | |
| *** mdrabe has joined #openstack-nova | 03:02 | |
| *** EinstCrazy has joined #openstack-nova | 03:02 | |
| *** ducttape_ has joined #openstack-nova | 03:02 | |
| *** ducttape_ has quit IRC | 03:04 | |
| *** ducttape_ has joined #openstack-nova | 03:05 | |
| *** EinstCrazy has quit IRC | 03:07 | |
| openstackgerrit | Angus Lees proposed openstack/nova: Add os-brick rootwrap filter for privsep https://review.openstack.org/277670 | 03:10 |
| openstackgerrit | Angus Lees proposed openstack/nova: Require privsep-based os-brick https://review.openstack.org/277671 | 03:10 |
| *** browne has joined #openstack-nova | 03:13 | |
| *** mylu has quit IRC | 03:14 | |
| *** thorst has quit IRC | 03:18 | |
| *** toure has quit IRC | 03:18 | |
| *** thorst has joined #openstack-nova | 03:18 | |
| *** unicell has joined #openstack-nova | 03:21 | |
| *** unicell1 has quit IRC | 03:22 | |
| *** dims has joined #openstack-nova | 03:27 | |
| *** thorst has quit IRC | 03:27 | |
| *** dims_ has quit IRC | 03:28 | |
| *** dims has quit IRC | 03:30 | |
| *** ccard_ has joined #openstack-nova | 03:31 | |
| *** doug-fish has joined #openstack-nova | 03:32 | |
| *** vilobhmm11 has joined #openstack-nova | 03:33 | |
| *** mylu has joined #openstack-nova | 03:33 | |
| *** su_zhang has joined #openstack-nova | 03:34 | |
| *** ccard__ has quit IRC | 03:34 | |
| *** achanda has joined #openstack-nova | 03:40 | |
| *** mdrabe has quit IRC | 03:40 | |
| *** links has joined #openstack-nova | 03:41 | |
| *** dims has joined #openstack-nova | 03:43 | |
| *** mylu has quit IRC | 03:45 | |
| *** achanda has quit IRC | 03:45 | |
| *** mylu has joined #openstack-nova | 03:46 | |
| *** dims has quit IRC | 03:47 | |
| *** mylu has quit IRC | 03:49 | |
| *** mylu has joined #openstack-nova | 03:49 | |
| *** ducttape_ has quit IRC | 03:51 | |
| *** ducttape_ has joined #openstack-nova | 03:53 | |
| *** Marga_ has joined #openstack-nova | 03:54 | |
| *** doug-fish has quit IRC | 03:56 | |
| *** doug-fish has joined #openstack-nova | 03:56 | |
| *** doug-fish has quit IRC | 03:57 | |
| *** doug-fis_ has joined #openstack-nova | 04:00 | |
| *** haomaiwang has quit IRC | 04:01 | |
| *** ducttape_ has quit IRC | 04:01 | |
| *** shz has quit IRC | 04:01 | |
| *** haomaiwa_ has joined #openstack-nova | 04:01 | |
| *** shz has joined #openstack-nova | 04:01 | |
| *** doug-fis_ has quit IRC | 04:04 | |
| *** ducttape_ has joined #openstack-nova | 04:09 | |
| *** doug-fish has joined #openstack-nova | 04:09 | |
| *** baoli_ has quit IRC | 04:11 | |
| *** apoorvad has joined #openstack-nova | 04:13 | |
| *** armax_ has joined #openstack-nova | 04:18 | |
| *** apoorvad has quit IRC | 04:19 | |
| *** mylu has quit IRC | 04:19 | |
| *** mylu has joined #openstack-nova | 04:20 | |
| *** armax has quit IRC | 04:21 | |
| *** armax_ is now known as armax | 04:21 | |
| *** thorst has joined #openstack-nova | 04:24 | |
| *** n0ano has quit IRC | 04:24 | |
| *** zenoway has joined #openstack-nova | 04:27 | |
| *** vilobhmm11 has quit IRC | 04:27 | |
| *** vilobhmm11 has joined #openstack-nova | 04:29 | |
| *** zenoway has quit IRC | 04:31 | |
| *** thorst has quit IRC | 04:32 | |
| *** daemontool has quit IRC | 04:32 | |
| *** daemontool has joined #openstack-nova | 04:32 | |
| *** mylu has quit IRC | 04:38 | |
| *** shyama has joined #openstack-nova | 04:40 | |
| *** salv-orl_ has joined #openstack-nova | 04:41 | |
| *** fawadkhaliq has quit IRC | 04:42 | |
| *** mylu has joined #openstack-nova | 04:42 | |
| *** salv-orlando has quit IRC | 04:43 | |
| *** mylu has quit IRC | 04:55 | |
| *** Sukhdev has joined #openstack-nova | 04:57 | |
| *** dulek has joined #openstack-nova | 05:00 | |
| *** haomaiwa_ has quit IRC | 05:01 | |
| *** haomaiwang has joined #openstack-nova | 05:01 | |
| *** sudipto has joined #openstack-nova | 05:02 | |
| *** funeutron has quit IRC | 05:03 | |
| *** mylu has joined #openstack-nova | 05:05 | |
| *** ducttape_ has quit IRC | 05:06 | |
| *** amotoki has joined #openstack-nova | 05:07 | |
| *** jgriffith is now known as jgriffith_away | 05:07 | |
| *** amotoki_ has joined #openstack-nova | 05:09 | |
| *** SridharG has joined #openstack-nova | 05:10 | |
| *** cfriesen has quit IRC | 05:11 | |
| *** amotoki has quit IRC | 05:12 | |
| *** mylu has quit IRC | 05:13 | |
| *** fawadkhaliq has joined #openstack-nova | 05:14 | |
| *** fawadkhaliq has quit IRC | 05:14 | |
| *** fawadkhaliq has joined #openstack-nova | 05:14 | |
| *** dave-mccowan has quit IRC | 05:14 | |
| *** amotoki_ has quit IRC | 05:15 | |
| *** chhavi has joined #openstack-nova | 05:17 | |
| *** mylu has joined #openstack-nova | 05:20 | |
| *** daemontool_ has joined #openstack-nova | 05:21 | |
| *** daemontool has quit IRC | 05:22 | |
| *** Marga_ has quit IRC | 05:24 | |
| *** rk4n has joined #openstack-nova | 05:24 | |
| *** rk4n has quit IRC | 05:29 | |
| *** thorst has joined #openstack-nova | 05:29 | |
| *** sudipto has quit IRC | 05:34 | |
| *** irenab has quit IRC | 05:36 | |
| *** Daisy has joined #openstack-nova | 05:36 | |
| *** thorst has quit IRC | 05:37 | |
| *** SridharG has quit IRC | 05:40 | |
| mikal | https://review.openstack.org/#/c/255516 is a quick and easy review for a core looking for a cheap +2 | 05:42 |
| *** mylu has quit IRC | 05:45 | |
| *** mylu has joined #openstack-nova | 05:46 | |
| *** doug-fish has quit IRC | 05:48 | |
| *** mylu has quit IRC | 05:49 | |
| *** amotoki has joined #openstack-nova | 05:49 | |
| *** sudipto has joined #openstack-nova | 05:50 | |
| *** mylu has joined #openstack-nova | 05:53 | |
| *** mylu has quit IRC | 05:55 | |
| *** doug-fish has joined #openstack-nova | 05:55 | |
| *** Daisy has quit IRC | 05:58 | |
| *** Daisy has joined #openstack-nova | 06:00 | |
| *** haomaiwang has quit IRC | 06:01 | |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova: virt: osinfo will report once if libosinfo is not loaded https://review.openstack.org/277565 | 06:01 |
| *** ljxiash has joined #openstack-nova | 06:01 | |
| *** haomaiwang has joined #openstack-nova | 06:01 | |
| *** sudipto has quit IRC | 06:03 | |
| *** ankit_ag has joined #openstack-nova | 06:04 | |
| *** Marga_ has joined #openstack-nova | 06:06 | |
| *** doug-fish has quit IRC | 06:08 | |
| *** sgordon has quit IRC | 06:09 | |
| *** sgordon has joined #openstack-nova | 06:11 | |
| *** irenab has joined #openstack-nova | 06:15 | |
| *** achanda has joined #openstack-nova | 06:18 | |
| *** sudipto has joined #openstack-nova | 06:19 | |
| *** esracelik has joined #openstack-nova | 06:20 | |
| *** ijw has joined #openstack-nova | 06:27 | |
| *** SridharG has joined #openstack-nova | 06:34 | |
| *** achanda has quit IRC | 06:35 | |
| *** thorst has joined #openstack-nova | 06:35 | |
| *** ekuris has joined #openstack-nova | 06:36 | |
| *** mnestratov|2 has quit IRC | 06:38 | |
| *** thorst has quit IRC | 06:42 | |
| *** vilobhmm11 has quit IRC | 06:43 | |
| *** achanda has joined #openstack-nova | 06:43 | |
| *** armax has quit IRC | 06:49 | |
| *** Daisy has quit IRC | 06:49 | |
| *** vilobhmm11 has joined #openstack-nova | 06:51 | |
| *** Daisy has joined #openstack-nova | 06:53 | |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (2) https://review.openstack.org/276087 | 06:53 |
| *** scheuran has joined #openstack-nova | 06:55 | |
| *** amotoki has quit IRC | 06:56 | |
| *** ljxiash has quit IRC | 06:56 | |
| *** amotoki has joined #openstack-nova | 06:58 | |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (3) https://review.openstack.org/276088 | 06:58 |
| *** achanda has quit IRC | 06:58 | |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (4) https://review.openstack.org/276090 | 06:59 |
| *** zenoway has joined #openstack-nova | 06:59 | |
| *** haomaiwang has quit IRC | 07:01 | |
| *** haomaiwang has joined #openstack-nova | 07:01 | |
| *** rk4n has joined #openstack-nova | 07:02 | |
| *** nkrinner has joined #openstack-nova | 07:03 | |
| *** lpetrut has joined #openstack-nova | 07:03 | |
| *** mjura has joined #openstack-nova | 07:06 | |
| *** Guest496 is now known as tdurakov | 07:06 | |
| tdurakov | hello nova | 07:06 |
| *** mubirru has joined #openstack-nova | 07:08 | |
| *** vilobhmm11 has quit IRC | 07:09 | |
| *** itamarl has joined #openstack-nova | 07:10 | |
| tdurakov | johnthetubaguy,hey, it's absolutely worth to try to *always* do cleanup, it will also fix this: https://bugs.launchpad.net/nova/+bug/1284719 | 07:12 |
| openstack | Launchpad bug 1284719 in OpenStack Compute (nova) "buggy live migration rollback when using shared storage" [Medium,Confirmed] - Assigned to Pushkar Umaranikar (pushkar-umaranikar) | 07:12 |
| *** ljxiash has joined #openstack-nova | 07:12 | |
| *** mubirru_ has joined #openstack-nova | 07:15 | |
| *** paul-carlton1 has joined #openstack-nova | 07:17 | |
| *** mubirru has quit IRC | 07:17 | |
| *** mubirru_ is now known as mubirru | 07:17 | |
| *** Sukhdev has quit IRC | 07:19 | |
| *** sudipto has quit IRC | 07:26 | |
| *** lpetrut has quit IRC | 07:31 | |
| *** sahid has joined #openstack-nova | 07:31 | |
| *** Daisy has quit IRC | 07:32 | |
| *** Daisy has joined #openstack-nova | 07:33 | |
| *** mubirru_ has joined #openstack-nova | 07:34 | |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (5) https://review.openstack.org/276095 | 07:35 |
| *** zenoway has quit IRC | 07:35 | |
| *** mubirru has quit IRC | 07:36 | |
| *** mubirru_ is now known as mubirru | 07:37 | |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (6) https://review.openstack.org/276099 | 07:37 |
| *** Daisy has quit IRC | 07:37 | |
| *** doug-fish has joined #openstack-nova | 07:38 | |
| *** thorst has joined #openstack-nova | 07:40 | |
| *** belmoreira has joined #openstack-nova | 07:41 | |
| *** doug-fis_ has joined #openstack-nova | 07:42 | |
| *** Sesso_ has joined #openstack-nova | 07:42 | |
| *** doug-fish has quit IRC | 07:43 | |
| esracelik | markus_z: hi | 07:43 |
| *** gjayavelu has joined #openstack-nova | 07:44 | |
| *** whenry_ has quit IRC | 07:45 | |
| *** thorst has quit IRC | 07:46 | |
| *** doug-fis_ has quit IRC | 07:47 | |
| tdurakov | markus_z, hey, any updates about ci failures? | 07:49 |
| *** rdopiera has joined #openstack-nova | 07:52 | |
| *** moshele has joined #openstack-nova | 07:53 | |
| *** su_zhang has quit IRC | 07:54 | |
| *** w_verdugo has quit IRC | 07:54 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: add release notes for the limits https://review.openstack.org/252870 | 07:55 |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (1) https://review.openstack.org/274698 | 07:55 |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (2) https://review.openstack.org/276087 | 07:56 |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (3) https://review.openstack.org/276088 | 07:57 |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (4) https://review.openstack.org/276090 | 07:57 |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (5) https://review.openstack.org/276095 | 07:57 |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (6) https://review.openstack.org/276099 | 07:58 |
| *** haomaiwang has quit IRC | 08:01 | |
| *** haomaiwa_ has joined #openstack-nova | 08:01 | |
| *** liverpooler has joined #openstack-nova | 08:03 | |
| *** mvk has joined #openstack-nova | 08:04 | |
| *** ssurana has joined #openstack-nova | 08:11 | |
| *** paul-carlton1 has quit IRC | 08:11 | |
| *** ssurana has left #openstack-nova | 08:13 | |
| *** zakora has joined #openstack-nova | 08:17 | |
| *** browne has quit IRC | 08:18 | |
| *** rotbeard has joined #openstack-nova | 08:23 | |
| *** paul-carlton1 has joined #openstack-nova | 08:23 | |
| *** Jeffrey4l has joined #openstack-nova | 08:23 | |
| *** sinese_ has joined #openstack-nova | 08:24 | |
| *** jlanoux has joined #openstack-nova | 08:27 | |
| *** deepakcs has joined #openstack-nova | 08:27 | |
| *** oomichi has quit IRC | 08:28 | |
| *** matrohon has joined #openstack-nova | 08:31 | |
| *** sudipto has joined #openstack-nova | 08:33 | |
| *** MCoLo has joined #openstack-nova | 08:33 | |
| bauzas | good morning Nova | 08:34 |
| *** zz_dimtruck is now known as dimtruck | 08:37 | |
| *** jaosorior has joined #openstack-nova | 08:37 | |
| *** thorst has joined #openstack-nova | 08:45 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes Python 3 unit tests for nova.compute https://review.openstack.org/247742 | 08:47 |
| *** jichen has joined #openstack-nova | 08:47 | |
| tdurakov | folks, who could tell me why we check this for shared block strage? https://github.com/openstack/nova/blob/f6e2c3327134ae7df5d52934ffef0138985e769b/nova/virt/libvirt/driver.py#L5564-L5568 | 08:48 |
| *** e0ne has joined #openstack-nova | 08:49 | |
| *** mvk has quit IRC | 08:49 | |
| *** dimtruck is now known as zz_dimtruck | 08:51 | |
| *** thorst has quit IRC | 08:53 | |
| *** mpavone has joined #openstack-nova | 08:54 | |
| bauzas | tdurakov: I'm thinking if you use for example NFS | 08:55 |
| *** ndipanov has joined #openstack-nova | 08:55 | |
| bauzas | tdurakov: I mean, take a qcow2 backfile | 08:55 |
| tdurakov | bauzas, well, yep, but it looks strange, as we also check it by temp_file | 08:56 |
| *** paul-carlton1 has quit IRC | 08:56 | |
| bauzas | if the instance path is shared on NFS, then there is a shared storage | 08:56 |
| *** rk4n has quit IRC | 08:56 | |
| bauzas | here, the conditional is about backing files, IIRC ? | 08:56 |
| bauzas | right? | 08:57 |
| tdurakov | bauzas, not sure | 08:57 |
| bauzas | so, that's not about temp files | 08:57 |
| openstackgerrit | Ryan Moore proposed openstack/nova: Add better help to rdp options https://review.openstack.org/253556 | 08:57 |
| * tdurakov checking shared storage flags code | 08:57 | |
| bauzas | no | 08:58 |
| bauzas | say that you use a QCOW2 image | 08:59 |
| *** bkumar has joined #openstack-nova | 08:59 | |
| *** rk4n has joined #openstack-nova | 08:59 | |
| bauzas | by default, it will create a backing file and a second file for writing | 08:59 |
| bauzas | all of them will be in the instance path | 08:59 |
| tdurakov | bauzas, true | 08:59 |
| bauzas | some operators will share that instance path by using a NFS export | 08:59 |
| bauzas | that's one way to have shared storage | 09:00 |
| bauzas | (not volume-backed, to be precise) | 09:00 |
| *** haomaiwa_ has quit IRC | 09:01 | |
| tdurakov | bauzas, there is a check before for rbd in this method, and as for me shared block storage is about block devices, not nfs file based shared storage, no? | 09:01 |
| bauzas | lemme check too | 09:01 |
| *** gjayavelu has quit IRC | 09:01 | |
| *** haomaiwang has joined #openstack-nova | 09:01 | |
| bauzas | so | 09:02 |
| tdurakov | imo, such interference make code a bit harder to understand, working on johnthetubaguy about removing _live_migration_cleanup_flags method from compute manager. | 09:03 |
| bauzas | is_shared_instance_path attribute verifies by _check_shared_storage_test_file that the temp file is shared between source and destination for the same instance path | 09:03 |
| *** rk4n has quit IRC | 09:03 | |
| tdurakov | right | 09:03 |
| bauzas | ie. it writes it on the source and tries to read it on the dest | 09:03 |
| bauzas | so, that means it's a shared storage righT ? | 09:03 |
| *** jaosorior has quit IRC | 09:03 | |
| tdurakov | it means that it's file-based shared storage, i.e. nfs | 09:04 |
| bauzas | because you write a file on a path and you expect it to be readable on another host with the same path | 09:04 |
| tdurakov | yup | 09:04 |
| bauzas | that's a shared storage | 09:04 |
| bauzas | file-based or not | 09:04 |
| bauzas | I mean | 09:04 |
| bauzas | you can see the files to be consistent, but that doesn't mean it's NFS | 09:05 |
| tdurakov | there is actually a big diff between ceph/rbd and nfs, so I'd separate this:) | 09:05 |
| bauzas | so | 09:05 |
| bauzas | whatever the backend is | 09:05 |
| bauzas | you know that CONF.instance_path is shared, that's what you can infer | 09:05 |
| bauzas | now, the second part of the conditional | 09:05 |
| *** claudiub has joined #openstack-nova | 09:06 | |
| bauzas | see the libvirt code | 09:06 |
| bauzas | nova.virt.libvirt.imagebackend | 09:06 |
| tdurakov | checking | 09:07 |
| *** moorryan has quit IRC | 09:07 | |
| bauzas | you'll see that this method is overrided for each backend | 09:07 |
| bauzas | well, not overrided | 09:07 |
| *** PaulMurray has joined #openstack-nova | 09:07 | |
| bauzas | rather implemented | 09:07 |
| bauzas | now you can assert a 2nd thing :) | 09:08 |
| *** jaosorior has joined #openstack-nova | 09:08 | |
| *** jaosorior has quit IRC | 09:08 | |
| *** jaosorior has joined #openstack-nova | 09:09 | |
| bauzas | I mean | 09:09 |
| bauzas | by default, it's False for an image backend but for QCOW2 and RAW (which makes sense...) | 09:09 |
| bauzas | because of what I explained before | 09:09 |
| *** lucas-dinner is now known as lucasagomes | 09:10 | |
| PaulMurray | Hi alex_xu are you around? | 09:10 |
| bauzas | tdurakov: so, here you're testing that CONF.instance_path is shared source and dest and that your image backend stores its backing files on the path | 09:10 |
| bauzas | see ? | 09:11 |
| tdurakov | bauzas, yes, walking through all this code, thanks for details | 09:11 |
| bauzas | np | 09:11 |
| bauzas | thanks for remembering me my old days with OpenStack when I was operating it :) | 09:12 |
| tdurakov | bauzas, welcome:) | 09:12 |
| *** danpb has joined #openstack-nova | 09:13 | |
| *** mvk has joined #openstack-nova | 09:17 | |
| *** e0ne has quit IRC | 09:19 | |
| *** ihrachys has joined #openstack-nova | 09:19 | |
| *** mgoddard has joined #openstack-nova | 09:21 | |
| *** paul-carlton1 has joined #openstack-nova | 09:24 | |
| *** yassine has joined #openstack-nova | 09:26 | |
| *** mvk has quit IRC | 09:29 | |
| kashyap | markus_z: Morning, (recovering after 10 days of travel), looking at this bug, it appears that the person triaging marked it as 'Invalid' following the process "Close as "invalid" if it is a support request or feature request." - https://bugs.launchpad.net/nova/+bug/1509304 | 09:29 |
| openstack | Launchpad bug 1509304 in OpenStack Compute (nova) "Add support for 'l2-cache-size' (a QCOW2 run time option for metadata cache size) for drives" [Undecided,Invalid] | 09:30 |
| *** mvk has joined #openstack-nova | 09:30 | |
| kashyap | markus_z: In this case, as bug has enough information, and clear enough, I think updating the description to RFE & placing it back to 'New' is okay. Lest it can give an impression to those skimming that the bug itself is invalid... | 09:30 |
| kashyap | I'm going update it and add a polite comment saying so. | 09:31 |
| *** jwcroppe has joined #openstack-nova | 09:33 | |
| johnthetubaguy | bauzas: tdurakov: when we start tracking shared storage in the scheduler, we will actually know which hosts shared the instances dir via the aggregates, so I am not sure we need to fixate on that too much | 09:35 |
| bauzas | johnthetubaguy: agreed | 09:36 |
| bauzas | johnthetubaguy: I was just trying to help tdurakov on his comment for https://github.com/openstack/nova/blob/f6e2c3327134ae7df5d52934ffef0138985e769b/nova/virt/libvirt/driver.py#L5564-L5568 | 09:36 |
| tdurakov | johnthetubaguy, when will this feature be implemented? | 09:36 |
| johnthetubaguy | kashyap: unsure, neutron does features through bugs, we generally try to avoid that in Nova | 09:36 |
| *** jwcroppe has quit IRC | 09:36 | |
| bauzas | but I'm not a libvirt expert, just reading code and a previous user :) | 09:36 |
| johnthetubaguy | tdurakov: I think its looking like Newton at this point | 09:36 |
| *** jwcroppe has joined #openstack-nova | 09:37 | |
| kashyap | johnthetubaguy: Yeah, I haven't added a comment yet there - but since the blueprint isn't filed yet, wouldn't the bug fall out of radar as 'Invalid' bugs are ignored? | 09:37 |
| *** efoley has joined #openstack-nova | 09:37 | |
| johnthetubaguy | kashyap: it will, but we don't track un targeted blueprints either, all thats left to google really | 09:37 |
| bauzas | johnthetubaguy: FWIW, I also considered again https://review.openstack.org/#/c/200870/ and I feel it's a bit bad to merge it | 09:38 |
| *** ociuhandu has quit IRC | 09:38 | |
| johnthetubaguy | kashyap: ideas we need to remember are backlog specs, really, I don't think this is one of those | 09:38 |
| kashyap | johnthetubaguy: So, for this, what is the action item? File a blueprint I guess? | 09:38 |
| johnthetubaguy | if its wanted, yes | 09:39 |
| *** daemontool has joined #openstack-nova | 09:39 | |
| kashyap | johnthetubaguy: Yes, at least two large operators say it's wanted. There was a discussion about it on operators-list some months ago. | 09:39 |
| *** derekh has joined #openstack-nova | 09:39 | |
| johnthetubaguy | kashyap: OK, do we have anyone who can work on that, and describe why its important? | 09:39 |
| *** daemontool_ has quit IRC | 09:40 | |
| markus_z | esracelik: hi, good morning, what's up? | 09:40 |
| markus_z | tdurakov: I guess mtreinish solved it with this https://bugs.launchpad.net/nova/+bug/1543253 | 09:41 |
| openstack | Launchpad bug 1543253 in OpenStack Compute (nova) "cells test_rebuild_instance_with_volume Tempest test fails" [Critical,Fix released] - Assigned to Andrew Laski (alaski) | 09:41 |
| kashyap | johnthetubaguy: Why it's important -- see the first two lines of the bug. | 09:41 |
| esracelik | markus_z: fine markus, you? | 09:41 |
| *** jwcroppe has quit IRC | 09:41 | |
| tdurakov | markus_z, cool, thanks for update! | 09:41 |
| *** Daisy has joined #openstack-nova | 09:41 | |
| kashyap | johnthetubaguy: About anyone working on it - not yet, as it's not a big priority item, nontheless a performance improvement. | 09:42 |
| esracelik | markus_z: happy to see that cells options are reviewed, I splitted cells options help texts into 6 patchsets. But I didn't split up configdrive and hyperv options. What do you think about them? | 09:43 |
| johnthetubaguy | kashyap: so, feels like a blueprint would be a better way to track it, I am off to the operators meetup in Manchester next week to see how we can work better with those folks | 09:43 |
| markus_z | kashyap: Tracing feature request with bugs doesn't feel right to me IMO | 09:43 |
| markus_z | esracelik: sec, let me have a look at them | 09:43 |
| *** zenoway has joined #openstack-nova | 09:44 | |
| kashyap | markus_z: johnthetubaguy: Right, I just went back and re-read the thread on operators list, seems like it was my fault - as I asked them to report the bug - http://lists.openstack.org/pipermail/openstack-operators/2015-October/008509.html | 09:45 |
| *** fawadkhaliq has quit IRC | 09:47 | |
| markus_z | johnthetubaguy: to be honest, I don't yet fully understand how such RFEs should be addressed to us. I understood the "spec backlog" thing a bit like the way to go. | 09:47 |
| *** fawadkhaliq has joined #openstack-nova | 09:47 | |
| bauzas | blueprints | 09:47 |
| bauzas | markus_z: http://docs.openstack.org/developer/nova/process.html#overview nope? | 09:48 |
| markus_z | bauzas: AFAIK the blueprint drafter is supposed to drive the implementation too, right? | 09:48 |
| bauzas | markus_z: why so ? | 09:49 |
| bauzas | that's just people are not looking at BPs like they do with bugs | 09:49 |
| bauzas | but if the question is "I have an idea, I want it to be implemented", it certainly needs some level of commitment more than just filling a bug | 09:50 |
| *** thorst has joined #openstack-nova | 09:50 | |
| bauzas | or speak to subteams | 09:50 |
| johnthetubaguy | yeah, we don't have to tie the blueprint author to the implementer | 09:50 |
| markus_z | bauzas: agreed to that | 09:50 |
| johnthetubaguy | problem is, most of the time, people have other things they are interested in more / told to do, so they do those instead | 09:50 |
| *** lpetrut has joined #openstack-nova | 09:51 | |
| bauzas | that's why having people sharing interest in subteams is nice | 09:51 |
| johnthetubaguy | There is a general "performance" question, including os-profiler, and other things in there | 09:51 |
| johnthetubaguy | I really would like a more... focused effort around that area, if people are interested (I am trying to rally some folks for control plane optimisations next release) | 09:51 |
| johnthetubaguy | the real answer is we attack this from many angles | 09:52 |
| johnthetubaguy | and hope one of them sticks | 09:52 |
| johnthetubaguy | which doesn't help the operators too much | 09:52 |
| johnthetubaguy | we did try to create an operators sub team, but not 100% thats working out too well yet | 09:52 |
| bauzas | but just filling a bug for asking a feature seems pretty low-committing | 09:53 |
| bauzas | johnthetubaguy: and given the talk you made at Summit, you know that an idea needs to be discussed | 09:53 |
| bauzas | that's why I think that if someone wants to add a feature, he needs somehow to discuss with some other people | 09:53 |
| bauzas | and not just be like posting an idea | 09:54 |
| *** rk4n has joined #openstack-nova | 09:54 | |
| bauzas | a bug is different, something is broken | 09:54 |
| johnthetubaguy | bauzas: yep, hence the backlog spec | 09:55 |
| johnthetubaguy | that process triggers that debate | 09:56 |
| bauzas | johnthetubaguy: then, we should amend http://docs.openstack.org/developer/nova/process.html#overview | 09:56 |
| bauzas | johnthetubaguy: what I'd like to avoid is having backlog specs a wall of ideas | 09:56 |
| bauzas | and people not committed to them | 09:56 |
| johnthetubaguy | bauzas: yeah, thats the tricky balance we haven't really worked out | 09:56 |
| bauzas | in particular, since we don't do backlog specs per-release | 09:57 |
| *** thorst has quit IRC | 09:57 | |
| johnthetubaguy | the idea is for them to be unassigned | 09:57 |
| bauzas | so the backlog of those backlog specs could dramatically increase | 09:57 |
| johnthetubaguy | I was thinking about culling them over time | 09:57 |
| bauzas | I know what you'd like :) | 09:57 |
| markus_z | esracelik: the configdrive help texts are good to go, the hyper-v ones could be split up I guess. | 09:57 |
| esracelik | markus_z: ok, I will be working on it this afternoon, thanks | 09:58 |
| markus_z | esracelik: cool, thanks. It worthwhile to get input from claudiub on the hyper-v related options. | 09:59 |
| openstackgerrit | jichenjc proposed openstack/nova: Not using mox stubs in nova.tests.unit.volume https://review.openstack.org/268081 | 09:59 |
| openstackgerrit | jichenjc proposed openstack/nova: not use mox for conductor/test_conductor.py https://review.openstack.org/271312 | 10:00 |
| markus_z | johnthetubaguy: bauzas: So, closing RFEs in bug reports and pointing them to the blueprint process is still OK? | 10:00 |
| claudiub | markus_z: esracelik: sure, will do, as soon as I can. | 10:00 |
| openstackgerrit | jichenjc proposed openstack/nova: not use mox for conductor.tasks https://review.openstack.org/272705 | 10:00 |
| *** haomaiwang has quit IRC | 10:01 | |
| esracelik | claudiub: thanks! | 10:01 |
| *** fawadkhaliq has quit IRC | 10:01 | |
| *** haomaiwa_ has joined #openstack-nova | 10:01 | |
| bauzas | markus_z: that's a rough answer, but the short answer is, yes | 10:01 |
| bauzas | markus_z: that said | 10:01 |
| bauzas | markus_z: what we need to explain is that rather just filling a blueprint and waiting for someone to magically take it, operators should at least come to us (in IRC, mail, etc.) and identify some volunteers to sheperd that in the community | 10:02 |
| markus_z | bauzas: yes, that's fair. | 10:03 |
| johnthetubaguy | markus_z: I can't think of a better option, I would love to point them to a specific "here is how to get your idea considered" doc though | 10:03 |
| bauzas | indeed | 10:04 |
| PaulMurray | pkoniszewski, I think your plan to address danpb's comment in https://review.openstack.org/#/c/227278 makes sense. Do you agree danpb ? | 10:04 |
| bauzas | I totally second that | 10:04 |
| johnthetubaguy | markus_z: there are project differences we probably need to iron out here too, Neutron is using bugs to capture feature requests in some way | 10:04 |
| openstackgerrit | jichenjc proposed openstack/nova: Not using mox stubs in nova.tests.unit.consoleauth https://review.openstack.org/266903 | 10:04 |
| bauzas | johnthetubaguy: we also have wishlist bugs | 10:04 |
| bauzas | although we try to refrain that | 10:05 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes network py3 unit tests https://review.openstack.org/253990 | 10:05 |
| johnthetubaguy | bauzas: yeah, its kinda similar, they are more broken things that are features, but its a dumb distinction | 10:05 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes cells py3 unit tests https://review.openstack.org/255824 | 10:05 |
| *** Jeffrey4l has quit IRC | 10:07 | |
| *** Guest15184 is now known as BobBall | 10:07 | |
| *** snikitin has quit IRC | 10:10 | |
| markus_z | johnthetubaguy: Would an additional entry at http://docs.openstack.org/developer/nova/blueprints.html make sense? | 10:11 |
| johnthetubaguy | markus_z: yeah, thats what we need | 10:11 |
| markus_z | johnthetubaguy: regarding the "here is how to get your idea considered" | 10:11 |
| johnthetubaguy | markus_z: yes, thats my proposal, add a new thing for that | 10:11 |
| *** Marga_ has quit IRC | 10:12 | |
| * johnthetubaguy hmm, we have a blueprints page, I was thinking http://docs.openstack.org/developer/nova/process.html | 10:12 | |
| tdurakov | johnthetubaguy, here is pottential issue in post_live_migration step for libvirt: https://github.com/openstack/nova/blob/197bd6dd1231f1f57cdd6c0acb1dfbdc3b2b0989/nova/virt/libvirt/driver.py#L1117 - because is_shared_block_storage is true for nfs either | 10:12 |
| johnthetubaguy | oh, dear, we do | 10:12 |
| johnthetubaguy | thats confusing | 10:12 |
| johnthetubaguy | I should take a look at this later | 10:12 |
| johnthetubaguy | I want to get something I can show the operators on Monday | 10:12 |
| johnthetubaguy | tdurakov: oops, I was in mid typing | 10:17 |
| *** tojuvone has quit IRC | 10:18 | |
| *** tojuvone has joined #openstack-nova | 10:18 | |
| markus_z | johnthetubaguy: I can add something to the FAQ section to argue about. | 10:19 |
| johnthetubaguy | markus_z: we could, I am thinking about tidying up those bits this afternoon, but getting something up there would be great | 10:20 |
| johnthetubaguy | tdurakov: oh dear... not is_shared_block_storage seems better there? | 10:21 |
| markus_z | johnthetubaguy: Then I just wait for your clean up, maybe after that it's easier to add. | 10:21 |
| *** e0ne has joined #openstack-nova | 10:21 | |
| johnthetubaguy | tdurakov: but I get your point, thats a massive upgrade issue | 10:21 |
| tdurakov | johnthetubaguy, there is a mess in flags here, working on your idea for cleanup, for me it's worth to try to clean this up | 10:22 |
| johnthetubaguy | tdurakov: I think we will only be able to call that all the time, once the min_service version is new enough, or RPC pin is high enough | 10:22 |
| tdurakov | johnthetubaguy, yep, keep in mind upgrades while implementing | 10:22 |
| johnthetubaguy | tdurakov: yeah, I think its worth tidying this mess up | 10:22 |
| johnthetubaguy | tdurakov: deleting instances (or not deleting them) is about the worst thing we can mess up | 10:23 |
| tdurakov | ++ | 10:23 |
| *** sfinucan has joined #openstack-nova | 10:28 | |
| *** jaosorior has quit IRC | 10:29 | |
| *** jaosorior has joined #openstack-nova | 10:29 | |
| *** deva_ has joined #openstack-nova | 10:30 | |
| *** brad[] has quit IRC | 10:33 | |
| *** PaulMurray has quit IRC | 10:33 | |
| *** samueldmq1 has joined #openstack-nova | 10:33 | |
| *** brad[] has joined #openstack-nova | 10:33 | |
| *** amotoki has quit IRC | 10:34 | |
| *** samueldmq1 has quit IRC | 10:38 | |
| *** deva_ has quit IRC | 10:38 | |
| ccard | Is there a way to use the python novaclient to do a request to the MetadataRequestHandler? | 10:39 |
| *** salv-orlando has joined #openstack-nova | 10:40 | |
| *** cdent has joined #openstack-nova | 10:41 | |
| *** salv-orl_ has quit IRC | 10:43 | |
| *** zenoway has quit IRC | 10:44 | |
| *** rfolco has joined #openstack-nova | 10:44 | |
| *** dims has joined #openstack-nova | 10:45 | |
| *** shyama has quit IRC | 10:47 | |
| *** omnipresent has joined #openstack-nova | 10:49 | |
| bauzas | johnthetubaguy: jaypipes: I just saw that http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/resource-objects.html was still there | 10:49 |
| *** dims has quit IRC | 10:49 | |
| bauzas | johnthetubaguy: jaypipes: we should probably cut it off | 10:49 |
| johnthetubaguy | bauzas: is that invalid now? | 10:49 |
| bauzas | or put it in the backlog specs repo | 10:49 |
| *** amotoki has joined #openstack-nova | 10:49 | |
| johnthetubaguy | or just not happening this release? | 10:49 |
| *** Daisy has quit IRC | 10:49 | |
| *** dims has joined #openstack-nova | 10:50 | |
| bauzas | johnthetubaguy: that was the early spec for explaining the issue, but jaypipes exploded it in 6 specs | 10:50 |
| johnthetubaguy | bauzas: so there are lots of patches for that: https://blueprints.launchpad.net/nova/+spec/resource-objects | 10:50 |
| *** Daisy has joined #openstack-nova | 10:50 | |
| johnthetubaguy | lets just be sure to call that finished soon | 10:50 |
| *** fawadkhaliq has joined #openstack-nova | 10:50 | |
| bauzas | johnthetubaguy: I know, jaypipes and cdent should clean them up or abandon them | 10:50 |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: libvirt: prevent serial console port exhaustion https://review.openstack.org/275800 | 10:51 |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: libvirt: Acquire TCP ports for console during live migration https://review.openstack.org/275801 | 10:51 |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: libvirt: Add serial ports to the migration data object https://review.openstack.org/215102 | 10:51 |
| bauzas | cdent: btw. I'm just about fencing https://review.openstack.org/#/c/200870/ in favor or resource-providers, it would be easier for me if there was some devref epic or backlog spec I could provide | 10:51 |
| bauzas | johnthetubaguy: ^ | 10:51 |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: libvirt: move "_get_serial_ports_from_guest" to Guest class https://review.openstack.org/236051 | 10:51 |
| cdent | bauzas: don't look at me, I've abandoned my aborted effort on a db-backed resource class and everything else that is "owned" by me is active. | 10:52 |
| bauzas | hah | 10:52 |
| *** deepakcs has quit IRC | 10:52 | |
| *** efoley has quit IRC | 10:52 | |
| bauzas | who would be foolish enough for providing an epic story about the overall effort ? | 10:52 |
| bauzas | at least describing the problem and the 10k-feet effort | 10:53 |
| jaypipes | johnthetubaguy, bauzas: yeah, I should clean up the resource objects stuff. | 10:53 |
| jaypipes | lemme do that now. | 10:53 |
| bauzas | because how could I explain to https://review.openstack.org/#/c/200870/ that his patch is not really cool given the effort we work on ? | 10:53 |
| johnthetubaguy | jaypipes: I marked the BP as superseded, not sure whats left | 10:53 |
| jaypipes | johnthetubaguy: I'll check and mark stuff abandoned if needed. | 10:54 |
| johnthetubaguy | jaypipes: bauzas: thanks | 10:54 |
| bauzas | FWIW, I just pointed http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/resource-providers.html | 10:54 |
| *** thorst has joined #openstack-nova | 10:54 | |
| bauzas | jaypipes: drop me a ping once you're done with cleaning that up, I'll +2 quickly | 10:55 |
| *** amotoki has quit IRC | 10:55 | |
| openstackgerrit | John Garbutt proposed openstack/nova: Get a ReqSpec in evacuate API and pass it to scheduler https://review.openstack.org/273104 | 10:56 |
| jaypipes | bauzas: I abandoned all of my open WIP patches involving resource objects and providers. | 10:56 |
| *** erlon has joined #openstack-nova | 10:56 | |
| jaypipes | bauzas: since cdent now has the master patch series.. | 10:56 |
| bauzas | _o/ | 10:57 |
| bauzas | \o/ | 10:57 |
| bauzas | roger. | 10:57 |
| jaypipes | ten four. | 10:57 |
| * cdent uses the power of greyskull | 10:57 | |
| jaypipes | bauzas, johnthetubaguy: your review on the gen-resource-pools spec would be great: https://review.openstack.org/#/c/253187/ | 10:57 |
| johnthetubaguy | bauzas: I guess you have all the patches for this up now? https://blueprints.launchpad.net/nova/+spec/check-destination-on-migrations | 10:57 |
| * jaypipes goes off to clean up the compute-node-inventory spec | 10:57 | |
| bauzas | jaypipes: in my queue for today :) | 10:57 |
| jaypipes | sweet, thank you sir | 10:57 |
| bauzas | at least once I'm ironing out my live-migrate patch :) | 10:57 |
| danpb | jaypipes: hey, need your input on this os-vif patch https://review.openstack.org/#/c/275281/ | 10:57 |
| johnthetubaguy | jaypipes: do we want to give priority to the node-inventory one though? | 10:57 |
| bauzas | johnthetubaguy: unfortunately no | 10:58 |
| jaypipes | danpb: heh, I have that half-reviewed in gertty. lemme finish that up before tackling the compute-node-inventory. | 10:58 |
| johnthetubaguy | bauzas: but we need that data migration this cycle really | 10:58 |
| *** PaulMurray has joined #openstack-nova | 10:58 | |
| jaypipes | johnthetubaguy: dansmith does, as do I. | 10:58 |
| bauzas | johnthetubaguy: that's my target indeed | 10:58 |
| danpb | jaypipes: the follow up linux-bridge & ovs patches show real-world exmaple of its usage | 10:58 |
| jaypipes | danpb: yup. gimme 30 minutes I should have the series reviewed. | 10:59 |
| bauzas | johnthetubaguy: I'm just writing a patch for live-migrate, I'll do that too for unshelve and cold-migrate | 10:59 |
| johnthetubaguy | bauzas: ah, got it | 10:59 |
| johnthetubaguy | bauzas: I think you typoed the blueprint link, I just fixed my patch finding script | 10:59 |
| danpb | jaypipes: in particular notice the (slightly disgusting) hack i use with 'deprecated_group' to make it upgrade friendly for existing nova.conf :-) | 11:00 |
| bauzas | johnthetubaguy: probably lemme check | 11:00 |
| bauzas | mmm | 11:00 |
| bauzas | https://review.openstack.org/#/c/273104/ says Partially-Implements: blueprint check-destination-on-migrations | 11:00 |
| jaypipes | danpb: yes, I noticed that, but understand the reason for it :) | 11:00 |
| bauzas | johnthetubaguy: oh, you fixed a typo | 11:00 |
| *** haomaiwa_ has quit IRC | 11:01 | |
| bauzas | meh | 11:01 |
| johnthetubaguy | bauzas: I just found this bug, that I think the scheduler team might need to dig into: https://review.openstack.org/#/c/245619/1 | 11:01 |
| johnthetubaguy | Move disk allocation ratio to ResourceTracker | 11:01 |
| *** haomaiwang has joined #openstack-nova | 11:01 | |
| bauzas | I was probably jammed by http://www.independent.co.uk/news/world/europe/french-grammar-uproar-over-decision-to-implement-26-year-old-reform-of-spelling-rules-a6854441.html when I made this typo | 11:01 |
| openstackgerrit | Abhilash Goyal proposed openstack/nova: session.py file was initially using eval() which is not recommended for security purposes, hence I used other safe equivalent function. Close-bug: 1414529 Change-Id: Ib6145408360fa57cccc8d77c590a203e5088b193 https://review.openstack.org/275513 | 11:02 |
| openstack | bug 1414529 in OpenStack Compute (nova) "eval being used in session.py" [Low,In progress] https://launchpad.net/bugs/1414529 - Assigned to Abhilash Goyal (abhilash-goyal) | 11:02 |
| bauzas | johnthetubaguy: oh great ! | 11:02 |
| bauzas | johnthetubaguy: thanks for finding that gem | 11:02 |
| *** thorst has quit IRC | 11:02 | |
| *** efoley has joined #openstack-nova | 11:02 | |
| johnthetubaguy | bauzas: I guess we missed one, oops | 11:02 |
| bauzas | yeah, I was fooled | 11:02 |
| bauzas | that would even be a low-hanging-fruit | 11:03 |
| bauzas | johnthetubaguy: meh this is a bug https://bugs.launchpad.net/nova/+bug/1513335, categorizing it as wishlist | 11:04 |
| openstack | Launchpad bug 1513335 in OpenStack Compute (nova) "disk allocation ratio should move to resource tracker" [Undecided,In progress] - Assigned to Wen Zhi Yu (yuywz) | 11:04 |
| bauzas | johnthetubaguy: but I added the low-hanging-fruit tag | 11:04 |
| johnthetubaguy | bauzas: it will slow down the scheduler plans though, if we don't fix that right? | 11:05 |
| bauzas | mmm | 11:05 |
| bauzas | not sure | 11:05 |
| johnthetubaguy | jaypipes: ^ | 11:05 |
| johnthetubaguy | bauzas: we need to keep the config working the same as it was last cycle | 11:05 |
| bauzas | if we introduce it in Mitaka, it means we can drop the compat in O right? | 11:06 |
| bauzas | meh no | 11:06 |
| bauzas | we can even drop that in early Newton | 11:06 |
| bauzas | because mitaka computes will report that | 11:06 |
| bauzas | so | 11:07 |
| jaypipes | danpb: top patch reviewed. | 11:07 |
| *** mvk has quit IRC | 11:07 | |
| johnthetubaguy | bauzas: yes, newton | 11:07 |
| bauzas | johnthetubaguy: that's a big thing to do, quite now | 11:07 |
| *** mvk has joined #openstack-nova | 11:08 | |
| *** chhavi has quit IRC | 11:08 | |
| *** chhavi has joined #openstack-nova | 11:08 | |
| ndipanov | johnthetubaguy, so all non-prio features not merged are dead at this point | 11:08 |
| ndipanov | no more exceptions right? | 11:08 |
| johnthetubaguy | ndipanov: yes | 11:09 |
| johnthetubaguy | ndipanov: just trying to find them all | 11:09 |
| *** efoley_ has joined #openstack-nova | 11:09 | |
| johnthetubaguy | ndipanov: actually there are a few special cases where we have screwed them over, or they are on going things we want to wrap up, but trying to shut them down soon ish too | 11:09 |
| *** efoley has quit IRC | 11:11 | |
| ndipanov | johnthetubaguy, this one falls in the latter category at least https://review.openstack.org/262583 but as discussed not really high-prio (though would enable further convos in neutron) | 11:11 |
| cdent | bauzas, jaypipes: Just to make sure I'm on the same page as you guys: The outcome of last night's discussion is that we are targetting the compute-node inventory migration (and its prereqs) as the must-have for mitake and the other resource-provider stuff is bonus if we make some headway? | 11:11 |
| jaypipes | cdent: not quite. we are targeting resource-classes, resource-providers, generic-rsource-pools as must have for Mitaka. compute-node-inventory as a Really Strong Priority to get done in Mitaka, and resource-providers-allocations and resource-providers-scheduler for Newton. | 11:13 |
| bauzas | cdent: tbc, I feel the already two approved specs are doable, while bonus points could be https://review.openstack.org/#/c/253187/ | 11:13 |
| danpb | jaypipes: cool thanks | 11:13 |
| *** whenry has joined #openstack-nova | 11:13 | |
| cdent | bauzas, jaypipes: this is why I asked: you two don't agree | 11:13 |
| bauzas | well | 11:14 |
| cdent | and neither of you seem to agree with dansmith | 11:14 |
| jaypipes | cdent: yes, I am aware of that :) | 11:14 |
| bauzas | lemme be clear, generic-resource-pools is not yet approved, right? | 11:14 |
| bauzas | so, yes, it's doable but it not only depends on code, that's it | 11:14 |
| johnthetubaguy | ndipanov: yeah, that one is blocked now I am afraid, I did intend to get to that one, but it didn't happen :( | 11:15 |
| * cdent will write some tests for the existing code | 11:16 | |
| johnthetubaguy | jaypipes: bauzas: why do we want generic-resource-pools over compute-node-inventory? | 11:16 |
| johnthetubaguy | it does feel like getting the nasty migrations all out the way in one release, would be a big, big, win | 11:17 |
| jaypipes | johnthetubaguy: because generic-resource-pools solves the long-standing issue of incorrect resource reporting for shared disk. plus it paves the way for compute-node-inventory. | 11:17 |
| ndipanov | johnthetubaguy, ah well... | 11:17 |
| johnthetubaguy | jaypipes: I think it was more just the adding the uuid, that sucked the most? | 11:18 |
| johnthetubaguy | jaypipes: but I need to re-read that | 11:18 |
| jaypipes | johnthetubaguy: you mean compute-node-inventory? the adding of the UUID to the compute node is only a small part of the overall migration needed in that one. | 11:19 |
| *** fawadkhaliq has quit IRC | 11:19 | |
| *** fawadkhaliq has joined #openstack-nova | 11:19 | |
| johnthetubaguy | jaypipes: but most of the other bits feel like the same update a new compute node would make, so its much easier, once we have the uuid, no need to "force" any of the other bits before we add unique constrains, etc | 11:19 |
| *** fawadkhaliq has quit IRC | 11:19 | |
| openstackgerrit | Rossella Sblendido proposed openstack/nova: Wait for neutron events when using xen driver https://review.openstack.org/268258 | 11:19 |
| *** fawadkhaliq has joined #openstack-nova | 11:20 | |
| johnthetubaguy | cdent: that might be the other way of looking at it, maybe we just add the uuid into the compute node this release, rather than the whole spec. although I do feel like I am missing another piece | 11:20 |
| jaypipes | johnthetubaguy: sorry, I'm not following you... | 11:20 |
| *** ihrachys has quit IRC | 11:20 | |
| johnthetubaguy | jaypipes: so the compute node uuid, to do it online, its a total faf that takes two cycles | 11:21 |
| *** paul-carlton2 has quit IRC | 11:21 | |
| jaypipes | johnthetubaguy: why? | 11:21 |
| johnthetubaguy | jaypipes: you have to add the column, then start filling it in "online", at somepoint we then the next release we add a force migration to say it must be complete, and we make it non-nullable and unique | 11:22 |
| jaypipes | johnthetubaguy: yes, of course, but I don't see how that is affected by doing the inventory migration in the same online migration? | 11:23 |
| johnthetubaguy | jaypipes: I was really just meaning, I think its the uuid thing thats the hard bit we should do this release | 11:23 |
| johnthetubaguy | jaypipes: the other bits are not migrations, they are just updating to a new location (and possible creating that location), I think? | 11:24 |
| johnthetubaguy | the reader just needs to know about looking at the old location, till the compute is upgraded | 11:24 |
| jaypipes | johnthetubaguy: yes, that is true. | 11:25 |
| jaypipes | johnthetubaguy: I would still really like to shoot for the inventory migrations in Mitaka, though... | 11:25 |
| johnthetubaguy | jaypipes: I am OK with that, I just think the uuid bit is on the must have list | 11:26 |
| johnthetubaguy | above the generic-resource-pools, I think, but thats just splitting hairs | 11:26 |
| johnthetubaguy | cdent: did that make any sense? | 11:27 |
| cdent | I parsed you both in what seems to be a sensible fashion... | 11:28 |
| cdent | But I'm not clear on the _actual_ decision. | 11:28 |
| openstackgerrit | garyk proposed openstack/nova: Virt: add context to attach and detach interface https://review.openstack.org/209362 | 11:28 |
| cdent | (if there was one) | 11:28 |
| openstackgerrit | garyk proposed openstack/nova: Compute: remove object_compat from get_instance_diagnostics https://review.openstack.org/218623 | 11:28 |
| cdent | code already exists (because dansmith wrote it) for the uuid migration | 11:28 |
| johnthetubaguy | cdent: ah, OK, I think my take is, lets try get that uuid migration in ASAP | 11:29 |
| johnthetubaguy | cdent: and do our best for the rest of it, largely | 11:30 |
| johnthetubaguy | it gets is to a simpler place quicker, thats all | 11:30 |
| jaypipes | johnthetubaguy, bauzas: can I approved the ResourceClassEnum patch? is there a reason that isn't +Wallaby'd yet? | 11:31 |
| cdent | dan added to the resource providers migration: https://review.openstack.org/#/c/275432/3/nova/db/sqlalchemy/migrate_repo/versions/314_add_resource_provider_tables.py | 11:31 |
| cdent | jaypipes: its okay to go, it was blocked briefly because dansmith wanted real code | 11:31 |
| cdent | (that used it) | 11:32 |
| cdent | and now there is, so he said he was happy | 11:32 |
| jaypipes | cdent: k, off to the tarlac pit it goes then. | 11:32 |
| *** bkumar has quit IRC | 11:32 | |
| jaypipes | cdent: BTW, I left a suggestion in there for you to move that code to oslo.versionedobjects in a future patch. | 11:32 |
| cdent | roger that | 11:33 |
| bauzas | jaypipes: go for it | 11:33 |
| jaypipes | cdent: should have been Sarlacc pit. | 11:33 |
| cdent | that's a good idea | 11:34 |
| cdent | (the base change) | 11:34 |
| *** efoley__ has joined #openstack-nova | 11:34 | |
| cdent | jaypipes: except as I've said before, property getter can't take an argument | 11:35 |
| cdent | (or am I misunderstanding your sugestion) | 11:35 |
| *** baoli has joined #openstack-nova | 11:36 | |
| *** efoley_ has quit IRC | 11:36 | |
| *** baoli has quit IRC | 11:37 | |
| *** dims_ has joined #openstack-nova | 11:40 | |
| *** paul-carlton2 has joined #openstack-nova | 11:40 | |
| *** dims has quit IRC | 11:40 | |
| openstackgerrit | jichenjc proposed openstack/nova: Forbidden non-authorized user to do resize actions https://review.openstack.org/277797 | 11:41 |
| jaypipes | cdent: oh, duh, yeah. | 11:41 |
| cdent | but it's still a good idea to get it in the base | 11:41 |
| jaypipes | ya. | 11:41 |
| cdent | I've stuck it on my rather meandering to do list ;) | 11:41 |
| *** omnipresent has quit IRC | 11:42 | |
| *** zenoway has joined #openstack-nova | 11:43 | |
| *** boris-42 has quit IRC | 11:43 | |
| jaypipes | cdent: heh, welcome to the club :) | 11:44 |
| cdent | oh I've been in this club for a very very long time, it's just new things on the list | 11:45 |
| cdent | over in a nother channel I'm discussing with a friend something we've been planning to in various way for about 16 years, still haven't started. is just a simple thing, mostly. | 11:45 |
| cdent | but I seem to be operating with a stack not a queue | 11:46 |
| jaypipes | :) | 11:46 |
| *** zenoway has quit IRC | 11:47 | |
| pkoniszewski | danpb: hi, are you around? | 11:48 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: WIP: Provide ReqSpec to live-migrate conductor task https://review.openstack.org/277800 | 11:49 |
| *** zenoway has joined #openstack-nova | 11:49 | |
| *** sudipto has quit IRC | 11:50 | |
| *** xek_ is now known as xek | 11:51 | |
| bauzas | jaypipes: johnthetubaguy: alaski: would you mind review https://review.openstack.org/#/c/273104/ ? I'll add it to the prio etherpad for the sched stuff | 11:52 |
| bauzas | oh, some gentleman added that to the etherpad | 11:52 |
| *** gszasz has joined #openstack-nova | 11:53 | |
| *** shyama has joined #openstack-nova | 11:53 | |
| jaypipes | bauzas: yup, will do as soon as I wrap up cdent's patch series. | 11:55 |
| cdent | cdent hasn't had a chance to look at that yet, since dan's visit | 11:56 |
| bauzas | I just sorted out the etherpad of doom | 11:56 |
| bauzas | johnthetubaguy: jaypipes: so we still have 2 "epics/BP" to merge before M-3, I raised them in the etherpad for the sched section | 11:57 |
| bauzas | I also cleaned what was merged before | 11:57 |
| johnthetubaguy | sweet, thanks | 11:57 |
| *** gmann_ has joined #openstack-nova | 11:57 | |
| jaypipes | bauzas: cheers | 11:57 |
| cdent | jaypipes: are you one of the reasons we don't just default to utf8 everywhere? ;) | 11:58 |
| bauzas | jaypipes: cdent: given it's hard to follow all the resouce-providers BPs, would you mind making sure the etherpad is good ? | 11:58 |
| jaypipes | cdent: we do default to utf8 everywhere, unfortunately. | 11:58 |
| *** gmann_ has quit IRC | 11:59 | |
| cdent | there's that "unfortunately" word I was looking for | 11:59 |
| jaypipes | bauzas: yes, I will. I am also writing a summary email to the ML about it, ala alaski's recent thread on cellsv2 progress. | 11:59 |
| *** gmann_ has joined #openstack-nova | 11:59 | |
| cdent | (summary emails)++ | 11:59 |
| johnthetubaguy | +1 | 11:59 |
| *** thorst has joined #openstack-nova | 12:00 | |
| *** gmann_ has quit IRC | 12:00 | |
| openstackgerrit | Andrea Rosa proposed openstack/nova: Allow setting a specific value for migration state https://review.openstack.org/277804 | 12:00 |
| *** gmann_ has joined #openstack-nova | 12:00 | |
| *** haomaiwang has quit IRC | 12:01 | |
| *** haomaiwang has joined #openstack-nova | 12:01 | |
| bauzas | jaypipes: <3 | 12:02 |
| gmann_ | johnthetubaguy: sdague do we have API meeting today ? | 12:03 |
| jichen | gmann_: same question, remember sdague said he will do it | 12:04 |
| gmann_ | jichen: i did not check mail if it was canceled, just reminding if we forgot :) | 12:05 |
| johnthetubaguy | gmann_: I think we said its Chinese new year, so it was a maybe | 12:05 |
| jichen | gmann_ : me too :), let's check mail archieve | 12:05 |
| gmann_ | johnthetubaguy: ok. | 12:05 |
| gmann_ | johnthetubaguy: jichen just one thing- on Bug 1539351 | 12:06 |
| openstack | bug 1539351 in OpenStack Compute (nova) "Authorization by user_id does not work in V2.1 API" [Undecided,In progress] https://launchpad.net/bugs/1539351 - Assigned to jichenjc (jichenjc) | 12:06 |
| johnthetubaguy | I would kick it off, but I think without sdague around, not sure we have enough people to talk | 12:06 |
| gmann_ | https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bug/1539351 | 12:06 |
| gmann_ | johnthetubaguy: yea, | 12:07 |
| sdague | oh, crap | 12:07 |
| gmann_ | johnthetubaguy: jichen do we have same issue in all other places too? | 12:07 |
| sdague | yeh, my bad | 12:07 |
| *** thorst has quit IRC | 12:07 | |
| jichen | gmann_ : yes, I think so | 12:07 |
| jichen | gmann_: I tried in my devstack and seems other actions have similar issue | 12:07 |
| *** garyk has joined #openstack-nova | 12:07 | |
| sdague | 7am catches up on me really quickly | 12:08 |
| jichen | so I am working on them now | 12:08 |
| garyk | johnthetubaguy: https://review.openstack.org/#/c/252870/ - these were the patches merged in Mitaka. | 12:08 |
| garyk | no idea what you want and expect here | 12:08 |
| gmann_ | jichen: ok, Thanks | 12:08 |
| johnthetubaguy | garyk: it depends on two patches that are blocked by feature freeze | 12:08 |
| johnthetubaguy | garyk: need to rebase that so its independent of those, so we can merge it, and make sure it doesn't talk about features that didn't merge in mitaka | 12:09 |
| garyk | johnthetubaguy: those two patches are relevant for the resize, which irrespective if they land or not the release note is relevant | 12:09 |
| garyk | but a few months ago you asked that the release note be reabsed on top of these | 12:10 |
| *** ihrachys has joined #openstack-nova | 12:10 | |
| garyk | if you wan i can rebase on master | 12:10 |
| garyk | please advise | 12:10 |
| johnthetubaguy | garyk: thats cool then, just need the rebase, so the release note doesn't depend on the two blocked patches | 12:10 |
| garyk | ok | 12:10 |
| openstackgerrit | John Garbutt proposed openstack/nova: VMware: add release notes for the limits https://review.openstack.org/252870 | 12:10 |
| *** dims_ has quit IRC | 12:11 | |
| johnthetubaguy | garyk: turns out gerrit has a new button for that, my bad, +W | 12:11 |
| *** dstepanenko has quit IRC | 12:11 | |
| garyk | johnthetubaguy: ok, thanks | 12:11 |
| *** lucasagomes is now known as lucas-hungry | 12:13 | |
| bauzas | johnthetubaguy: you still okay if I'm providing the patches for https://bugs.launchpad.net/nova/+bug/1513335 by just a wishlist bug ? :( | 12:13 |
| openstack | Launchpad bug 1513335 in OpenStack Compute (nova) "disk allocation ratio should move to resource tracker" [Wishlist,In progress] - Assigned to Wen Zhi Yu (yuywz) | 12:13 |
| bauzas | johnthetubaguy: I think I should rather create a specless BP | 12:13 |
| johnthetubaguy | bauzas: meh, its a priority affecting bug/feature-ett, I am fine with whatever gets folks to review it, and we don't forget it | 12:14 |
| bauzas | okay | 12:14 |
| bauzas | I will port those patches for the disk-related stuff today | 12:14 |
| bauzas | and will create a BP | 12:14 |
| * bauzas jumps off for eating | 12:15 | |
| *** dstepanenko has joined #openstack-nova | 12:17 | |
| *** smatzek has joined #openstack-nova | 12:18 | |
| *** dims_ has joined #openstack-nova | 12:19 | |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova-specs: Libvirt: Use the virtlogd deamon https://review.openstack.org/234291 | 12:20 |
| *** zenoway has quit IRC | 12:20 | |
| rdopiera | tdurakov: would you have a moment to talk about https://blueprints.launchpad.net/nova/+spec/remove-compute-compute-communication ? | 12:21 |
| tdurakov | rdopiera, hey, sure | 12:22 |
| jaypipes | cdent: ok, your resource providers patch series fully reviewed, including the WIP last patch. | 12:23 |
| cdent | thanks jaypipes | 12:23 |
| ndipanov | cdent, jaypipes which series is that? | 12:23 |
| * ndipanov interested | 12:24 | |
| ndipanov | johnthetubaguy, was there a devref doc on our upgrade policy or was it just my imagination :) | 12:24 |
| *** zhangjn has quit IRC | 12:24 | |
| johnthetubaguy | ndipanov: it needs updating, but we have this one: http://docs.openstack.org/developer/nova/upgrade.html | 12:24 |
| ndipanov | johnthetubaguy, perfect! thanks! | 12:25 |
| johnthetubaguy | ndipanov: in terms of code review, we do have this guide: http://docs.openstack.org/developer/nova/code-review.html#upgrade-related-concerns | 12:25 |
| johnthetubaguy | I think that one is more up to date, right now | 12:26 |
| *** efoley__ has quit IRC | 12:26 | |
| *** zenoway has joined #openstack-nova | 12:26 | |
| ndipanov | johnthetubaguy, this is the reason I was asking https://bugs.launchpad.net/nova/+bug/1543562 | 12:27 |
| openstack | Launchpad bug 1543562 in OpenStack Compute (nova) "mitaka pci_request object needs a migration script for an online data migration" [Undecided,New] | 12:27 |
| ndipanov | johnthetubaguy, we were discussing this I think | 12:27 |
| *** zhangjn has joined #openstack-nova | 12:27 | |
| johnthetubaguy | ndipanov: yeah, I think at the midcycle(?) we were saying to create a new nova-manage command to always do that kind of thing (in batches). cdent will need to share that for the compute node uuid stuff | 12:28 |
| cdent | ndipanov: the series starts here: https://review.openstack.org/#/c/273100/ | 12:28 |
| *** mtanino has joined #openstack-nova | 12:28 | |
| ndipanov | cdent, got it | 12:29 |
| ndipanov | johnthetubaguy, well yeah ideally we'd want to have some kind of a framework but we're not there now | 12:29 |
| johnthetubaguy | ndipanov: yep | 12:29 |
| johnthetubaguy | ndipanov: the flavor one established the pattern, but it was flavor specific | 12:29 |
| ndipanov | johnthetubaguy, let me think about it for a bit, maybe I can come up with something | 12:29 |
| bauzas | jaypipes: just a comment on https://review.openstack.org/#/c/275432/3/nova/db/sqlalchemy/migrate_repo/versions/314_add_resource_provider_tables.py | 12:30 |
| *** gmann_ has quit IRC | 12:31 | |
| *** zenoway has quit IRC | 12:31 | |
| *** gmann_ has joined #openstack-nova | 12:31 | |
| rdopiera | tdurakov: so as far as I understand, you plan to keep the status of migration on the compute itself, and have the conductor call some api to get it | 12:31 |
| cdent | man there is a _ton_ of latent assumptions running around this resource provider stuff... | 12:31 |
| bauzas | ndipanov: johnthetubaguy: cdent: for the compute uuid, why do we need a nova-manage ? the RT can provide that for us | 12:31 |
| johnthetubaguy | ndipanov: this was the last one, I think: https://github.com/openstack/nova/blob/stable/kilo/nova/cmd/manage.py#L966 | 12:31 |
| *** gmann_ has quit IRC | 12:32 | |
| bauzas | ndipanov: johnthetubaguy: cdent: that's something I did for adding the host field to the ComputeNode object | 12:32 |
| *** gmann_ has joined #openstack-nova | 12:32 | |
| johnthetubaguy | bauzas: because we need to do it for *all* db entries, even if the computenode is dead | 12:32 |
| tdurakov | rdopiera, there is a plan to get rid of compute-compute communication, this part will be required to save migration state in case of compute-failure | 12:32 |
| jaypipes | bauzas: answereed :) | 12:32 |
| johnthetubaguy | bauzas: so we can later change the DB schema to be non-null, and unique uuids, etc | 12:33 |
| bauzas | johnthetubaguy: mmm, I had the same concern when I added the host field to compute_nodes | 12:33 |
| rdopiera | tdurakov: do you have any idea how that would be done? | 12:33 |
| bauzas | johnthetubaguy: but since an host can come back, then its RT will provide its field | 12:33 |
| bauzas | but meh, I have to go for lunch | 12:34 |
| johnthetubaguy | bauzas: right, we would only need that if we made the schema non-null | 12:34 |
| johnthetubaguy | bauzas: which we might want to do alongside the uuid stuff, actually | 12:34 |
| tdurakov | rdopiera, I've started to work on this bp. some kind of storage will fit this requirement, something small and easy to maintain | 12:34 |
| * cdent reviews jay's reviews of dan's changes to "his" code | 12:34 | |
| jaypipes | bauzas: evac reqspec patch reviewed. nice work. | 12:34 |
| tdurakov | cdent, lol, sounds complicated:) | 12:35 |
| rdopiera | tdurakov: sqlite? :) | 12:35 |
| tdurakov | rdopiera, not ready to answer, need to verify it first:) You are welcome to propose ideas for storage | 12:35 |
| pkoniszewski | rdopiera: we haven't agreed on a solution yet, so pretty hard to say which storage will be used | 12:36 |
| jaypipes | tdurakov: definitely Oracle RAC. | 12:36 |
| jaypipes | tdurakov: or Clipper 4. | 12:36 |
| tdurakov | jaypipes, yes, heavy metal thing is absolutely required:-D | 12:36 |
| rdopiera | tdurakov, pkoniszewski: is there anything more on this that I could read? The original bugs that are being solved by this, for instance? | 12:37 |
| tdurakov | rdopiera, well, I'm not sure there are a lot of bugs, I'll ping you once spec will be ready for review | 12:37 |
| rdopiera | tdurakov: that | 12:38 |
| tdurakov | ? | 12:38 |
| rdopiera | tdurakov: that's great | 12:38 |
| pkoniszewski | rdopiera: i can see at least 3 bugs that might be solved by this refactoring | 12:38 |
| rdopiera | tdurakov: by the way, I wonder if it could be also used to keep track of other tasks | 12:38 |
| rdopiera | tdurakov: like deleting | 12:38 |
| *** openstackgerrit_ has joined #openstack-nova | 12:39 | |
| *** tdurakov_ has joined #openstack-nova | 12:39 | |
| tdurakov | rdopiera, ndipanov said that you have at least one, that will be fixed by this change | 12:39 |
| *** tdurakov_ has quit IRC | 12:39 | |
| tdurakov | rdopiera, I'd rather say yes | 12:40 |
| rdopiera | tdurakov: yeah, if rabbit goes down during live-migrate | 12:40 |
| tdurakov | rdopiera, could you please file it so it will be posible to track it too | 12:40 |
| rdopiera | tdurakov: I have, and I think pkoniszewski even commented on it | 12:41 |
| tdurakov | oh, cool! | 12:41 |
| rdopiera | https://bugs.launchpad.net/nova/+bug/1536589 | 12:41 |
| openstack | Launchpad bug 1536589 in OpenStack Compute (nova) "Inconsistent state when connection to conductor is lost during live migration" [Undecided,New] - Assigned to Radomir Dopieralski (thesheep) | 12:41 |
| pkoniszewski | tdurakov: its already reported, yes | 12:41 |
| rdopiera | I have another one: https://bugs.launchpad.net/nova/+bug/1543511 | 12:41 |
| openstack | Launchpad bug 1543511 in OpenStack Compute (nova) "Instances stuck in DELETING state if delete fails" [Undecided,New] | 12:41 |
| rdopiera | could possibly work for this too | 12:42 |
| *** fawadkhaliq has quit IRC | 12:43 | |
| *** takashin has left #openstack-nova | 12:44 | |
| *** marcusvrn_ has joined #openstack-nova | 12:44 | |
| *** zenoway has joined #openstack-nova | 12:44 | |
| tdurakov | rdopiera, well, some kind of general storage | 12:45 |
| pkoniszewski | PaulMurray: are you around? | 12:45 |
| *** doug-fish has joined #openstack-nova | 12:45 | |
| *** thorst has joined #openstack-nova | 12:46 | |
| *** efoley__ has joined #openstack-nova | 12:46 | |
| *** kylek3h has joined #openstack-nova | 12:47 | |
| *** thorst_ has joined #openstack-nova | 12:47 | |
| *** doug-fish has quit IRC | 12:47 | |
| *** doug-fish has joined #openstack-nova | 12:48 | |
| *** zenoway has quit IRC | 12:49 | |
| *** thorst has quit IRC | 12:51 | |
| *** zenoway has joined #openstack-nova | 12:51 | |
| *** ducttape_ has joined #openstack-nova | 12:53 | |
| *** Marga_ has joined #openstack-nova | 12:55 | |
| cdent | jaypipes: you and dansmith seem to have some difference of opinon on versionedobjects tropes | 12:58 |
| *** haomaiwang has quit IRC | 13:01 | |
| jaypipes | cdent: to be expected. | 13:01 |
| *** haomaiwang has joined #openstack-nova | 13:01 | |
| PaulMurray | pkoniszewski, hi | 13:03 |
| *** cdent has quit IRC | 13:04 | |
| *** cdent has joined #openstack-nova | 13:05 | |
| *** ildikov has quit IRC | 13:06 | |
| *** garyk1 has joined #openstack-nova | 13:06 | |
| *** garyk has quit IRC | 13:07 | |
| *** lucas-hungry is now known as lucasagomes | 13:07 | |
| *** dane-fichter has joined #openstack-nova | 13:08 | |
| *** ducttape_ has quit IRC | 13:10 | |
| *** mfedosin has joined #openstack-nova | 13:10 | |
| openstackgerrit | Ludovic Beliveau proposed openstack/nova: Failed migration shoudn't be reported as in progress https://review.openstack.org/277451 | 13:12 |
| openstackgerrit | Bob Ball proposed openstack/nova: XenAPI:Resolve Nova/Neutron race condition https://review.openstack.org/241127 | 13:14 |
| openstackgerrit | Bob Ball proposed openstack/nova: XenAPI: Fix VIF plug and unplug problem https://review.openstack.org/254022 | 13:14 |
| *** dave-mccowan has joined #openstack-nova | 13:14 | |
| PaulMurray | pkoniszewski, i'm around if you need me | 13:14 |
| *** zenoway has quit IRC | 13:16 | |
| *** links has quit IRC | 13:17 | |
| pkholkin_ | johnthetubaguy, sdague, alaski: thank you guys for help with enginefacade bp, you were our main reviewers, happy that it was done) | 13:18 |
| sdague | pkholkin_: thanks for all the work on that | 13:18 |
| pkholkin_ | mdbooth: ^^ | 13:19 |
| *** garyk1 has quit IRC | 13:20 | |
| mdbooth | pkholkin_: Congrats :) That's a lot of work. | 13:20 |
| *** dedery has joined #openstack-nova | 13:21 | |
| pkholkin_ | yes, hope it will work well :) | 13:21 |
| *** fawadkhaliq has joined #openstack-nova | 13:22 | |
| johnthetubaguy | pkholkin_: that was a big push, thank you for sticking with it! | 13:22 |
| *** dedery_ has joined #openstack-nova | 13:22 | |
| pkholkin_ | no problem) | 13:23 |
| *** deepakcs has joined #openstack-nova | 13:24 | |
| *** jlanoux has quit IRC | 13:24 | |
| *** jlanoux has joined #openstack-nova | 13:25 | |
| alaski | pkholkin_: yeah, thanks for pushing that forward | 13:25 |
| *** dedery has quit IRC | 13:25 | |
| *** dedery_ is now known as dedery | 13:25 | |
| ankit_ag | sdauge: Hi could you please check this patch https://review.openstack.org/#/c/185549/, you already given +2 but I have fixed minor comment from Kevin and submitted a new patch | 13:31 |
| ankit_ag | sdauge: sorry for trouble and thank you for support | 13:31 |
| *** zenoway has joined #openstack-nova | 13:35 | |
| *** ivase has joined #openstack-nova | 13:35 | |
| *** ankit_ag has quit IRC | 13:37 | |
| *** nelsnels_ has joined #openstack-nova | 13:37 | |
| *** nelsnelson has quit IRC | 13:37 | |
| sdague | done | 13:38 |
| *** speller has joined #openstack-nova | 13:40 | |
| *** ihrachys has quit IRC | 13:40 | |
| *** mriedem has joined #openstack-nova | 13:41 | |
| johnthetubaguy | alaski: I thinking some of those cells configs are used in nova-compute, not just nova-cells. I think thats correct? https://review.openstack.org/#/c/274698/5 | 13:41 |
| *** daemontool has quit IRC | 13:42 | |
| *** jlanoux has quit IRC | 13:43 | |
| *** jlanoux has joined #openstack-nova | 13:43 | |
| *** garyk has joined #openstack-nova | 13:43 | |
| *** eharney has joined #openstack-nova | 13:45 | |
| alaski | hmm, let me check | 13:45 |
| *** haomaiwang has quit IRC | 13:46 | |
| alaski | johnthetubaguy: responded. one is used elsewhere the other is not | 13:47 |
| *** haomaiwa_ has joined #openstack-nova | 13:47 | |
| *** zenoway has quit IRC | 13:47 | |
| *** zenoway has joined #openstack-nova | 13:48 | |
| openstackgerrit | John Garbutt proposed openstack/nova: Improve the help text for cells options (1) https://review.openstack.org/274698 | 13:49 |
| johnthetubaguy | alaski: thanks, yeah, that makes sense | 13:50 |
| *** tdurakov has quit IRC | 13:51 | |
| *** w_verdugo has joined #openstack-nova | 13:51 | |
| *** tdurakov has joined #openstack-nova | 13:51 | |
| bauzas | do we know if we have some grenade problem with n-api not stopping correctly? | 13:51 |
| *** zenoway has quit IRC | 13:52 | |
| mriedem | bauzas: link? | 13:54 |
| sdague | bauzas: probably oslo.service issue again | 13:55 |
| bauzas | mriedem: http://logs.openstack.org/00/277800/1/check/gate-grenade-dsvm/72f8464/logs/grenade.sh.txt.gz#_2016-02-09_12_41_22_774 | 13:55 |
| sdague | oslo.service continues to break shutting down | 13:55 |
| bauzas | I'm just looking at my change, so I'll logstash | 13:55 |
| sdague | bauzas: I think there is a tracking bug | 13:55 |
| sdague | but we should figure out if there is a new oslo.service break | 13:55 |
| bauzas | http://logs.openstack.org/00/277800/1/check/gate-grenade-dsvm/72f8464/logs/old/screen-n-api.txt.gz#_2016-02-09_12_39_06_994 | 13:55 |
| sdague | because about every 6 months it stops doing shutdown correctly | 13:56 |
| tdurakov | PaulMurray, hi, is l-m meetng on schedule? | 13:56 |
| *** edmondsw has joined #openstack-nova | 13:56 | |
| PaulMurray | yes, in 5 mins | 13:56 |
| *** openstackgerrit_ has quit IRC | 13:56 | |
| sdague | ok, log might be tripping it up? | 13:56 |
| tdurakov | acked | 13:56 |
| sdague | bauzas: can you file a critical bug with oslo.service | 13:57 |
| bauzas | sdague: sure, I'm just checking logstash | 13:57 |
| sdague | bauzas: yeh | 13:57 |
| *** ljxiash has quit IRC | 13:57 | |
| mriedem | oslo.service hasn't done a release since 1/14 fwiw | 13:57 |
| bauzas | but lemme write the bug first and invalidate it if needed | 13:57 |
| PaulMurray | live migration meeting on #openstack-meeting-3 in a couple of minutes PaulMurray paul-carlton2 eliqiao shaohe_feng pkoniszewski jlanoux yuntongjin mdbooth kashyap kmartin claudiub | 13:57 |
| dane-fichter | mriedem: Any idea when the next glance-store release will be? We need it for image-signing in Glance | 13:58 |
| sdague | mriedem: it's been around for a few weeks | 13:58 |
| sdague | there was also an eventlet bump | 13:58 |
| mriedem | dane-fichter: you'd have to ask flaper87 or jokke probably | 13:58 |
| *** ljxiash has joined #openstack-nova | 13:58 | |
| mriedem | sdague: yeah something started spiking http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22AssertionError%3A%20Cannot%20switch%20to%20MAINLOOP%20from%20MAINLOOP%5C%22&from=7d | 13:58 |
| dane-fichter | mriedem: ok, thanks. Just saw that you did the last couple | 13:58 |
| * bauzas writing the bug report | 13:58 | |
| *** daemontool has joined #openstack-nova | 13:59 | |
| bauzas | mriedem: I don't see the spikes, just something occuring regularly. Am | 13:59 |
| bauzas | Am I blind ? | 13:59 |
| mriedem | well, 'AssertionError: Cannot switch to MAINLOOP from MAINLOOP' is new | 14:00 |
| mriedem | but that just might be new eventlet | 14:00 |
| *** haomaiwa_ has quit IRC | 14:01 | |
| mriedem | actually i'm not seeing a new eventlet | 14:01 |
| mriedem | it's using 0.17.4 | 14:01 |
| sdague | mriedem: could be | 14:01 |
| *** garyk has quit IRC | 14:01 | |
| *** haomaiwa_ has joined #openstack-nova | 14:01 | |
| mriedem | i refreshed that query, message:"AssertionError: Cannot switch to MAINLOOP from MAINLOOP" doesn't look new after all | 14:02 |
| *** xyang1 has joined #openstack-nova | 14:02 | |
| *** ildikov has joined #openstack-nova | 14:02 | |
| bauzas | yeah | 14:02 |
| *** Daisy has quit IRC | 14:02 | |
| sdague | it's not fatal all the time, however there is a real issue here | 14:02 |
| sdague | oslo.service really only has a few jobs, and shutting down is one of them | 14:02 |
| *** dims_ has quit IRC | 14:03 | |
| mriedem | could we just be timing out? | 14:03 |
| mriedem | n-api is waiting for 6 children to exit, http://logs.openstack.org/00/277800/1/check/gate-grenade-dsvm/72f8464/logs/old/screen-n-api.txt.gz#_2016-02-09_12_39_06_988 | 14:04 |
| mriedem | 5 do | 14:04 |
| sdague | it's supposed to have a hard stop | 14:05 |
| bauzas | sdague: shall I write another bug or use this https://bugs.launchpad.net/bugs/1538204 | 14:06 |
| openstack | Launchpad bug 1538204 in oslo.service "Failed to stop nova-api in grenade tests" [Undecided,New] | 14:06 |
| bauzas | it seems the stack is different | 14:06 |
| openstackgerrit | John Garbutt proposed openstack/nova: Config options: centralize options in conductor api https://review.openstack.org/274259 | 14:07 |
| *** deepakcs has quit IRC | 14:07 | |
| *** mtanino has quit IRC | 14:07 | |
| *** haomaiwa_ has quit IRC | 14:08 | |
| mriedem | it looks like grenade is actually waiting up to 2 minutes for the service to die | 14:08 |
| mriedem | http://logs.openstack.org/00/277800/1/check/gate-grenade-dsvm/72f8464/logs/grenade.sh.txt.gz#_2016-02-09_12_39_15_226 | 14:08 |
| mriedem | 2016-02-09 12:39:14.223 | The following services are still running: nova-api... sleeping and trying again 2016-02-09 12:39:14.223 | + sleep 1 2016-02-09 12:39:15.226 | + wait_for=119 | 14:08 |
| doffm | alaski: Do you have time to discuss the resource-pool spec again? I spent a good deal of time going over it last night, i'm worried about how it fits with cellsV2. | 14:08 |
| sdague | mriedem: right, and the hard stop in oslo.service is supposed to be 60 seconds I think | 14:09 |
| *** zenoway has joined #openstack-nova | 14:09 | |
| sdague | so that is obviously not working | 14:09 |
| *** dims has joined #openstack-nova | 14:10 | |
| bauzas | sdague: mriedem: https://bugs.launchpad.net/oslo.service/+bug/1543593 | 14:11 |
| openstack | Launchpad bug 1543593 in oslo.service "Grenade old n-api doesn't stop because of eventlet" [Undecided,New] | 14:11 |
| sdague | bauzas: I don't think it's eventlet | 14:11 |
| sdague | but that's fine | 14:11 |
| bauzas | sdague: fair enough, I can modify the title | 14:11 |
| *** andrey-mp has joined #openstack-nova | 14:11 | |
| sdague | make sure the oslo.service folks know about this | 14:11 |
| *** mdrabe has joined #openstack-nova | 14:11 | |
| andrey-mp | Hello | 14:11 |
| mriedem | sdague: bauzas: fyi https://bugs.launchpad.net/nova/+bug/1029727 | 14:12 |
| openstack | Launchpad bug 1029727 in OpenStack Compute (nova) "Using logging module from signal handler can wedge process" [Low,Fix released] - Assigned to Johannes Erdfelt (johannes.erdfelt) | 14:12 |
| mriedem | ^ very old | 14:12 |
| *** omnipresent has joined #openstack-nova | 14:12 | |
| bauzas | mmm | 14:12 |
| *** ctrath has joined #openstack-nova | 14:12 | |
| bauzas | indee | 14:13 |
| bauzas | indeed* | 14:13 |
| *** cdent has quit IRC | 14:13 | |
| bauzas | mriedem: e-r query ? | 14:13 |
| mdbooth | Do we care with Citrix XenServer CI thinks? | 14:13 |
| bauzas | btw. how is elastic-recheck today ? | 14:13 |
| andrey-mp | sdague: thank you for review. Can someone else review this - https://review.openstack.org/#/c/266425/ ? | 14:13 |
| * mdbooth can't remember | 14:13 | |
| mdbooth | If so, is there any way to tell if a patch is currently in its queue? | 14:13 |
| mriedem | bauzas: first day back from vacation so i'm going to assume everything is wonderful | 14:14 |
| bauzas | mdbooth: see http://ci-watch.tintri.com/project?project=nova&time=7+days if that's a CI issue | 14:14 |
| *** zenoway has quit IRC | 14:14 | |
| bauzas | mriedem: ah :) | 14:14 |
| bauzas | well, broncos won | 14:14 |
| mriedem | mdbooth: you care what xenserver ci thinks if your change touches the xenapi driver | 14:14 |
| *** mylu has joined #openstack-nova | 14:14 | |
| *** dave-mccowan has quit IRC | 14:14 | |
| *** baoli has joined #openstack-nova | 14:14 | |
| *** rook has quit IRC | 14:14 | |
| *** knikolla has joined #openstack-nova | 14:15 | |
| *** rook has joined #openstack-nova | 14:15 | |
| bauzas | mriedem: so, I'll take the reformulation card : how can I know if e-r is delayed ? | 14:15 |
| bauzas | or ill ? | 14:15 |
| mriedem | bauzas: it says it's up to date http://status.openstack.org/elastic-recheck/ | 14:15 |
| mriedem | Data Last Updated:Â Tue Feb 09 2016 08:00:02 GMT-0600 (Central Standard Time) | 14:16 |
| bauzas | ack | 14:16 |
| *** daemontool_ has joined #openstack-nova | 14:16 | |
| *** links has joined #openstack-nova | 14:16 | |
| *** omnipresent has quit IRC | 14:16 | |
| *** baoli_ has joined #openstack-nova | 14:17 | |
| mdbooth | mriedem: It's voted -1 on these 2, which is a bit rude as they only touch libvirt: https://review.openstack.org/#/c/270999/3 https://review.openstack.org/#/c/270997/3 | 14:17 |
| *** rlrossit has joined #openstack-nova | 14:17 | |
| *** daemontool has quit IRC | 14:17 | |
| bauzas | mdbooth: as per http://ci-watch.tintri.com/project?project=nova&time=7+days it seems the XenServer CI had some issue recently | 14:18 |
| bauzas | BobBall ? | 14:18 |
| mriedem | Could not stat /dev/vdb --- No such file or directory | 14:19 |
| *** eharney has quit IRC | 14:19 | |
| *** dane-fichter has quit IRC | 14:20 | |
| *** baoli has quit IRC | 14:20 | |
| mriedem | bauzas: mdbooth: should be fixed | 14:20 |
| mriedem | https://github.com/openstack/tempest/commit/bbeab27ab8f05ad1287907222fb6b02990021885 | 14:20 |
| mriedem | it was a bad test in tempest, failed in the gate also apparently | 14:20 |
| bauzas | mriedem: that's what I can see | 14:20 |
| *** ducttape_ has joined #openstack-nova | 14:21 | |
| mriedem | mdbooth: so just recheck those | 14:21 |
| *** jgriffith_away is now known as jgriffith | 14:21 | |
| *** yamahata has joined #openstack-nova | 14:21 | |
| *** toure has joined #openstack-nova | 14:21 | |
| *** ducttape_ has quit IRC | 14:21 | |
| mdbooth | mriedem: Cool, thanks! Could we have reasonably excluded them, btw? I don't know if we're doing path-based checks for driver-specific CI. | 14:21 |
| *** tcammann has joined #openstack-nova | 14:21 | |
| *** amotoki has joined #openstack-nova | 14:21 | |
| mriedem | mdbooth: we aren't. | 14:22 |
| *** sfinucan has quit IRC | 14:22 | |
| mdbooth | k | 14:22 |
| mriedem | personally i don't think it's a bad idea | 14:22 |
| *** andrey-mp has quit IRC | 14:22 | |
| *** sfinucan has joined #openstack-nova | 14:22 | |
| bauzas | johnthetubaguy: FWIW https://blueprints.launchpad.net/nova/+spec/disk-allocation-ratio-to-rt logged | 14:23 |
| bauzas | johnthetubaguy: could you please approve it ? | 14:23 |
| *** gszasz has quit IRC | 14:24 | |
| *** efoley__ is now known as efoley | 14:24 | |
| openstackgerrit | Rossella Sblendido proposed openstack/nova: Wait for neutron events when using xen driver https://review.openstack.org/268258 | 14:24 |
| *** ferest has joined #openstack-nova | 14:24 | |
| johnthetubaguy | bauzas: sure, OK | 14:24 |
| *** vladikr has joined #openstack-nova | 14:25 | |
| *** gszasz has joined #openstack-nova | 14:25 | |
| *** mrkz has joined #openstack-nova | 14:25 | |
| bauzas | johnthetubaguy: thanks | 14:27 |
| *** ferest has quit IRC | 14:28 | |
| *** ferest_ has joined #openstack-nova | 14:28 | |
| *** tcammann has quit IRC | 14:28 | |
| *** gszasz has quit IRC | 14:29 | |
| *** jlibosva has joined #openstack-nova | 14:30 | |
| *** amotoki has quit IRC | 14:31 | |
| rpodolyaka | dansmith: hey! are you around? got a quick question on https://review.openstack.org/#/c/242206 | 14:32 |
| dansmith | rdopiera: yep, shoot | 14:32 |
| rpodolyaka | dansmith: is it expected that services are shown in nova service-list output? or we are missing another filter? | 14:32 |
| rpodolyaka | I mean osapi_compute and metadata | 14:32 |
| rpodolyaka | it's a bit confusing, that they are shown as DOWN, and you just got a response from this very API service :) | 14:33 |
| *** ferest has joined #openstack-nova | 14:33 | |
| rdopiera | dansmith: ok, so we have those bugs about "I tried to delete an instance, and then rabbit crashed/I unplugged the network cable/whatever and now it's stuck in DELETING" | 14:33 |
| dansmith | rdopiera: ah, no, I meant to filter them out like all the other places | 14:33 |
| rdopiera | rpodolyaka: ^^ | 14:33 |
| rpodolyaka | oops :) | 14:33 |
| rpodolyaka | dansmith: ah, ok. I'll propose a quick fix then | 14:33 |
| rpodolyaka | thank you! | 14:34 |
| dansmith | rpodolyaka: thanks! | 14:34 |
| rdopiera | ah, sorry | 14:34 |
| *** mylu has quit IRC | 14:34 | |
| rdopiera | dansmith: let me know when you are caffeinated :) | 14:34 |
| * dansmith is so foncused :P | 14:34 | |
| *** ferest_ has quit IRC | 14:34 | |
| johnthetubaguy | sfinucan: I was wondering about your patch, do these nits/ideas make sense? https://review.openstack.org/#/c/263812/7 | 14:34 |
| *** dave-mccowan has joined #openstack-nova | 14:34 | |
| dansmith | rdopiera: okay, given me 90 seconds :D | 14:34 |
| rdopiera | 79 78 77 76 75 74... | 14:35 |
| *** jgriffith is now known as jgriffith_away | 14:35 | |
| *** amotoki has joined #openstack-nova | 14:35 | |
| *** ferest_ has joined #openstack-nova | 14:35 | |
| *** mylu has joined #openstack-nova | 14:35 | |
| *** smatzek has quit IRC | 14:35 | |
| esracelik | markus_z: can I ask a question | 14:35 |
| markus_z | esracelik: sure | 14:35 |
| sfinucan | johnthetubaguy: Agree on renaming the variable, though best done as a separate change IMO | 14:36 |
| sfinucan | Not sure on listing the opts: this stuff is already "documented" as part of 'genconfig' so do we need more? | 14:36 |
| sfinucan | I'm really not that bothered either way though :) | 14:36 |
| esracelik | cells commits +2'ed with comments. So do I commit a new patch? I mean why +2'ed if there are comments? | 14:37 |
| *** ferest_ has quit IRC | 14:37 | |
| dansmith | rdopiera: okay, caffeination in process, feel free to proceed | 14:37 |
| *** amotoki has quit IRC | 14:37 | |
| esracelik | markus_z: like this one https://review.openstack.org/#/c/260181/ | 14:37 |
| *** gjayavelu has joined #openstack-nova | 14:37 | |
| *** ferest has quit IRC | 14:38 | |
| rdopiera | dansmith: so we have those instances stuck in DELETING state, and we can't reissue delete, because we don't know if there is a message in the queue somewhere or something | 14:38 |
| *** mubirru has quit IRC | 14:38 | |
| markus_z | esracelik: what's the question there? | 14:38 |
| rdopiera | dansmith: there is some code in the compute init that cleans them up on restart, though | 14:38 |
| dansmith | rdopiera: correct | 14:39 |
| *** ijw has quit IRC | 14:39 | |
| esracelik | markus_z: cells commits +2'ed with comments. So do I commit a new patch? I mean why +2'ed if there are comments? | 14:39 |
| rdopiera | dansmith: and we have some periodic code that also cleans up DELETED instances from libvirt if they have been deleted long enough ago | 14:39 |
| *** tongli has joined #openstack-nova | 14:40 | |
| dansmith | rdopiera: yeah, although the periodic doesn't operate if it's just task_state, IIRC right? | 14:40 |
| rdopiera | dansmith: I wonder if we could do something similar with DELETING -- check if the delete was issued two weeks ago, and if so, repeat the delete | 14:40 |
| rdopiera | dansmith: what I have a problem with is that I have no idea about the possible interactions and race conditions | 14:40 |
| markus_z | esracelik: that's ok. A simple follow up patch which addresses the comments is fine. Sometimes it doesn't make sense to block a patch if a simple follow up patch is good enough. | 14:41 |
| dansmith | rdopiera: and you'd detect this from just instance.updated_at? | 14:41 |
| *** amotoki has joined #openstack-nova | 14:41 | |
| *** amotoki has quit IRC | 14:41 | |
| esracelik | markus_z: ooh got it, thanks.. | 14:42 |
| markus_z | esracelik: In that follow up patch just explain in the commit message that this is a follow up to commit xxx | 14:42 |
| esracelik | markus_z: ok, thanks | 14:42 |
| markus_z | esracelik: I would wait until https://review.openstack.org/#/c/260181 is merged in a few hours and then push the new change based on master | 14:43 |
| markus_z | should be easier I guess | 14:43 |
| *** jgriffith_away is now known as jgriffith | 14:43 | |
| *** ociuhandu has joined #openstack-nova | 14:43 | |
| esracelik | markus_z: how do we know that it will be merged in a few hours? | 14:44 |
| rdopiera | dansmith: deleted_at | 14:44 |
| *** zz_dimtruck is now known as dimtruck | 14:44 | |
| rdopiera | dansmith: like the DELETE cleanup code does | 14:44 |
| dansmith | rdopiera: I don't think we set deleted at until the compute node actually does the delete, right? | 14:44 |
| *** moshele has quit IRC | 14:44 | |
| *** gszasz has joined #openstack-nova | 14:45 | |
| dansmith | rdopiera: i.e. not when we set task_state=DELETING | 14:45 |
| rdopiera | dansmith: that's terminated_at | 14:45 |
| markus_z | esracelik: http://status.openstack.org/zuul/ enter the review id 260181 there and you see the queues the change is going through | 14:45 |
| dansmith | rdopiera: deleted_at should only be set when we do instance.destroy() | 14:45 |
| *** ociuhandu has quit IRC | 14:46 | |
| rdopiera | ah, I see | 14:46 |
| *** dmsimard has left #openstack-nova | 14:46 | |
| *** moshele has joined #openstack-nova | 14:46 | |
| esracelik | markus_z: oh, thanks markus_z I have learned much from you | 14:46 |
| rdopiera | I guess that only leaves updated_at | 14:46 |
| markus_z | esracelik: the least I can do. Your work on the options is really great. | 14:47 |
| esracelik | markus_z: thank you.. I hope I can contribute more | 14:47 |
| dansmith | rdopiera: well, my thought on how to make this more robust, was to have a known TTL on messages in rabbit, which ensures that they don't get delivered past a certain time, | 14:48 |
| dansmith | rdopiera: and then have the compute node able to scan task states and restart some operations if 2*TTL has passed | 14:48 |
| *** cfriesen has joined #openstack-nova | 14:49 | |
| rdopiera | dansmith: sounds reasonable | 14:49 |
| *** cdent has joined #openstack-nova | 14:50 | |
| dansmith | rdopiera: I wonder if for delete, since it's special, | 14:50 |
| rdopiera | dansmith: delete can be repeated safely, right? | 14:50 |
| rdopiera | dansmith: I mean, deleteing an already deleted instance doesn't hurt | 14:50 |
| dansmith | rdopiera: if we could put a delete timer in system_metadata or something before we make that call and then check it on the compute and let the thing that normally runs deletes, but skips the DELETING task state go ahead with it | 14:50 |
| dansmith | rdopiera: usually, yeah :) | 14:50 |
| *** lbeliveau has quit IRC | 14:51 | |
| PaulMurray | bauzas, server groups are check in build_run_instance - that's why it doesn't get checked on migration - my mistake | 14:51 |
| johnthetubaguy | sfinucan: yeah, I am not sure really about the reno stuff, agreed the rename is a separate change | 14:51 |
| rdopiera | dansmith: there is some effort to make compute stateful for solving other stuck states problems | 14:51 |
| johnthetubaguy | sfinucan: I am just not sure if folks reading the release notes get all the info they need. Maybe they do. | 14:51 |
| rdopiera | dansmith: but that's going to take a while | 14:51 |
| bauzas | PaulMurray: yeah, that's why I was a bit unclear, I was not remembering me adding there | 14:52 |
| dansmith | rdopiera: yeah | 14:52 |
| rdopiera | dansmith: why the need for a separate timer? | 14:52 |
| *** fawadkhaliq has quit IRC | 14:52 | |
| *** pradk has joined #openstack-nova | 14:52 | |
| dansmith | rdopiera: meaning why can't we apply the system_metadata timer to other operations? | 14:52 |
| gjayavelu | johnthetubaguy: Hi, I see you have blocked this patch https://review.openstack.org/#/c/196171/. It has been tracked under https://etherpad.openstack.org/p/mitaka-nova-priorities-tracking I thought the freeze was for non-priority items | 14:53 |
| gjayavelu | johnthetubaguy: also, you asked this patch to be part of same chain as https://review.openstack.org/252870 | 14:53 |
| dansmith | johnthetubaguy: super important patch that requires your immediate attention: https://review.openstack.org/#/c/277589/ | 14:53 |
| rdopiera | dansmith: why we can't use updated_at | 14:53 |
| *** tpatzig has quit IRC | 14:53 | |
| *** mtanino has joined #openstack-nova | 14:54 | |
| dansmith | rdopiera: because anything else can change updated_at, including info_cache updates, which happen on a timer | 14:54 |
| dansmith | rdopiera: updated_at is probably never more than a few hours old in reality | 14:54 |
| rdopiera | I see | 14:54 |
| *** links has quit IRC | 14:54 | |
| PaulMurray | bauzas, I thought russellb did it? | 14:54 |
| johnthetubaguy | gjayavelu: the priority items are those associated with this agreed list: http://specs.openstack.org/openstack/nova-specs/priorities/mitaka-priorities.html | 14:54 |
| PaulMurray | did you move it | 14:54 |
| dansmith | well, I guess I'm not positive about info_cache, but regardless, probably not a safe thing to use | 14:54 |
| dansmith | for repeatability | 14:54 |
| *** mkoderer has quit IRC | 14:55 | |
| rdopiera | dansmith: and why do we have separate terminated_at and deleted_at? | 14:55 |
| dansmith | rdopiera: I think because we have various levels of delete | 14:55 |
| gjayavelu | johnthetubaguy: oh ok | 14:56 |
| bauzas | PaulMurray: yeah it was a joint effort, I just rebased his series | 14:56 |
| *** esracelik has quit IRC | 14:56 | |
| *** mkoderer has joined #openstack-nova | 14:56 | |
| rdopiera | dansmith: ok, thank you, I will read up on system_metadata and try to write a blueprint for this? | 14:56 |
| *** tpatzig has joined #openstack-nova | 14:56 | |
| *** mylu has quit IRC | 14:56 | |
| johnthetubaguy | gjayavelu: more details here: http://docs.openstack.org/developer/nova/process.html#non-priority-feature-freeze | 14:56 |
| sfinucan | johnthetubaguy: OK, I'll wait on someone else to voice their opinion before making any changes RE: reno | 14:56 |
| johnthetubaguy | sfinucan: its fine, I didn't see the wsgi group name in there before somehow, I think thats specific enough | 14:57 |
| dansmith | rdopiera: sure.. alternately you could also shoot for the general purpose infrastructure of enabling TTLs on the messages too, of you want | 14:57 |
| *** mylu has joined #openstack-nova | 14:57 | |
| dedery | hi cores, can you please review https://review.openstack.org/#/c/270686 ? it's the backport to Liberty of the "nova boot" fails on duplicate vNICs (_gather_port_ids_and_networks) | 14:57 |
| *** ociuhandu has joined #openstack-nova | 14:58 | |
| rdopiera | dansmith: that'd be oslo.messaging, right? | 14:58 |
| *** lbeliveau has joined #openstack-nova | 14:58 | |
| *** burt has joined #openstack-nova | 14:58 | |
| dansmith | rdopiera: o.msg changes might be required, yeah, but it's also something we'd need to build in logic for in nova.. knowing that the ttl is $ttl, we could do certain things | 14:58 |
| dansmith | rdopiera: we also probably don't want the ttl to be the same on all things, or maybe even set on all operations, at least in the beginning | 14:58 |
| rdopiera | dansmith: sounds like a lot of talking to people | 14:59 |
| *** mgoddard_ has joined #openstack-nova | 14:59 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Move Disk allocation ratio to ResourceTracker https://review.openstack.org/277866 | 14:59 |
| *** sudipto has joined #openstack-nova | 14:59 | |
| *** smatzek has joined #openstack-nova | 15:00 | |
| *** ducttape_ has joined #openstack-nova | 15:01 | |
| *** dimtruck is now known as zz_dimtruck | 15:02 | |
| *** ducttape_ has quit IRC | 15:02 | |
| *** mgoddard has quit IRC | 15:02 | |
| *** ducttape_ has joined #openstack-nova | 15:02 | |
| PaulMurray | bauzas, I don't think live migrate has the same retry loop that boot does | 15:03 |
| *** Daisy has joined #openstack-nova | 15:03 | |
| bauzas | PaulMurray: I was on another stuff, but looking | 15:03 |
| dansmith | jaypipes: I think maybe you and cdent and I should have a quick hangout about your comments on that object patch | 15:03 |
| edleafe | bauzas: Why didn't you just rebase https://review.openstack.org/#/c/245619 ? | 15:04 |
| dansmith | jaypipes: because a lot of the things you're asking for make these objects behave differently than the rest | 15:04 |
| *** Daisy has quit IRC | 15:04 | |
| jaypipes | dansmith: unfortunately, I am in meetings for next 3 hours :( | 15:04 |
| PaulMurray | bauzas, in a call now | 15:04 |
| *** Daisy has joined #openstack-nova | 15:04 | |
| dansmith | jaypipes: that is unfortunate | 15:04 |
| jaypipes | dansmith: can we just use IRC | 15:04 |
| jaypipes | ? | 15:04 |
| cdent | I can IRC now or audio later | 15:05 |
| dansmith | jaypipes: I guess | 15:05 |
| *** su_zhang has joined #openstack-nova | 15:05 | |
| dansmith | jaypipes: your desire for _from_db_object() to be a classmethod breaks the usage of it to refresh the objects, and deviates from every other object in the tree | 15:05 |
| *** irenab has quit IRC | 15:05 | |
| dansmith | jaypipes: it means we can't use that code inside create() or save() or anything else, other than a get() type query method | 15:06 |
| openstackgerrit | Moshe Levi proposed openstack/os-vif: Adding VIFHostDevicePCI object https://review.openstack.org/277871 | 15:06 |
| mdbooth | Kinda circling back to where I was earlier, does anybody fancy giving another +2 to these 2 trivial libvirt patches which xenserver CI voted -1 on? https://review.openstack.org/#/c/270999/ https://review.openstack.org/#/c/270997/3 | 15:07 |
| bauzas | edleafe: that's actually a good point | 15:07 |
| bauzas | edleafe: I could recycle this one | 15:07 |
| openstackgerrit | Marc-Andre Gatien proposed openstack/nova: is there anything I need to do please advise.. https://review.openstack.org/277872 | 15:07 |
| openstackgerrit | Marc-Andre Gatien proposed openstack/nova: is there anything I need to do please advise.. https://review.openstack.org/277873 | 15:07 |
| bauzas | edleafe: that said, the proposer is missing a lot of things | 15:07 |
| mdbooth | I xenserver: rechecked them, but it seems like a bit of a waste of resources. | 15:07 |
| bauzas | lemme do that | 15:07 |
| edleafe | bauzas: I was just working on it, and found a bunch of issues, too | 15:07 |
| bauzas | edleafe: I'm handling that | 15:08 |
| bauzas | the whole series should be up soon | 15:08 |
| edleafe | bauzas: be my guest! | 15:08 |
| *** ijw has joined #openstack-nova | 15:08 | |
| *** SridharG has quit IRC | 15:09 | |
| BobBall | mdbooth: Sorry about that - looking at ci-watch it seems there was a broken period earlier today but it's fixed itself before I even realised | 15:09 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Move Disk allocation ratio to ResourceTracker https://review.openstack.org/245619 | 15:09 |
| mdbooth | BobBall: Yeah, it's fine. TBH I was kinda hoping that if they got shoved through the gate then xenserver might not bother? | 15:10 |
| * mdbooth doesn't know how the queues work. | 15:10 | |
| BobBall | Indeed it wont | 15:10 |
| mdbooth | It really doesn't make sense for xenserver ci to run against those 2 patches. | 15:10 |
| BobBall | uhhh actually - what I mean is the gate won't care, but the XS CI will still run the jobs. | 15:10 |
| mdbooth | Ah, ok. I guess it doesn't matter all that much from a resource pov, then. | 15:10 |
| BobBall | I know; but the CI system in use there doesn't have the ability to filter based on files modified (we're still looking to move to jenkins+zuul-based... watch this spacE) | 15:11 |
| *** edtubill has joined #openstack-nova | 15:11 | |
| mdbooth | BobBall: with bated breath :) | 15:12 |
| mdbooth | Mostly, though, I'd just like to clear the easy things from the bottom of my patch queue. | 15:13 |
| BobBall | by watching this space I would still breathe normally if I were you... it's likely to be a month or two before it's switched over :) | 15:13 |
| mdbooth | hehe | 15:13 |
| jaypipes | dansmith: sorry, gimme a few to digest. | 15:14 |
| dansmith | jaypipes: okay, that's the easier discussion point.... | 15:14 |
| dansmith | cdent: do you have changes on top of this set? I need to revise the middle patch to add the uuids | 15:14 |
| *** cdelatte has joined #openstack-nova | 15:14 | |
| cdent | dansmith: I haven't done anything yet, meetings and reviewing so far | 15:14 |
| cdent | so feel free | 15:14 |
| dansmith | cdent: okay | 15:14 |
| openstackgerrit | Andrea Rosa proposed openstack/nova: Add tests to attach/detach vols for shelved server https://review.openstack.org/277463 | 15:15 |
| *** ildikov has quit IRC | 15:19 | |
| *** ildikov has joined #openstack-nova | 15:19 | |
| *** Daisy has quit IRC | 15:20 | |
| openstackgerrit | Roman Podoliaka proposed openstack/nova: Filter APIs out from services list https://review.openstack.org/277881 | 15:20 |
| *** Daisy has joined #openstack-nova | 15:20 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] Trying py27/34 with oslo-master https://review.openstack.org/277884 | 15:23 |
| bauzas | dansmith: correct me if I'm wrong but now that we have the version manifests, I no longer need to bump a version for a dependent NovaObject like ComputeNodeList being incremented if ComputeNode is updated ? | 15:23 |
| dansmith | bauzas: correct | 15:23 |
| bauzas | cool, just wanted to be 100% sure | 15:23 |
| bauzas | because we're keeping the docstrings like # Version 1.14 ComputeNode version 1.14 | 15:24 |
| johnthetubaguy | bauzas: does the test still create a new hash for those bumps? | 15:24 |
| *** Daisy has quit IRC | 15:25 | |
| bauzas | johnthetubaguy: just for the one which is modified I thik | 15:25 |
| dansmith | bauzas: right, which is what we want | 15:25 |
| johnthetubaguy | yeah, that sounds all self-consistent | 15:25 |
| *** mgoddard_ has quit IRC | 15:26 | |
| jaypipes | dansmith: ok, off call. we can do a call now for 35 minutes? | 15:26 |
| *** mgoddard has joined #openstack-nova | 15:26 | |
| jaypipes | google hangout? | 15:26 |
| dansmith | okay | 15:26 |
| bauzas | could I lurk it ? | 15:26 |
| jaypipes | sure | 15:27 |
| jaypipes | https://talkgadget.google.com/hangouts/_/roew7s6ui42qt6nlcyqv2b6ebia?authuser=0&hl=en | 15:27 |
| jaypipes | https://hangouts.google.com/call/roew7s6ui42qt6nlcyqv2b6ebia | 15:27 |
| *** pushkaru has joined #openstack-nova | 15:27 | |
| dansmith | cdent: ^ | 15:28 |
| jaypipes | cdent: ^^ | 15:28 |
| * cdent headphones up | 15:29 | |
| *** jlanoux has quit IRC | 15:29 | |
| *** zakora has quit IRC | 15:30 | |
| *** zakora has joined #openstack-nova | 15:30 | |
| *** zz_dimtruck is now known as dimtruck | 15:31 | |
| *** farid has joined #openstack-nova | 15:34 | |
| *** mragupat has joined #openstack-nova | 15:38 | |
| openstackgerrit | Pushkar Umaranikar proposed openstack/nova: Config options: Centralize debugger options https://review.openstack.org/274298 | 15:38 |
| openstackgerrit | Moshe Levi proposed openstack/os-vif: Remove duplicate os_vif.initialize() https://review.openstack.org/277891 | 15:39 |
| openstackgerrit | Andrea Rosa proposed openstack/nova: Remove useless call to terminate_connection https://review.openstack.org/257853 | 15:39 |
| *** ijw has quit IRC | 15:41 | |
| mfedosin | sdague: hi there :) do you have 5 minutes to discuss glance v2 support in xen plugin? | 15:42 |
| *** krtaylor has quit IRC | 15:42 | |
| jaypipes | ndipanov: ooh, nice catch on the not dirtying the uuid field. | 15:42 |
| dansmith | jaypipes: sometimes I leave little obvious things in there for ndipanov to find.. on purpose of course | 15:42 |
| jaypipes | like a little Easter egg hunt. | 15:43 |
| ndipanov | of course | 15:43 |
| dansmith | :) | 15:43 |
| cdent | I aspire to be a different sized deal from you guys. Being a big deal sounds pretty messy. | 15:43 |
| ndipanov | btw I am writing a kind of lame framework for doing the nova-manage side of this dance | 15:43 |
| ndipanov | will post later today I hope | 15:43 |
| dansmith | cool | 15:43 |
| *** ljxiash has quit IRC | 15:45 | |
| bauzas | ndipanov: dansmith: jaypipes: about providing a nova-manage command for the compute uuid, I was not that convinced, couldn't we just leave the RT provide it and have the object layer initiate it if needed ? | 15:46 |
| dansmith | bauzas: we don't need it for compute node uuid, but we will for other things | 15:47 |
| dansmith | compute node inventory is probably also not really required | 15:47 |
| ndipanov | bauzas, well the idea is that we want to remove any compat code one version later | 15:47 |
| *** marcusvrn_ has quit IRC | 15:47 | |
| bauzas | dansmith: okay cool then | 15:47 |
| bauzas | ndipanov: well, I don't see your point | 15:48 |
| ndipanov | and there could be edge cases with this I think... | 15:48 |
| bauzas | we should already have some compat code in the object later | 15:48 |
| bauzas | layer | 15:48 |
| ndipanov | bauzas, we want to remove that too once the release is doen | 15:48 |
| bauzas | ndipanov: well, I did that with the compute.host field without trouble | 15:48 |
| bauzas | by doing online migrations | 15:48 |
| dansmith | bauzas: right, like flavor compat was removed | 15:49 |
| bauzas | yup | 15:49 |
| openstackgerrit | Dan Smith proposed openstack/nova: Implement an indexed ResourceClass Enum object https://review.openstack.org/273100 | 15:49 |
| openstackgerrit | Dan Smith proposed openstack/nova: resource_providers, allocations and inventories models https://review.openstack.org/275432 | 15:49 |
| openstackgerrit | Dan Smith proposed openstack/nova: WIP resource-provider versioned objects https://review.openstack.org/277466 | 15:49 |
| openstackgerrit | Dan Smith proposed openstack/nova: Allocate UUID for compute node https://review.openstack.org/277554 | 15:49 |
| dansmith | bauzas: the flavor code was definitely something we wanted to remove immediately | 15:49 |
| dansmith | cdent: ^ | 15:49 |
| bauzas | I see | 15:49 |
| bauzas | and not wait for one cycle ? | 15:49 |
| bauzas | then I see | 15:49 |
| dansmith | no, we do | 15:49 |
| ndipanov | so if you keep the compute node turned of for a whole release and then turn it on again when compat code is gone... won't that make it unmigratable? | 15:49 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: [functional] Move code for boot vm to base testcase https://review.openstack.org/277901 | 15:50 |
| cdent | thanks dansmith | 15:50 |
| dansmith | bauzas: the nova-manage thing is for pushing migrations in the background for things that might not be touched | 15:50 |
| dansmith | bauzas: we needed that for flavor | 15:50 |
| markus_z | dims: could you help me to understand the upper-constraint thing? | 15:50 |
| dansmith | bauzas: and I think what ndipanov is saying is a generic framework for pushing those in nova-manage | 15:50 |
| markus_z | dims: This one pinned eventlet to 0.18.1 https://review.openstack.org/#/c/267962/14/upper-constraints.txt | 15:50 |
| dims | markus_z : what's up? | 15:50 |
| bauzas | dansmith: sure, I understood that - I was just asking for the uuid thing | 15:50 |
| dansmith | bauzas: like nova-manage migrate_flavor_data | 15:50 |
| markus_z | dims: but that version is not on pypi | 15:50 |
| dansmith | bauzas: okay | 15:50 |
| markus_z | dims: https://pypi.python.org/pypi/eventlet/ | 15:50 |
| bauzas | dansmith: but yeah, sure a nova-manage framework for the 2nd item to add is a great idea | 15:51 |
| *** angdraug has joined #openstack-nova | 15:51 | |
| dims | markus_z : weird, looks like someone pulled a release | 15:51 |
| ndipanov | dansmith, right so this would allow you to not write all of the boilerplate for those | 15:51 |
| bauzas | just not sure that this 2nd item is the uuid thing :) | 15:51 |
| dansmith | ndipanov: yes, which I should have done in the first place | 15:51 |
| ndipanov | uuid thing is trivial enough that we may even want to leave it | 15:51 |
| ndipanov | but even the pci_device one we got in is quite confusing and should be gone asap | 15:52 |
| dims | markus_z : can you fix it up with 0.18.2 and propose a review? | 15:52 |
| dansmith | ndipanov: well, we can't because we need to land a migration to enforce unique/non-null on that column in the next cycle | 15:52 |
| dims | markus_z : please :) | 15:52 |
| bauzas | that | 15:52 |
| dansmith | ndipanov: so in this case we need to force it in newton, but agreed for other easy things | 15:52 |
| ndipanov | dansmith, yeah ok even better | 15:52 |
| *** ekuris has quit IRC | 15:52 | |
| *** gszasz has quit IRC | 15:52 | |
| markus_z | dims: sure, I was just a bit confused why hell didn't break loose until now | 15:52 |
| dansmith | ndipanov: for liberty we had that barrier migration that just failed if you hadn't done all your migrations.. we need another one of those for this | 15:52 |
| ndipanov | and the pci_device one is like 4 lines and it still should go away asap | 15:52 |
| *** liverpooler has quit IRC | 15:53 | |
| dims | markus_z : so that release is still on our pypi mirrors | 15:53 |
| dims | markus_z : probably | 15:53 |
| *** itamarl has quit IRC | 15:54 | |
| diana_clarke | eventlet | 15:55 |
| *** ljxiash has joined #openstack-nova | 15:55 | |
| diana_clarke | Oops, so sorry. I thought I typed that into search ;( | 15:55 |
| *** nikhil is now known as nikhil_k | 15:56 | |
| markus_z | dims: global-requirements can keep its "eventlet>=0.18.1" or must it be in sync with "upper-constraints.txt"? | 15:56 |
| dims | markus_z : let's bump that too | 15:57 |
| diana_clarke | There are worse things I could be caught publicly searching for... lol | 15:57 |
| markus_z | dims: okidoki. just checking tox and push then in a few moments | 15:57 |
| dims | markus_z : thanks for noticing and fixing :) | 15:58 |
| *** jdurgin1 has joined #openstack-nova | 15:58 | |
| *** moshele has quit IRC | 15:59 | |
| edleafe | diana_clarke: :) | 16:00 |
| *** ljxiash has quit IRC | 16:00 | |
| johnthetubaguy | dansmith: is that save needing to set compute node uuid because its a read_only field? | 16:01 |
| rlrossit | dansmith: cdent: Not sure if this falls under the "resource provider" umbrella, but is there something being done about formalizing resources (https://github.com/openstack/nova/blob/master/nova/objects/compute_node.py#L301-L313)? This stuff is making me rip my hair out for DictCompat | 16:01 |
| dansmith | johnthetubaguy: I'm not sure what you mean | 16:02 |
| *** n0ano has joined #openstack-nova | 16:02 | |
| dansmith | johnthetubaguy: line 310? | 16:02 |
| *** eharney has joined #openstack-nova | 16:02 | |
| dansmith | rlrossit: that's exactly what we're doing homey | 16:02 |
| dansmith | rlrossit: those fields will DIAF soonly | 16:02 |
| cdent | rlrossit: if you want be a nova guy apparently you _have_ to pull out your hair. | 16:02 |
| cdent | all those big deal guys are a bit thin on top | 16:03 |
| rlrossit | dansmith: please finish while I still have hair | 16:03 |
| dansmith | cdent: excuse me? :) | 16:03 |
| * cdent scampers away | 16:03 | |
| edleafe | dansmith: that toupee isn't fooling anyone | 16:03 |
| *** mjura has quit IRC | 16:03 | |
| *** gokrokve has joined #openstack-nova | 16:04 | |
| *** sinese_ has quit IRC | 16:04 | |
| *** lpetrut has quit IRC | 16:04 | |
| dansmith | edleafe: in austin, I will permit one tug on my hair to defend my dignity | 16:05 |
| *** jlanoux has joined #openstack-nova | 16:05 | |
| *** ijw has joined #openstack-nova | 16:05 | |
| edleafe | dansmith: lol | 16:05 |
| PaulMurray | edleafe, bring a camera with wifi | 16:06 |
| * PaulMurray or is that called a phone these days | 16:07 | |
| edleafe | PaulMurray: I'll post it on YouTube | 16:07 |
| *** mvk has quit IRC | 16:07 | |
| *** armax has joined #openstack-nova | 16:08 | |
| *** fawadkhaliq has joined #openstack-nova | 16:08 | |
| johnthetubaguy | dansmith: sorry, looking again | 16:08 |
| *** vilobhmm11 has joined #openstack-nova | 16:08 | |
| *** danpb has left #openstack-nova | 16:09 | |
| johnthetubaguy | dansmith: yeah, 310 | 16:09 |
| *** gmann_ has quit IRC | 16:09 | |
| johnthetubaguy | dansmith: just wondering what cases are not covered by from_db and create | 16:09 |
| dansmith | johnthetubaguy: that's just avoiding generating a uuid if we're saving with a uuid already set | 16:09 |
| dansmith | johnthetubaguy: you mean how could we get to this case? | 16:09 |
| diana_clarke | Anyhoo, the real reason I stopped by was to thank you all for reviewing my unit test profiling patches. | 16:10 |
| diana_clarke | Here's the last one in the series that needs a review: https://review.openstack.org/#/c/277208/ Thanks! | 16:10 |
| *** burgerk has joined #openstack-nova | 16:10 | |
| *** mdrabe has quit IRC | 16:10 | |
| dansmith | johnthetubaguy: older compute node pulls a ComputeNode, we strip uuid from the response when we generate it because it's backported, then when save is called, we need to make sure that we have a uuid there again | 16:11 |
| dansmith | johnthetubaguy: although you have a good point.. that uuid will keep changing until we're upgraded... | 16:11 |
| *** boris-42 has joined #openstack-nova | 16:11 | |
| *** sc68cal has joined #openstack-nova | 16:11 | |
| dansmith | johnthetubaguy: I probably need to do a less efficient-but-only-once thing there | 16:11 |
| dansmith | johnthetubaguy: not sure if that was what you were getting at, but .. good catch! :) | 16:11 |
| *** daemontool__ has joined #openstack-nova | 16:12 | |
| *** PaulMurray has quit IRC | 16:12 | |
| *** daemontool__ is now known as daemontool | 16:13 | |
| *** krtaylor has joined #openstack-nova | 16:14 | |
| bauzas | dansmith: now that old computes all use the object layer, they're always getting an uuid, right? | 16:14 |
| bauzas | I mean | 16:14 |
| bauzas | sec | 16:14 |
| dansmith | bauzas: no, because if they ask for an older version, we won't pass uuid back to them, since they won't know what to do with it | 16:14 |
| bauzas | sec | 16:15 |
| bauzas | just thinking of something | 16:15 |
| *** daemontool_ has quit IRC | 16:15 | |
| bauzas | if that's a remotable method, you can save and persist an uuid | 16:15 |
| bauzas | that's just the old compute which will no longer know that it has an uuid, right? | 16:16 |
| johnthetubaguy | dansmith: thats basically it, I wonder if in load_from_db we should do the save once we generate the uuid, for a similar reason, but maybe thats the same way around | 16:17 |
| bauzas | so the updates are not including the uuid, but we could generate once | 16:17 |
| dansmith | bauzas: yes, but the code was only looking at uuid in the received object, not in the db, and the received one will never have uuid if coming from an older node | 16:17 |
| bauzas | I see | 16:18 |
| dansmith | johnthetubaguy: we could, but for flavor we did it this way | 16:18 |
| dansmith | johnthetubaguy: it's more awkward to do it there | 16:18 |
| openstackgerrit | Dan Smith proposed openstack/nova: WIP resource-provider versioned objects https://review.openstack.org/277466 | 16:20 |
| openstackgerrit | Dan Smith proposed openstack/nova: Allocate UUID for compute node https://review.openstack.org/277554 | 16:20 |
| dansmith | johnthetubaguy: ^ | 16:21 |
| *** belmoreira has quit IRC | 16:21 | |
| *** mdrabe has joined #openstack-nova | 16:24 | |
| *** _ducttape_ has joined #openstack-nova | 16:26 | |
| *** knikolla has quit IRC | 16:27 | |
| *** knikolla has joined #openstack-nova | 16:27 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Move Disk allocation ratio to ResourceTracker https://review.openstack.org/245619 | 16:28 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Add disk_allocation_ratio to ComputeNode https://review.openstack.org/277934 | 16:28 |
| *** ducttape_ has quit IRC | 16:29 | |
| *** ijw has quit IRC | 16:29 | |
| johnthetubaguy | dansmith: OK, will take another look, I just worry about two people generating different uuid, and using that in other tables, but I am probably worrying too much | 16:32 |
| *** browne has joined #openstack-nova | 16:32 | |
| *** zenoway has joined #openstack-nova | 16:32 | |
| dansmith | johnthetubaguy: yeah, but I don't think anything other than the one compute node will be doing get..save operations | 16:33 |
| johnthetubaguy | dansmith: true | 16:33 |
| *** _ducttape_ has quit IRC | 16:36 | |
| *** zenoway has quit IRC | 16:37 | |
| *** ducttape_ has joined #openstack-nova | 16:37 | |
| openstackgerrit | Ryan Rossiter proposed openstack/nova: Remove DictCompat from ComputeNode https://review.openstack.org/265521 | 16:40 |
| *** salv-orl_ has joined #openstack-nova | 16:41 | |
| *** zakora has quit IRC | 16:41 | |
| * rlrossit throws a hammer at gerrit | 16:42 | |
| hemna | gus, ping | 16:42 |
| openstackgerrit | Pushkar Umaranikar proposed openstack/nova: Config options: Centralize debugger options https://review.openstack.org/274298 | 16:43 |
| cdent | rlrossit: "Because getattr() needs AttributeError in order to return the default, your pattern doesn't work." | 16:44 |
| * cdent cries | 16:44 | |
| *** krtaylor has quit IRC | 16:44 | |
| * rlrossit hands cdent tissues | 16:44 | |
| *** salv-orlando has quit IRC | 16:44 | |
| rlrossit | I'm right there with ya, bud | 16:44 |
| cdent | It feels like you've found a bad UX in ovo? | 16:44 |
| rlrossit | makes me want to ask dansmith if we want to change that behavior | 16:44 |
| *** PaulMurray has joined #openstack-nova | 16:44 | |
| cdent | yeah | 16:44 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Update ComputeNode values with disk allocation ratios in the RT https://review.openstack.org/277953 | 16:45 |
| * cdent flutters his eyelashes at dansmith | 16:45 | |
| dansmith | cdent: rlrossit: I mildly remember this, but link me | 16:45 |
| dansmith | (also about to jump on a call) | 16:45 |
| dansmith | however, there are very specific reasons for the way we handle defaulting | 16:45 |
| dansmith | because objects are not always fully loaded from the database | 16:45 |
| cdent | (gerrit has gone slow) | 16:46 |
| *** mgoddard_ has joined #openstack-nova | 16:46 | |
| *** paul-carlton1 has quit IRC | 16:46 | |
| cdent | dansmith: https://review.openstack.org/#/c/265521/9/nova/scheduler/ironic_host_manager.py | 16:46 |
| rlrossit | dansmith: I will try to find it before I have to go eat some Polish pastry of some sort | 16:46 |
| rlrossit | thanks cdent. Gerrit hates me | 16:46 |
| *** yamahata has quit IRC | 16:47 | |
| cdent | gerrit hates all | 16:47 |
| bauzas | johnthetubaguy: all the changes for the disk ratio are in flight https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/disk-allocation-ratio-to-rt | 16:48 |
| dansmith | rlrossit: cdent: if you have non-lazy-loadable things, then just return an attribute error in your obj_load_attr() method for that thing | 16:48 |
| dansmith | if you don't implement that, and you try to lazy load a thing, | 16:48 |
| dansmith | then NotImplementError is the right thing... | 16:48 |
| *** jlibosva has quit IRC | 16:49 | |
| *** mgoddard has quit IRC | 16:49 | |
| rlrossit | dansmith: the problem is coming with the obj_attr_is_set() though (not sure if talking past you) | 16:49 |
| dansmith | rlrossit: not sure I follow | 16:50 |
| rlrossit | I'm not sure if I follow what I say either | 16:50 |
| dansmith | call now, lets resume in a few | 16:50 |
| *** speller has quit IRC | 16:50 | |
| *** knikolla has quit IRC | 16:50 | |
| *** vishwanathj has joined #openstack-nova | 16:51 | |
| johnthetubaguy | bauzas: do you remember what we did last time for letting people update the CONF files as the configuration moved between the services? | 16:52 |
| bauzas | no sorry | 16:52 |
| bauzas | I probably missed that | 16:52 |
| *** scheuran has quit IRC | 16:52 | |
| bauzas | oh you mean when we did the cpu and ram things | 16:53 |
| johnthetubaguy | bauzas: me neither, I think we had a way to let the config be in both places for one release | 16:53 |
| johnthetubaguy | bauzas: yeah, with the CPU and RAM | 16:53 |
| bauzas | the concensus was to say that the conf files had to be consistent per cloud | 16:53 |
| bauzas | hence my reno file | 16:53 |
| bauzas | that I'm overriding with the last change | 16:54 |
| bauzas | johnthetubaguy: that's the trick we did for the RAM and CPU | 16:54 |
| johnthetubaguy | bauzas: I couldn't remember where that landed, OK | 16:54 |
| *** Raildo has joined #openstack-nova | 16:55 | |
| *** Raildo is now known as raildo | 16:55 | |
| bauzas | johnthetubaguy: I could point you the old changes | 16:55 |
| bauzas | johnthetubaguy: I just mocked them for the above... | 16:55 |
| *** krtaylor has joined #openstack-nova | 16:55 | |
| bauzas | the strategy was (and is) : | 16:55 |
| *** apoorvad has joined #openstack-nova | 16:56 | |
| bauzas | #1 modify the disk filter to opt_import the compute namespace | 16:56 |
| bauzas | #2 add a persistence layer for those compute values | 16:56 |
| bauzas | #3 flag a specific value for knowing whether the opt was updated by the op or not | 16:57 |
| johnthetubaguy | bauzas: so I think we did actually let the scheduler value over-ride the compute value for a little bit. | 16:57 |
| bauzas | that's still the case, the logic is still in the object | 16:57 |
| bauzas | and the ram and disk values are still 0.0 | 16:57 |
| bauzas | that's something we promised to land in mitaka (ie. me modifying the default) | 16:58 |
| *** jgriffith is now known as jgriffith_away | 16:58 | |
| *** rdopiera has quit IRC | 16:58 | |
| *** dimtruck is now known as zz_dimtruck | 16:58 | |
| *** zz_dimtruck is now known as dimtruck | 16:58 | |
| bauzas | that said, I'm fine waiting for newton and update all of them | 16:58 |
| * bauzas needs to drop off | 16:59 | |
| *** jlanoux has quit IRC | 16:59 | |
| *** gjayavelu has quit IRC | 16:59 | |
| *** achanda has joined #openstack-nova | 16:59 | |
| johnthetubaguy | bauzas: yeah, lets do that all at once | 16:59 |
| *** raildo has quit IRC | 16:59 | |
| *** achanda has quit IRC | 16:59 | |
| johnthetubaguy | bauzas: I think the idea of that was to ensure the configuration not being set in the computes still works correctly | 17:00 |
| *** mragupat has quit IRC | 17:00 | |
| *** jlanoux has joined #openstack-nova | 17:00 | |
| bauzas | exactly | 17:00 |
| *** mragupat has joined #openstack-nova | 17:01 | |
| bauzas | that's what my 2nd patch adds for the disk | 17:01 |
| bauzas | all of that is when you access the computenode object | 17:01 |
| bauzas | depending on whether you can source it from the db or from the controller | 17:01 |
| *** mpavone has quit IRC | 17:01 | |
| bauzas | oh sec, I missed the sched bit | 17:02 |
| bauzas | I'll write that up later in the evening | 17:02 |
| *** raildo-afk is now known as raildo | 17:03 | |
| *** jaosorior has quit IRC | 17:04 | |
| johnthetubaguy | bauzas: yeah, you need the scheduler bit | 17:04 |
| bauzas | johnthetubaguy: so, the first change doesn't change the opt values | 17:04 |
| bauzas | it only changes how you access them | 17:04 |
| bauzas | johnthetubaguy: but I hope my last patch will clarify that | 17:05 |
| * bauzas needs to run | 17:05 | |
| *** daemontool has quit IRC | 17:06 | |
| *** vilobhmm11 has quit IRC | 17:07 | |
| cdent | dansmith: if we do the compute uuid migration in with the resource provider models migration, then we need at least some of your https://review.openstack.org/277554 in parent for the tests to be happy | 17:08 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: [functional] Move code for boot vm to base testcase https://review.openstack.org/277901 | 17:08 |
| *** mylu has quit IRC | 17:10 | |
| *** jlanoux has quit IRC | 17:10 | |
| openstackgerrit | Andrea Rosa proposed openstack/nova: WIP: Abort an ongoing live migration https://review.openstack.org/277971 | 17:11 |
| *** gjayavelu has joined #openstack-nova | 17:11 | |
| *** mfedosin has quit IRC | 17:11 | |
| *** yamahata has joined #openstack-nova | 17:13 | |
| *** paul-carlton1 has joined #openstack-nova | 17:14 | |
| mriedem | since we don't have quotas on disk usage, is there another way to restrict that per project? | 17:14 |
| johnthetubaguy | mriedem: I guess its in-directly restricted by the flavors? | 17:14 |
| johnthetubaguy | mriedem: i.e. if you set a quota on RAM only, and you can only use the ratio available in the flavour | 17:15 |
| johnthetubaguy | mriedem: assuming cinder does quota separately for all your volumes | 17:15 |
| *** ijw has joined #openstack-nova | 17:15 | |
| mriedem | yeah, cinder has quotas for volume size limits and total storage used | 17:16 |
| *** zenoway has joined #openstack-nova | 17:19 | |
| *** su_zhang has quit IRC | 17:21 | |
| *** Bjoern has joined #openstack-nova | 17:21 | |
| mriedem | thomasem: did you still need this for testing or can i drop it? https://review.openstack.org/#/c/240611/ | 17:22 |
| *** nic has joined #openstack-nova | 17:22 | |
| *** zenoway has quit IRC | 17:24 | |
| Bjoern | Hi Folks, does anyone know when numa_topology gets populated inside the nova/compute_nodes table ? That behavior seem to have changes with kilo. My problem is that the topology does not get updated from nova_compute altough the virsh capabilities list two cells | 17:24 |
| Bjoern | Expected something like https://github.com/openstack/nova/blob/1635086529c43edcb69966c48c41b5fca2d1b26c/doc/source/testing/libvirt-numa.rst | 17:24 |
| -openstackstatus- NOTICE: Gerrit is restarting now, to alleviate current performance impact and WebUI errors. | 17:24 | |
| *** matrohon has quit IRC | 17:25 | |
| *** shawnmaten has joined #openstack-nova | 17:26 | |
| *** mylu has joined #openstack-nova | 17:27 | |
| *** ssurana has joined #openstack-nova | 17:28 | |
| Bjoern | oh I see libvirt version is now checked and it needs to be version 1.2.7 before that code even looks at the topology | 17:30 |
| *** jdurgin1 has quit IRC | 17:30 | |
| *** ssurana has left #openstack-nova | 17:32 | |
| *** e0ne has quit IRC | 17:32 | |
| *** unicell has quit IRC | 17:32 | |
| *** gjayavelu has quit IRC | 17:32 | |
| *** mragupat has quit IRC | 17:32 | |
| *** mragupat has joined #openstack-nova | 17:33 | |
| *** knikolla has joined #openstack-nova | 17:33 | |
| *** gjayavelu has joined #openstack-nova | 17:33 | |
| *** mylu has quit IRC | 17:34 | |
| *** sfinucan_ has joined #openstack-nova | 17:34 | |
| *** gokrokve_ has joined #openstack-nova | 17:34 | |
| *** ijw has quit IRC | 17:36 | |
| dansmith | cdent: I'm not sure I follow.. we have to have that patch in front of any of the stuff to start moving inventory stuff into those tables, yes | 17:36 |
| *** jdurgin1 has joined #openstack-nova | 17:36 | |
| *** jdurgin1 has left #openstack-nova | 17:36 | |
| *** ssurana has joined #openstack-nova | 17:36 | |
| *** mylu has joined #openstack-nova | 17:36 | |
| *** mragupat has quit IRC | 17:36 | |
| *** ijw has joined #openstack-nova | 17:37 | |
| *** sfinucan has quit IRC | 17:37 | |
| *** gokrokve has quit IRC | 17:37 | |
| *** mragupat has joined #openstack-nova | 17:37 | |
| *** shawnmaten has quit IRC | 17:37 | |
| *** gjayavelu has quit IRC | 17:37 | |
| cdent | dansmith: nova.tests.unit.db.test_db_api.ComputeNodeTestCase.test_compute_node_get fails in https://review.openstack.org/#/c/275432/ | 17:38 |
| cdent | because uuid has been added | 17:38 |
| cdent | (there are some other failures which appear to be entirely unrelated, investigating those concurrently) | 17:38 |
| *** paul-carlton1 has quit IRC | 17:39 | |
| dansmith | cdent: oh, thought you mean the objects patch | 17:39 |
| *** jaosorior has joined #openstack-nova | 17:39 | |
| cdent | dansmith: so I'm going ahead and putting the uuid sentinel in | 17:40 |
| cdent | but the PING PONG test failures are...weird | 17:40 |
| *** zenoway has joined #openstack-nova | 17:40 | |
| cdent | ah, they are present in master too | 17:40 |
| dansmith | cdent: Just filter it out of the things that we load on compute node get, and then let the patch above remove the exclusion and add in the code | 17:41 |
| *** zenoway has quit IRC | 17:41 | |
| *** ijw has quit IRC | 17:41 | |
| *** mrkz has quit IRC | 17:41 | |
| *** zenoway has joined #openstack-nova | 17:41 | |
| *** mrkz has joined #openstack-nova | 17:42 | |
| notmorgan | mriedem, sdague: going to revisit novaclient conversion ot OCC today, will ping you when it's posted so we can have novaclient happily on the new shiney stuff. | 17:42 |
| cdent | dansmith: srsly? that's...awkward, but you are a big deal ;) | 17:42 |
| dansmith | cdent: seriously what? seriously filter? | 17:43 |
| *** samueldmq has joined #openstack-nova | 17:44 | |
| dansmith | hang on maybe I'm not fully grasping the fail.. let me run it locally | 17:44 |
| cdent | k | 17:44 |
| johnthetubaguy | dansmith: added another comment on the latest patch, not sure how that works, I could be missing something: https://review.openstack.org/#/c/277554/4 | 17:45 |
| dansmith | johnthetubaguy: okay just a sec | 17:45 |
| *** browne has quit IRC | 17:46 | |
| *** sinese_ has joined #openstack-nova | 17:46 | |
| dansmith | cdent: oh, I thought this was the object test, sorry :) | 17:46 |
| cdent | dansmith: we can just stick uuid in _ignored_keys and then pull it back out | 17:46 |
| *** BobBall is now known as BobBall_AWOL | 17:46 | |
| cdent | but that feels like a dirty hack or something | 17:46 |
| dansmith | cdent: but you can just add it to compute_node_dict anyway, like I do in the next patch | 17:47 |
| cdent | yeah, that's what I said above: "so I'm going ahead and putting the uuid sentinel in" | 17:47 |
| cdent | so that's cool? | 17:47 |
| cdent | (are you getting the PING and PONG failures?) | 17:47 |
| dansmith | cdent: yeah, sorry I was thinking this was the object test, not db_api so yeah you're good | 17:47 |
| cdent | cool | 17:47 |
| dansmith | cdent: nope: Failed 2 tests - output below: | 17:47 |
| dansmith | just those two | 17:48 |
| cdent | is your venv stale? | 17:48 |
| dansmith | no just rebuilt it this morning | 17:48 |
| cdent | that's pretty stale ;) | 17:48 |
| * dansmith gives it the sniff test | 17:48 | |
| *** sfinucan_ has quit IRC | 17:49 | |
| cdent | I'll push this up again and see if it gives the unrelated failures again (there's in the py27 logs if you want to see) | 17:49 |
| dansmith | cdent: oh, I only ran the db tests, not that ssl one so I wouldn't know | 17:49 |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/nova: Revert "Revert "Added new scheduler filter: AggregateTypeExtraSpecsAffinityFilter"" https://review.openstack.org/275128 | 17:49 |
| cdent | no worries, I'll push that down the road a bit | 17:49 |
| *** achanda has joined #openstack-nova | 17:50 | |
| *** eil397 has joined #openstack-nova | 17:50 | |
| dansmith | johnthetubaguy: so the new code will generate the uuid in load (from_db) and then it will be dirty (because we're setting it after the reset_changes), so that the save after the fact will save it out | 17:51 |
| dansmith | johnthetubaguy: however, if you want to save it in _from_db (sounds like you do) I'll just do that :) | 17:51 |
| *** Swami has joined #openstack-nova | 17:51 | |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/nova: Revert "Revert "Added new scheduler filter: AggregateTypeExtraSpecsAffinityFilter"" https://review.openstack.org/275128 | 17:52 |
| *** ildikov has quit IRC | 17:53 | |
| openstackgerrit | Chris Dent proposed openstack/nova: resource_providers, allocations and inventories models https://review.openstack.org/275432 | 17:53 |
| markus_z | FYI: nova-bugs-team meeting in ~10min in #openstack-meeting-4 | 17:53 |
| *** jaosorior has quit IRC | 17:53 | |
| cdent | Sigh, yup, looks like nova.tests.unit.test_wsgi.TestWSGIServerWithSSL is failing everywhere | 17:53 |
| dansmith | sdague: ^ | 17:53 |
| *** knikolla has quit IRC | 17:54 | |
| johnthetubaguy | dansmith: I was thinking that the uuid would be set, so the second save would not happen | 17:55 |
| *** lpetrut has joined #openstack-nova | 17:55 | |
| *** jordanP has joined #openstack-nova | 17:55 | |
| *** mylu has quit IRC | 17:56 | |
| markus_z | dansmith: Maybe interesting for you: https://bugs.launchpad.net/nova/+bug/1539342 | 17:57 |
| openstack | Launchpad bug 1539342 in OpenStack Compute (nova) "AttributeError raised during roll_back_live_migration_at_destination" [Undecided,New] - Assigned to Zhenyu Zheng (zhengzhenyu) | 17:57 |
| *** mylu has joined #openstack-nova | 17:57 | |
| dansmith | markus_z: ugh | 17:58 |
| *** knikolla has joined #openstack-nova | 17:58 | |
| *** mylu has quit IRC | 17:58 | |
| *** derekh has quit IRC | 17:58 | |
| *** jwang has joined #openstack-nova | 18:00 | |
| *** jwang_ has quit IRC | 18:00 | |
| *** sudipto has quit IRC | 18:00 | |
| *** mylu has joined #openstack-nova | 18:00 | |
| *** penick has joined #openstack-nova | 18:00 | |
| *** vishwanathj has quit IRC | 18:00 | |
| thomasem | mriedem: actually, can you assign that to me? I'd like to investigate more... we had some prod issues come up that have taken up all of my time the past week or so. | 18:00 |
| *** angdraug has quit IRC | 18:02 | |
| *** mgoddard_ has quit IRC | 18:02 | |
| *** mgoddard has joined #openstack-nova | 18:02 | |
| *** bnemec has joined #openstack-nova | 18:02 | |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/nova: Revert "Revert "Added new scheduler filter: AggregateTypeExtraSpecsAffinityFilter"" https://review.openstack.org/275128 | 18:02 |
| *** cdent has quit IRC | 18:02 | |
| mriedem | thomasem: np, i'll just leave it, rebase whenever you want/need | 18:02 |
| thomasem | mriedem: awesome, thank you | 18:02 |
| *** jordanP has quit IRC | 18:03 | |
| *** mylu has quit IRC | 18:03 | |
| * mriedem wishes people would use tags or git hashes when putting code links into bug reports | 18:05 | |
| *** penick has quit IRC | 18:05 | |
| *** su_zhang has joined #openstack-nova | 18:05 | |
| *** ctrath has quit IRC | 18:06 | |
| *** eil397 has quit IRC | 18:07 | |
| *** mylu has joined #openstack-nova | 18:08 | |
| *** ctrath has joined #openstack-nova | 18:08 | |
| *** eil397 has joined #openstack-nova | 18:08 | |
| *** efoley has quit IRC | 18:09 | |
| *** gokrokve_ has quit IRC | 18:11 | |
| *** unicell has joined #openstack-nova | 18:11 | |
| *** lucasagomes is now known as lucas-dinner | 18:11 | |
| *** rotbeard has quit IRC | 18:11 | |
| openstackgerrit | Tracy Jones proposed openstack/nova: config options: centralize section "service" https://review.openstack.org/256569 | 18:15 |
| *** sahid has quit IRC | 18:16 | |
| *** e0ne has joined #openstack-nova | 18:17 | |
| openstackgerrit | Dan Smith proposed openstack/nova: WIP resource-provider versioned objects https://review.openstack.org/277466 | 18:17 |
| openstackgerrit | Dan Smith proposed openstack/nova: Allocate UUID for compute node https://review.openstack.org/277554 | 18:17 |
| dansmith | johnthetubaguy: ^ | 18:18 |
| dansmith | johnthetubaguy: I think it's probably better and safer this way anyway, so .. | 18:18 |
| johnthetubaguy | dansmith: did you mean to leave that note in that says it needs to come after the thing it comes before? | 18:19 |
| dansmith | johnthetubaguy: say what now? | 18:19 |
| dansmith | johnthetubaguy: oh | 18:20 |
| dansmith | no | 18:20 |
| dansmith | :D | 18:20 |
| johnthetubaguy | dansmith: yeah, added a comment | 18:20 |
| johnthetubaguy | dansmith: I think that makes your save() do too much saving actually, because I guess all the fields are dirty at that point? | 18:21 |
| dansmith | dammit | 18:21 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Update import_file to provide a general-purpose interface https://review.openstack.org/267662 | 18:21 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Implement import_file for Qcow2 backend https://review.openstack.org/278010 | 18:21 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Implement import_file for LVM backend https://review.openstack.org/278011 | 18:21 |
| openstackgerrit | Matthew Booth proposed openstack/nova: ploop import_file https://review.openstack.org/278012 | 18:21 |
| openstackgerrit | Matthew Booth proposed openstack/nova: abstract method https://review.openstack.org/278013 | 18:21 |
| dansmith | johnthetubaguy: I kinda liked removing that extra reset, but yeah, maybe that's the thing to do | 18:22 |
| johnthetubaguy | dansmith: yeah, its annoying to re-call the reset for the uuid thing, I wondered about splitting up the two sides of that if statement, but maybe thats worse? | 18:22 |
| openstackgerrit | Tracy Jones proposed openstack/nova: config options: centralize section "volume" https://review.openstack.org/253747 | 18:22 |
| dansmith | johnthetubaguy: well, we can't really because we can't set the potentially-NULL uuid on the object | 18:23 |
| dansmith | unless we break it up and do the test again, but meh | 18:23 |
| dansmith | johnthetubaguy: we get to remove this in newton, so it's probably easier to keep it small | 18:23 |
| *** delattec has joined #openstack-nova | 18:23 | |
| johnthetubaguy | dansmith: yeah, I was thinking one side only set if not null, but like you say, thats harder to remove | 18:24 |
| *** PaulMurray has quit IRC | 18:24 | |
| *** daemontool has joined #openstack-nova | 18:25 | |
| *** rk4n has quit IRC | 18:25 | |
| *** cdelatte has quit IRC | 18:26 | |
| *** browne has joined #openstack-nova | 18:26 | |
| dansmith | johnthetubaguy: actually, it might make it easier if I change one other thing.. just a sec | 18:27 |
| dansmith | I'm adding notes about when to remove each piece | 18:27 |
| johnthetubaguy | dansmith: oh, good plan | 18:27 |
| openstackgerrit | Dan Smith proposed openstack/nova: WIP resource-provider versioned objects https://review.openstack.org/277466 | 18:29 |
| openstackgerrit | Dan Smith proposed openstack/nova: Allocate UUID for compute node https://review.openstack.org/277554 | 18:29 |
| *** su_zhang has quit IRC | 18:29 | |
| dansmith | it's two pieces now, which looks like more, but both will just be deleted in newton, so maybe that's better | 18:29 |
| *** su_zhang has joined #openstack-nova | 18:29 | |
| alaski | dansmith: https://review.openstack.org/#/c/277589/ should reserve some spots for api db migrations as well | 18:32 |
| openstackgerrit | Tracy Jones proposed openstack/nova: config options: centralize section "service" https://review.openstack.org/256569 | 18:32 |
| *** aginwala has joined #openstack-nova | 18:32 | |
| dansmith | alaski: oh, interesting, didn't think about that | 18:32 |
| *** mragupat_ has joined #openstack-nova | 18:32 | |
| alaski | I hadn't either until now, but we probably should have done that this cycle | 18:33 |
| *** irenab has joined #openstack-nova | 18:33 | |
| alaski | 10 placeholders is probably a bit much though | 18:33 |
| *** jgriffith_away is now known as jgriffith | 18:34 | |
| alaski | considering there are currently 4 migrations there | 18:34 |
| dansmith | okay | 18:34 |
| dansmith | Maybe I'll just make an option to do api ones, | 18:35 |
| dansmith | and then you can do --api -n 2, or something | 18:35 |
| *** ijw has joined #openstack-nova | 18:35 | |
| alaski | works for me | 18:36 |
| *** mragupat has quit IRC | 18:36 | |
| *** ildikov has joined #openstack-nova | 18:36 | |
| *** ociuhandu has quit IRC | 18:39 | |
| openstackgerrit | Stephen Finucane proposed openstack/nova: config options: Remove 'wsgi_' prefix from opts https://review.openstack.org/278023 | 18:39 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: trivial: Fix alignment of wsgi options https://review.openstack.org/278024 | 18:39 |
| openstackgerrit | Dan Smith proposed openstack/nova: Add a tool for reserving migration placeholders during release time https://review.openstack.org/277589 | 18:40 |
| *** baoli_ has quit IRC | 18:40 | |
| dansmith | alaski: ^ | 18:40 |
| *** eil397 has quit IRC | 18:41 | |
| *** shawnmaten has joined #openstack-nova | 18:42 | |
| openstackgerrit | Jay Pipes proposed openstack/nova-specs: resource-providers: generic resource pools https://review.openstack.org/253187 | 18:42 |
| *** cdent has joined #openstack-nova | 18:43 | |
| *** ijw has quit IRC | 18:43 | |
| johnthetubaguy | dansmith: alaski: so these new host related tables for the scheduler, they make sense living in the child cell DB, or should they be in a new scheduler db? | 18:44 |
| dansmith | johnthetubaguy: you mean the resource pool tables? | 18:44 |
| johnthetubaguy | yeah | 18:44 |
| dansmith | johnthetubaguy: we hashed that out yesterday | 18:45 |
| *** aginwala has quit IRC | 18:45 | |
| johnthetubaguy | ah, sorry, I missed all that, I should read that back | 18:45 |
| *** mylu has quit IRC | 18:45 | |
| dansmith | johnthetubaguy: one reason jaypipes wants this to be generic is so we can lift these tables out and into scheduler some day without a bunch of relations to break | 18:45 |
| dansmith | johnthetubaguy: but for now, in the cell | 18:45 |
| johnthetubaguy | Ok, I am cool with that | 18:45 |
| alaski | yeah, in the cell works for now | 18:45 |
| jaypipes | ++ | 18:45 |
| johnthetubaguy | I think a scheduler fetching from multiple cells, like the API, works just fine (in my head) | 18:46 |
| * cdent schwews | 18:46 | |
| alaski | I suggested a new scheduler db because I think that's closer to where we want to eventually be, but it also derails things to worry about that now | 18:46 |
| johnthetubaguy | alaski: yeah, was thinking along the same lines, cools | 18:46 |
| *** rlrossit has quit IRC | 18:46 | |
| dansmith | yeah, so *two* new required databases in mitaka, that would be ... awesome :D | 18:46 |
| *** matrohon has joined #openstack-nova | 18:46 | |
| * johnthetubaguy makes evil cackle noises | 18:46 | |
| *** chhavi has quit IRC | 18:47 | |
| johnthetubaguy | anyways, that works for me | 18:47 |
| alaski | aren't microservices all the rage now with little datastores everywhere | 18:47 |
| dansmith | yes, nova should store some of its information in SMTP queues around the world | 18:48 |
| dansmith | eventually consistent .. the emails will circle and come back to us over time -- free load balancing! | 18:48 |
| johnthetubaguy | alaski: I guess we do midi-services? | 18:48 |
| alaski | dansmith: lol, though I have seen something like that implemented | 18:48 |
| dansmith | alaski: I know :) | 18:48 |
| *** delatte has joined #openstack-nova | 18:48 | |
| dansmith | probably some python library we could import | 18:48 |
| dansmith | smtpdb | 18:48 |
| johnthetubaguy | did anyway do chinese whispers, did that teach us anything? | 18:49 |
| johnthetubaguy | my typing is terrible | 18:49 |
| johnthetubaguy | time for food | 18:49 |
| *** mriedem has quit IRC | 18:49 | |
| alaski | we called it "telephone" but yeah | 18:49 |
| johnthetubaguy | thats a nicer name I guess | 18:49 |
| *** aginwala has joined #openstack-nova | 18:49 | |
| dansmith | heh | 18:49 |
| *** delattec has quit IRC | 18:51 | |
| *** mriedem has joined #openstack-nova | 18:51 | |
| alaski | dansmith: thinking on this SMTP does have nicer guarantees than rabbitmq. this could be a nice april 1st proposal | 18:52 |
| dansmith | alaski: certainly more reliable | 18:53 |
| doffm | Thats just evil. Notifications using NNTP. | 18:53 |
| dansmith | alaski: maybe less performant | 18:53 |
| dims | alaski : dansmith : ouch :) | 18:53 |
| *** mfedosin has joined #openstack-nova | 18:53 | |
| dansmith | dims: disagree? | 18:53 |
| dims | dansmith : nope :) | 18:53 |
| dansmith | :D | 18:53 |
| *** angdraug has joined #openstack-nova | 18:53 | |
| *** rlrossit has joined #openstack-nova | 18:54 | |
| alaski | doffm: heh | 18:55 |
| *** krtaylor has quit IRC | 18:57 | |
| *** thorst_ is now known as thorst | 18:57 | |
| *** matrohon has quit IRC | 18:57 | |
| *** vilobhmm11 has joined #openstack-nova | 18:58 | |
| openstackgerrit | Jay Pipes proposed openstack/nova-specs: resource-providers: Migrate compute node inventory https://review.openstack.org/260048 | 18:58 |
| *** ctrath has quit IRC | 18:58 | |
| jaypipes | johnthetubaguy, dansmith, bauzas: ^^ | 18:58 |
| cdent | alaski: I always thought it might be fun to implement ODBC over IMAP | 18:59 |
| *** ccarmack has joined #openstack-nova | 18:59 | |
| *** ctrath has joined #openstack-nova | 19:01 | |
| dansmith | all protocols can be implemented over SMTP and IMAP | 19:01 |
| doffm | alaski: What will we do with the API elements of the resource pools? We will have to create a table in the API database that maps Resource pool ids to cells? | 19:01 |
| alaski | cdent: heh, that would be interesting. I've always found IMAP awkward though | 19:02 |
| *** raildo is now known as raildo-afk | 19:02 | |
| cdent | exactly | 19:02 |
| cdent | dansmith: this is true | 19:02 |
| rlrossit | alright dansmith, ya dun got me... how in the world do fields magically get lazy-loaded? I thought this always happened in obj_load_attr(), but that isn't overridden everywhere | 19:03 |
| *** ociuhandu has joined #openstack-nova | 19:03 | |
| alaski | doffm: as far as I can tell there won't be API elements for resource pools | 19:04 |
| alaski | the only thing that should need to be aware of them is the scheduler, and it will be reading from all of the cell dbs anyways | 19:04 |
| alaski | actually, I should double check how these are going to be managed, API vs nova-manage | 19:05 |
| *** tdurakov_ has joined #openstack-nova | 19:05 | |
| dansmith | rlrossit: right, very few objects actually do lazy loading | 19:05 |
| *** fawadkhaliq has quit IRC | 19:06 | |
| dansmith | rlrossit: so they don't implement that.. thus errors that would trigger a lazy load get found immediately | 19:06 |
| alaski | jaypipes: cdent what was the plan on setting up and managing resource pools, API or nova-manage commands | 19:06 |
| rlrossit | dansmith: so shouldn't something like PersistentObject implement that at the bottom? | 19:06 |
| doffm | * Trying to find the spec link. | 19:06 |
| rlrossit | or is it because there's so gdang many mixins it'll muck things up? | 19:06 |
| dansmith | rlrossit: implement it how? | 19:07 |
| *** eil397 has joined #openstack-nova | 19:07 | |
| rlrossit | lazy loading | 19:07 |
| dansmith | rlrossit: most objects don't _need_ lazy loading.. we certainly wouldn't want to do it generically | 19:07 |
| doffm | https://review.openstack.org/#/c/253187/8/specs/mitaka/approved/generic-resource-pools.rst | 19:07 |
| rlrossit | dansmith: but we do want it for DB-backed objects right? (which is what I imaging persistent objects are) | 19:07 |
| dansmith | rlrossit: no | 19:07 |
| doffm | Mentions new admin-only API methods to create and manage new resource pools. | 19:08 |
| dansmith | rlrossit: there's nothing to load on most objects -- everything is filled out all the time on a query | 19:08 |
| rlrossit | and here I thought I started understanding something... welp back to square one | 19:08 |
| openstackgerrit | Corey Wright proposed openstack/nova: Avoid lazy-loading flavor during usage audit https://review.openstack.org/274984 | 19:08 |
| alaski | doffm: yeah, just found it https://review.openstack.org/#/c/253187/8/specs/mitaka/approved/generic-resource-pools.rst | 19:08 |
| *** krtaylor has joined #openstack-nova | 19:08 | |
| *** daemontool_ has joined #openstack-nova | 19:10 | |
| alaski | doffm: I'm going to comment on that review that something like a mapping table is going to be needed due to the decision to put the tables in the cells | 19:10 |
| jaypipes | alaski: RESTful API | 19:10 |
| *** daemontool__ has joined #openstack-nova | 19:11 | |
| alaski | jaypipes: okay. There's going to need to be a way to indicate which cell a pool should be created in | 19:11 |
| doffm | alaski: I'm worried about the API and also what bauzas mentioned about links to aggregates. (Which are in the API db). I guess we can discuss possible aggregate solutions tomorrow in the meeting. | 19:11 |
| doffm | alaski: Yeah, it will need to take a cell id in the create api. | 19:12 |
| *** mylu has joined #openstack-nova | 19:12 | |
| *** daemontool has quit IRC | 19:12 | |
| alaski | doffm: yeah, this would all be good to discuss tomorrow when we have more interested parties | 19:13 |
| openstackgerrit | Chuck Carmack proposed openstack/nova: config options: centralize 'spice' options https://review.openstack.org/269910 | 19:13 |
| openstackgerrit | Chuck Carmack proposed openstack/nova: config options: centralize 'spice' options https://review.openstack.org/264271 | 19:13 |
| *** daemontool_ has quit IRC | 19:14 | |
| *** alex_xu has quit IRC | 19:15 | |
| *** penick has joined #openstack-nova | 19:17 | |
| *** mylu has quit IRC | 19:17 | |
| *** dims_ has joined #openstack-nova | 19:17 | |
| *** jcook has quit IRC | 19:17 | |
| *** cfcp1 has joined #openstack-nova | 19:17 | |
| *** dims has quit IRC | 19:17 | |
| *** alex_xu has joined #openstack-nova | 19:18 | |
| *** mylu has joined #openstack-nova | 19:20 | |
| *** jcook has joined #openstack-nova | 19:20 | |
| *** Bjoern has quit IRC | 19:21 | |
| *** mylu has quit IRC | 19:24 | |
| *** cfcp1 has quit IRC | 19:27 | |
| *** dims_ has quit IRC | 19:27 | |
| *** Sukhdev has joined #openstack-nova | 19:27 | |
| *** dedery has quit IRC | 19:27 | |
| *** dedery has joined #openstack-nova | 19:28 | |
| *** su_zhang has quit IRC | 19:29 | |
| *** dims has joined #openstack-nova | 19:30 | |
| *** eil397 has quit IRC | 19:30 | |
| *** eil397 has joined #openstack-nova | 19:31 | |
| *** su_zhang has joined #openstack-nova | 19:32 | |
| *** eil397 has quit IRC | 19:33 | |
| *** dedery has quit IRC | 19:35 | |
| *** dedery has joined #openstack-nova | 19:36 | |
| *** eil397 has joined #openstack-nova | 19:36 | |
| *** dedery has quit IRC | 19:38 | |
| *** dedery has joined #openstack-nova | 19:39 | |
| *** can8dnSix has joined #openstack-nova | 19:41 | |
| *** dims has quit IRC | 19:41 | |
| *** dims has joined #openstack-nova | 19:42 | |
| sdague | dansmith: is that eventlet related? | 19:43 |
| dansmith | sdague: what? | 19:44 |
| *** rk4n has joined #openstack-nova | 19:44 | |
| *** dedery has quit IRC | 19:44 | |
| sdague | the test fail you pointed me at | 19:44 |
| *** tdurakov_ has quit IRC | 19:45 | |
| *** mriedem has quit IRC | 19:45 | |
| dansmith | sdague: oh, I dunno | 19:46 |
| *** dedery has joined #openstack-nova | 19:46 | |
| sdague | eventlet release days, always exciting | 19:46 |
| dansmith | sdague: cdent was asking if it was known | 19:46 |
| dansmith | could be | 19:46 |
| *** dims_ has joined #openstack-nova | 19:46 | |
| sdague | 0.18.2 is out it appears | 19:46 |
| sdague | things like that I always assume are eventlet. Let me see if I can reproduce/fix | 19:46 |
| *** mylu has joined #openstack-nova | 19:46 | |
| sdague | was away at lunch so only just saw the carnage | 19:46 |
| cdent | sdague: I was assuming eventlet too | 19:47 |
| cdent | but haven't had a chance to check yet | 19:47 |
| *** mriedem has joined #openstack-nova | 19:47 | |
| *** vishwanathj has joined #openstack-nova | 19:47 | |
| * cdent checks now | 19:48 | |
| *** su_zhang has quit IRC | 19:48 | |
| *** su_zhang has joined #openstack-nova | 19:48 | |
| cdent | sdague: yup all good with "Successfully installed eventlet-0.17.4" | 19:48 |
| *** dims has quit IRC | 19:49 | |
| sdague | yeh, 0.18.2 | 19:49 |
| *** e0ne has quit IRC | 19:49 | |
| *** su_zhang has quit IRC | 19:49 | |
| *** aginwala has quit IRC | 19:49 | |
| * cdent sighs | 19:49 | |
| vilobhmm11 | sahid : ping | 19:49 |
| *** su_zhang has joined #openstack-nova | 19:50 | |
| sdague | ok, so "hold please" on patch approvals until we have this sorted | 19:50 |
| vilobhmm11 | bauzas : ping | 19:50 |
| *** su_zhang has quit IRC | 19:51 | |
| *** dedery has quit IRC | 19:51 | |
| *** su_zhang has joined #openstack-nova | 19:52 | |
| cdent | sdague: anything I can do to assist? | 19:52 |
| *** dedery has joined #openstack-nova | 19:52 | |
| sdague | cdent: maybe send an email to the list letting people know nova patches are currently blocked | 19:52 |
| sdague | and not to approve things until we sort it | 19:52 |
| cdent | sdague: roger | 19:52 |
| *** mylu has quit IRC | 19:53 | |
| *** aginwala has joined #openstack-nova | 19:55 | |
| *** achanda has quit IRC | 19:55 | |
| edleafe | alaski: I'm starting to work on objectifying the build_request_spec method in scheduler utils. Should that go under this BP: https://blueprints.launchpad.net/nova/+spec/add-buildrequest-obj | 19:55 |
| edleafe | alaski: or should it be its own thing? | 19:55 |
| sdague | god this is goofy, it's not returning a bit of the wsgi response | 19:56 |
| *** dedery has quit IRC | 19:56 | |
| *** su_zhang has quit IRC | 19:57 | |
| cdent | sdague: I reckon that test as _unit_ test is goofy | 19:57 |
| sdague | yeh, well it all used to be unit | 19:57 |
| *** dims_ has quit IRC | 19:57 | |
| *** dedery has joined #openstack-nova | 19:57 | |
| *** can8dnSix has quit IRC | 19:59 | |
| *** ndipanov has quit IRC | 20:02 | |
| mriedem | wow yeah logging is borked | 20:04 |
| *** dedery has quit IRC | 20:04 | |
| *** jgriffith has quit IRC | 20:05 | |
| *** dedery has joined #openstack-nova | 20:05 | |
| mriedem | cdent: is there a bug tracking this yet? | 20:05 |
| cdent | mriedem: not yet, I had to step away for dinner | 20:05 |
| cdent | and also before I ever got to that stage I stumbled trying to make a query for logstash and utterly failed | 20:06 |
| cdent | mostly because I have no idea what I'm doing | 20:06 |
| *** eil397 has quit IRC | 20:06 | |
| *** jgriffith has joined #openstack-nova | 20:06 | |
| *** jgriffith is now known as Guest55421 | 20:06 | |
| *** ctrath has quit IRC | 20:06 | |
| mriedem | http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22testtools.matchers._impl.MismatchError%3A%20'OK%5C%5C%5C%5Cr%5C%5C%5C%5Cn'%20!%3D%20'PONG'%5C%22%20AND%20tags%3A%5C%22console%5C%22&from=7d | 20:07 |
| mriedem | you have to escape some things | 20:07 |
| *** rcernin has joined #openstack-nova | 20:08 | |
| *** rfolco has quit IRC | 20:08 | |
| cdent | mriedem: I think my error was I was trying to search on the test name | 20:08 |
| mriedem | https://bugs.launchpad.net/nova/+bug/1543766 | 20:09 |
| openstack | Launchpad bug 1543766 in OpenStack Compute (nova) "nova.tests.unit.test_wsgi.TestWSGIServerWithSSL fails with testtools.matchers._impl.MismatchError: 'OK\r\n' != 'PONG' and eventlet 0.18.2" [Undecided,New] | 20:09 |
| *** ctrath has joined #openstack-nova | 20:09 | |
| cdent | which thus has no hits. how is that the failure is a message but the test name is not? | 20:09 |
| mriedem | cdent: well you want to fingerprint the failure not the test name anyway | 20:10 |
| cdent | yeah, I get that in terms of fingerprinting | 20:11 |
| mriedem | it's only 44 hits today | 20:11 |
| cdent | but I just mean in terms of searching | 20:11 |
| mriedem | so it doesn't appear to be crippling everything | 20:11 |
| mriedem | and that's 2 hits per test run where it fails | 20:11 |
| mriedem | so it's really like 22 job failures | 20:11 |
| cdent | mriedem: I think the low number of hits is because the gate is slow today and eventlet was released in the afternoon, today | 20:12 |
| mriedem | yeah maybe, but you were talking about this pong failure hours ago | 20:13 |
| mriedem | surely we've run more than 22 py27 nova jobs in the check and gate queues in the last few hours | 20:13 |
| cdent | well, it is eventlet | 20:13 |
| cdent | which means it is unpredictable | 20:13 |
| cdent | anyway: what's your point? | 20:14 |
| cdent | :) | 20:14 |
| mriedem | my point is the sky doesn't appear to be falling | 20:14 |
| mriedem | and why wouldn't it fail on the py34 job? | 20:14 |
| mriedem | it's just weir | 20:14 |
| mriedem | *weird | 20:14 |
| cdent | the sky is never falling, it's just the gate | 20:14 |
| cdent | but we _act_ like it is falling because that's what we do | 20:14 |
| cdent | if I had my druthers we'd just let the gate fail and willingly let things come to an eventual stop so that we are forced to care | 20:16 |
| cdent | but hey, it's not my world | 20:16 |
| melwitt | mriedem: this says gates are working on a cached version of 0.18.1 which might be why? https://review.openstack.org/#/c/277912/ | 20:17 |
| melwitt | (from http://lists.openstack.org/pipermail/openstack-dev/2016-February/086161.html) | 20:18 |
| *** dedery has quit IRC | 20:19 | |
| *** krtaylor has quit IRC | 20:19 | |
| *** dedery has joined #openstack-nova | 20:19 | |
| mriedem | oh fun | 20:19 |
| sdague | mriedem: it is crippling everything | 20:19 |
| sdague | mriedem: we might skip it there | 20:20 |
| mriedem | yeah i see it locally too on a fresh venv | 20:20 |
| sdague | so, it's 2 tests with a small wsgi app | 20:20 |
| sdague | we're just never getting content from it | 20:20 |
| sdague | only the headers | 20:20 |
| cdent | probably need to throw in an arbitrary yield and it'll be fine/ | 20:20 |
| cdent | ? | 20:20 |
| mriedem | has anyone checked the eventlet 0.18.2 changelog? | 20:20 |
| * cdent tries it | 20:20 | |
| mriedem | * wsgi: Fix data loss on partial writes (socket.send); Thanks to Jakub Stasiak | 20:21 |
| mriedem | https://github.com/eventlet/eventlet/commit/b7380fdc70fde26777f4f141da3f01570e870cac | 20:22 |
| cdent | yeah, that fixes it | 20:22 |
| cdent | (a sleep before the assert) | 20:22 |
| mriedem | so they fix a thing for py3 and break py2 | 20:22 |
| cdent | sdague, mriedem: I'm actually surprised this test ever worked | 20:22 |
| sdague | cdent: it is a weird test | 20:23 |
| cdent | (without the yield | 20:23 |
| cdent | I can commit that change if that seems wise? | 20:23 |
| cdent | it's used elsewhere in the same file (the sleep trick) | 20:23 |
| *** Marga_ has quit IRC | 20:23 | |
| sdague | cdent: yeh, maybe, though what is the result if we just use requests in the client | 20:23 |
| sdague | I find it weird that we're using an eventlet specific connect | 20:24 |
| *** Guest55421 is now known as jgriffith | 20:24 | |
| cdent | yes, that is a bit weird | 20:24 |
| cdent | I see three options: rewrite the test to modern norms, work around this immediate problem, get rid of this bogus test | 20:25 |
| sdague | it's a kind of weird test, I'd almost lean on just delete. But using requests seems better | 20:26 |
| *** Sukhdev has quit IRC | 20:27 | |
| sdague | yeh, it works with requests I think fine, there is something about the explicit client socket bit | 20:28 |
| sdague | which we completely don't care about from nova | 20:28 |
| *** su_zhang has joined #openstack-nova | 20:28 | |
| cdent | and requests is already imported, so yeah, makes sense | 20:28 |
| cdent | you wanna, or shall I? | 20:29 |
| sdague | ok, just need to sort out requests syntax for the ssl bundle | 20:29 |
| sdague | cdent: if you want to, that would be cool | 20:29 |
| cdent | I can do it | 20:29 |
| sdague | I think the only thing that's really needed is doing the certs right with requests, I hit the non ssl port for ease locally to figure out if it would work | 20:29 |
| sdague | cdent: coolio, poke me as soon as it's up | 20:29 |
| cdent | see as you're supposed to be in a meeting | 20:29 |
| *** mylu has joined #openstack-nova | 20:29 | |
| *** raildo-afk is now known as raildo | 20:29 | |
| *** Sukhdev has joined #openstack-nova | 20:30 | |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: objects: Rename PciDevice _migrate_parent_addr method https://review.openstack.org/278076 | 20:30 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: db: Add a basic framework for generating online migration scripts https://review.openstack.org/278077 | 20:30 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: nova-manage: Add hooks for generating data-migration scripts https://review.openstack.org/278078 | 20:30 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: nova-manage: Declare a PciDevice online migration script https://review.openstack.org/278079 | 20:30 |
| sdague | yeh, I've got one eye on it ... but it is all networking ancronyms | 20:30 |
| *** ndipanov has joined #openstack-nova | 20:31 | |
| *** krtaylor has joined #openstack-nova | 20:31 | |
| *** su_zhang has quit IRC | 20:32 | |
| ndipanov | dansmith, https://review.openstack.org/#/c/278076/ - one patch is missing tests | 20:32 |
| *** tcammann has joined #openstack-nova | 20:33 | |
| ndipanov | but would like to hear thoughts early | 20:33 |
| *** shyama has quit IRC | 20:34 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: libvirt: check for interface when detach_interface fails https://review.openstack.org/270891 | 20:36 |
| openstackgerrit | Matt Riedemann proposed openstack/nova: libvirt: implement LibvirtConfigGuestInterface.parse_dom https://review.openstack.org/270981 | 20:36 |
| *** salv-orl_ has quit IRC | 20:36 | |
| *** tcammann has quit IRC | 20:39 | |
| *** mylu has quit IRC | 20:40 | |
| *** mylu has joined #openstack-nova | 20:40 | |
| *** aginwala has quit IRC | 20:40 | |
| *** rk4n has quit IRC | 20:43 | |
| *** aginwala has joined #openstack-nova | 20:43 | |
| *** shawnmaten has quit IRC | 20:43 | |
| *** omnipresent has joined #openstack-nova | 20:44 | |
| *** su_zhang has joined #openstack-nova | 20:44 | |
| *** Sukhdev has quit IRC | 20:45 | |
| *** zenoway has quit IRC | 20:45 | |
| *** zenoway_ has joined #openstack-nova | 20:45 | |
| *** lxsli has joined #openstack-nova | 20:47 | |
| openstackgerrit | Swaminathan Vasudevan proposed openstack/nova: Implement setup_networks_on_host for Neutron to support Migration https://review.openstack.org/275073 | 20:47 |
| *** mylu has quit IRC | 20:48 | |
| *** omnipresent has quit IRC | 20:48 | |
| *** tcammann has joined #openstack-nova | 20:48 | |
| bauzas | so, we have a gate issue with eventlet ? | 20:49 |
| *** knikolla has quit IRC | 20:49 | |
| cdent | bauzas: yeah, fix on its way | 20:49 |
| *** deva_ has joined #openstack-nova | 20:50 | |
| *** knikolla has joined #openstack-nova | 20:50 | |
| smcginnis | bauzas: Fix just went through. My CI can finally stack again, so looks like it worked. | 20:52 |
| *** salv-orlando has joined #openstack-nova | 20:53 | |
| *** krtaylor has quit IRC | 20:53 | |
| *** tcammann has quit IRC | 20:53 | |
| tjones_zzzz | ok to recheck now? | 20:53 |
| *** lxsli has quit IRC | 20:53 | |
| *** deva_ has quit IRC | 20:54 | |
| *** su_zhang has quit IRC | 20:57 | |
| *** cfcp1 has joined #openstack-nova | 20:57 | |
| *** aginwala has quit IRC | 20:57 | |
| *** cfcp1 has quit IRC | 20:58 | |
| cdent | smcginnis: which fix do you mean? | 20:58 |
| *** Marga_ has joined #openstack-nova | 20:59 | |
| mmedvede | cdent: https://review.openstack.org/277912 | 20:59 |
| smcginnis | cdent: What mmedvede said ^^ | 21:00 |
| sdague | that's a differen thing | 21:00 |
| cdent | mmedvede: that's a different problem. 0.18.2 causes nova unit test to fail | 21:00 |
| mmedvede | oh sorry | 21:00 |
| cdent | but fixing the test is relatively easy, so that's on the way | 21:00 |
| smcginnis | Ah, didn't catch that. | 21:00 |
| cdent | tjones_zzzz: not yet | 21:00 |
| alaski | edleafe: it should be its own thing | 21:01 |
| *** eil397 has joined #openstack-nova | 21:01 | |
| edleafe | alaski: thx | 21:01 |
| *** nikhil_k is now known as nikhil_ | 21:01 | |
| *** daemontool has joined #openstack-nova | 21:02 | |
| *** dedery has quit IRC | 21:02 | |
| *** nikhil_ is now known as nikhil_k | 21:02 | |
| *** daemontool__ has quit IRC | 21:02 | |
| *** dims has joined #openstack-nova | 21:03 | |
| *** krotscheck is now known as krotscheck_dcm | 21:03 | |
| *** e0ne has joined #openstack-nova | 21:03 | |
| *** knikolla has quit IRC | 21:03 | |
| sdague | cdent: how's it coming? | 21:05 |
| *** ctrath has quit IRC | 21:05 | |
| cdent | sdague: done, just running tests local for completeness | 21:05 |
| sdague | ok, cool | 21:05 |
| *** ctrath has joined #openstack-nova | 21:06 | |
| *** krtaylor has joined #openstack-nova | 21:07 | |
| *** eil397 has quit IRC | 21:08 | |
| sdague | you can always push early, we'll need test results upstream before we can pop it to top of gate and get rid of the carnage | 21:10 |
| *** smatzek has quit IRC | 21:13 | |
| dims | cdent : pushing a eventlet review? (since g-r has merged) | 21:13 |
| cdent | wait for it... | 21:15 |
| openstackgerrit | Chris Dent proposed openstack/nova: Replace eventlet-based raw socket client with requests https://review.openstack.org/278089 | 21:15 |
| cdent | wait for it... | 21:15 |
| cdent | there it is | 21:15 |
| cdent | sdague, dims, mriedem ^ | 21:16 |
| mriedem | cool, looking | 21:17 |
| *** lpetrut has quit IRC | 21:18 | |
| sdague | cdent: looks great | 21:20 |
| sdague | and passes locally | 21:20 |
| cdent | cool | 21:20 |
| dims | cdent : ah the OK vs PONG | 21:21 |
| cdent | Did you see that one of the servers wasn't getting shut down in the earlier code? I tried just that to see if it made any difference, it did not | 21:21 |
| dims | +1 | 21:21 |
| cfriesen | Its been a while since I fired up devstack, and now I'm getting this: "Could not determine a suitable URL for the plugin" Anyone know what's going on? | 21:22 |
| mriedem | cdent: is eventlet.wsgi import still used | 21:23 |
| mriedem | ? | 21:23 |
| cdent | mriedem: yes | 21:23 |
| cdent | for a constant | 21:23 |
| cfriesen | more specifically, I'm getting this: http://paste.openstack.org/show/486489/ | 21:23 |
| cdent | " eventlet.wsgi.MAX_HEADER_LINE" | 21:23 |
| mriedem | ah yeah | 21:24 |
| *** browne has quit IRC | 21:24 | |
| *** daemontool_ has joined #openstack-nova | 21:24 | |
| mriedem | +W | 21:24 |
| mriedem | we'll need that in liberty and kilo too, i can track those | 21:25 |
| cdent | thank you mriedem | 21:25 |
| *** erlon has quit IRC | 21:26 | |
| *** daemontool has quit IRC | 21:27 | |
| cfriesen | vgs | 21:27 |
| openstackgerrit | Mark Doffman proposed openstack/nova: WIP: Map instances to cell0 on failed schedule https://review.openstack.org/267827 | 21:27 |
| cfriesen | bah, wrong window | 21:27 |
| *** matrohon has joined #openstack-nova | 21:28 | |
| *** rk4n has joined #openstack-nova | 21:28 | |
| openstackgerrit | Drew Thorstensen proposed openstack/nova: Update the CONF import path for VNC https://review.openstack.org/277392 | 21:29 |
| *** jgriffith is now known as jgriffith_away | 21:31 | |
| openstackgerrit | Pushkar Umaranikar proposed openstack/nova: Modify VM's updated_at field on volume actions https://review.openstack.org/247176 | 21:32 |
| *** achanda has joined #openstack-nova | 21:33 | |
| *** browne has joined #openstack-nova | 21:34 | |
| *** baoli has joined #openstack-nova | 21:38 | |
| *** burt has quit IRC | 21:42 | |
| *** davideagnello has quit IRC | 21:42 | |
| mriedem | wow http://logs.openstack.org/12/276812/2/check/gate-manila-tempest-dsvm-neutron/1b42ac2/logs/screen-n-cpu.txt.gz?level=TRACE | 21:42 |
| *** davideagnello has joined #openstack-nova | 21:43 | |
| *** achanda has quit IRC | 21:43 | |
| *** nikhil_k is now known as nikhil | 21:45 | |
| mriedem | https://github.com/openstack/nova/blob/master/nova/virt/osinfo.py#L95 | 21:45 |
| mriedem | cripes | 21:45 |
| mriedem | vladikr: ^ | 21:45 |
| mtreinish | mriedem: yeah, I was gonna complain about that at some point soon. The n-cpu logs are full of that now | 21:47 |
| mriedem | you need to complain sooner | 21:47 |
| *** achanda has joined #openstack-nova | 21:48 | |
| mriedem | i'd like to complain about your lack of complaining | 21:48 |
| *** Marga_ has quit IRC | 21:49 | |
| mtreinish | I've been home for only ~40 hrs and spent a good 20+ of those sleeping :) | 21:49 |
| *** raildo is now known as raildo-afk | 21:49 | |
| cdent | mtreinish: I'm beginning to doubt your commitment to openstack motion | 21:50 |
| *** Marga_ has joined #openstack-nova | 21:50 | |
| mriedem | yeah c'mon | 21:50 |
| mriedem | https://bugs.launchpad.net/nova/+bug/1543800 | 21:50 |
| openstack | Launchpad bug 1543800 in OpenStack Compute (nova) "Error "Cannot find OS information - Reason: (Cannot load Libosinfo: (No module named gi.repository.Libosinfo))" - spamming n-cpu logs" [Undecided,New] | 21:50 |
| *** nkrinner has quit IRC | 21:50 | |
| cdent | damn, I got the quote wrong | 21:50 |
| cdent | close though | 21:51 |
| mtreinish | cdent: haha | 21:51 |
| *** Bjoern has joined #openstack-nova | 21:51 | |
| mriedem | https://github.com/openstack/nova/blob/master/nova/virt/osinfo.py#L24-L28 | 21:52 |
| doffm | I don't know what the problem with GObject is. Or GObject Introspection. | 21:54 |
| mriedem | so we could install libosinfo in devstack, | 21:54 |
| mriedem | doffm: this is the problem https://github.com/openstack/nova/blob/master/nova/virt/osinfo.py#L44 | 21:54 |
| mriedem | the package isn't installed | 21:54 |
| mriedem | we'd have to add libosinfo to the list of native distro packages that devstack installs | 21:55 |
| mriedem | since there isn't a pip for this | 21:55 |
| mriedem | but, given this code is dependent on a native package that may not be installed, we shouldn't probably be logging an error every time we hit it | 21:55 |
| *** su_zhang has joined #openstack-nova | 21:55 | |
| mriedem | we could log a warning, once | 21:55 |
| melwitt | the package used to always be there? | 21:56 |
| *** claudiub has quit IRC | 21:56 | |
| *** smatzek has joined #openstack-nova | 21:56 | |
| *** Marga_ has quit IRC | 21:57 | |
| mriedem | melwitt: i don't think so | 21:57 |
| *** Marga_ has joined #openstack-nova | 21:57 | |
| mriedem | melwitt: http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/libvirt-hardware-policy-from-libosinfo.html#other-deployer-impact | 21:57 |
| mriedem | it's supposed to still work if the package isn't there | 21:57 |
| mriedem | and it must, because otherwise things would be failing | 21:58 |
| mriedem | or could be we just don't test the image props in the gate used here in such a way that we'd know if we fail | 22:00 |
| melwitt | hm | 22:02 |
| *** jwcroppe has joined #openstack-nova | 22:02 | |
| *** Sukhdev has joined #openstack-nova | 22:03 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Update ComputeNode values with disk allocation ratios in the RT https://review.openstack.org/277953 | 22:03 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Add disk_allocation_ratio to ComputeNode https://review.openstack.org/277934 | 22:03 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Update HostManager and DiskFilter to use ComputeNode disk ratio https://review.openstack.org/278112 | 22:03 |
| *** matrohon has quit IRC | 22:03 | |
| *** mrkz has quit IRC | 22:04 | |
| *** yassine has quit IRC | 22:07 | |
| *** mylu has joined #openstack-nova | 22:08 | |
| *** mylu has quit IRC | 22:09 | |
| *** smatzek has quit IRC | 22:11 | |
| *** sinese_ has quit IRC | 22:12 | |
| *** burgerk_ has joined #openstack-nova | 22:13 | |
| *** whenry is now known as whenry_away | 22:13 | |
| *** thorst has quit IRC | 22:14 | |
| *** ccarmack has quit IRC | 22:15 | |
| *** thorst has joined #openstack-nova | 22:15 | |
| *** ccarmack has joined #openstack-nova | 22:15 | |
| doffm | mriedem: Its super confusing trying to work out in which situations the osinfo info is going to be neccessary. | 22:16 |
| *** jichen has quit IRC | 22:16 | |
| *** burgerk has quit IRC | 22:16 | |
| doffm | We could just refactor this to pump out one warning on import failure as you suggested. | 22:16 |
| doffm | Assume its working. | 22:16 |
| *** thorst_ has joined #openstack-nova | 22:16 | |
| *** burgerk_ has quit IRC | 22:17 | |
| *** achanda has quit IRC | 22:19 | |
| *** mylu has joined #openstack-nova | 22:19 | |
| *** thorst has quit IRC | 22:19 | |
| *** achanda has joined #openstack-nova | 22:21 | |
| *** thorst_ has quit IRC | 22:21 | |
| *** dedery has joined #openstack-nova | 22:22 | |
| *** ducttape_ has quit IRC | 22:31 | |
| *** aginwala has joined #openstack-nova | 22:33 | |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Remove version hardcoding for 'detail', 'download' and 'delete' methods https://review.openstack.org/238309 | 22:36 |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Remove version hardcoding for 'create' and 'update' methods https://review.openstack.org/259097 | 22:38 |
| *** thorst has joined #openstack-nova | 22:39 | |
| *** dims_ has joined #openstack-nova | 22:39 | |
| mriedem | doffm: i'm working on a change | 22:39 |
| mriedem | trying to write the unit test first | 22:39 |
| mriedem | which basically falls down b/c we don't have libosinfo... | 22:39 |
| *** salv-orl_ has joined #openstack-nova | 22:40 | |
| *** dims has quit IRC | 22:40 | |
| *** gibi_ has quit IRC | 22:41 | |
| *** ctrath has quit IRC | 22:41 | |
| *** gibi has joined #openstack-nova | 22:41 | |
| *** jwcroppe has quit IRC | 22:42 | |
| *** rlrossit has quit IRC | 22:42 | |
| lbeliveau | ndipanov: Are you around ? | 22:42 |
| *** thorst has quit IRC | 22:43 | |
| *** w_verdugo has quit IRC | 22:43 | |
| vladikr | mriedem, sorry I missed that.. here is the fix, hope soe | 22:43 |
| vladikr | so | 22:43 |
| vladikr | https://review.openstack.org/#/c/277565/ | 22:43 |
| *** salv-orlando has quit IRC | 22:43 | |
| *** w_verdugo has joined #openstack-nova | 22:45 | |
| *** achanda has quit IRC | 22:45 | |
| lbeliveau | ndipanov: I replied to one of your comment on https://review.openstack.org/#/c/242573. Can you please have a look when you have a chance ? Need your opinion before make any changes. | 22:46 |
| *** farid has quit IRC | 22:46 | |
| *** Sukhdev has quit IRC | 22:47 | |
| *** gibi has quit IRC | 22:47 | |
| *** gibi has joined #openstack-nova | 22:48 | |
| *** pradk has quit IRC | 22:49 | |
| *** baoli has quit IRC | 22:50 | |
| *** ctrath has joined #openstack-nova | 22:50 | |
| mriedem | vladikr: one nit in there but +2 | 22:51 |
| mriedem | vladikr: if you wanted to update quick i could +2 again before i leave | 22:51 |
| vladikr | mriedem, sure, looking | 22:51 |
| dedery | hi cores, can you please review https://review.openstack.org/#/c/270686 ? It's a backport of the duplicate NICs issue (failing nova boot) to Liberty (mriedem, it's according to what we've talked about a month or so ago) | 22:52 |
| *** rk4n_ has joined #openstack-nova | 22:52 | |
| cdent | sdague or mriedem: I'm off to bed, so can somebody else followup when https://review.openstack.org/#/c/278089/ merges | 22:54 |
| cdent | it's almost there | 22:54 |
| mriedem | dedery: done | 22:54 |
| *** cdent has quit IRC | 22:55 | |
| *** rk4n has quit IRC | 22:55 | |
| *** baoli has joined #openstack-nova | 22:55 | |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova: virt: osinfo will report once if libosinfo is not loaded https://review.openstack.org/277565 | 22:55 |
| *** mylu has quit IRC | 22:56 | |
| *** tongli has quit IRC | 22:56 | |
| *** Sukhdev has joined #openstack-nova | 22:57 | |
| mriedem | tonyb: fyi https://review.openstack.org/#/q/I00c1c1e93eb7c66c3370434db4e3e6c00087f8aa,n,z | 22:57 |
| *** dimtruck is now known as zz_dimtruck | 22:57 | |
| dedery | mriedem: thank you :) | 22:58 |
| *** rcernin has quit IRC | 22:58 | |
| tonyb | mriedem: Thanks. | 22:58 |
| *** su_zhang has quit IRC | 22:58 | |
| *** su_zhang_ has joined #openstack-nova | 22:58 | |
| *** Sukhdev has quit IRC | 22:58 | |
| tonyb | mriedem: I did see them land in my inbox. | 22:58 |
| *** mdrabe has quit IRC | 22:59 | |
| *** Sukhdev has joined #openstack-nova | 22:59 | |
| *** mfedosin has quit IRC | 22:59 | |
| *** mylu has joined #openstack-nova | 22:59 | |
| *** su_zhang_ has quit IRC | 22:59 | |
| *** su_zhang has joined #openstack-nova | 23:00 | |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova: virt: osinfo will report once if libosinfo is not loaded https://review.openstack.org/277565 | 23:00 |
| *** jaypipes has quit IRC | 23:00 | |
| *** thorst has joined #openstack-nova | 23:00 | |
| vladikr | mriedem, all done :) | 23:00 |
| *** su_zhang has quit IRC | 23:00 | |
| *** su_zhang has joined #openstack-nova | 23:00 | |
| *** edtubill has quit IRC | 23:01 | |
| mriedem | vladikr: thanks | 23:01 |
| *** Sukhdev has quit IRC | 23:01 | |
| *** edtubill has joined #openstack-nova | 23:02 | |
| *** ccarmack has quit IRC | 23:02 | |
| tonyb | mriedem: Hehe that change is in both check and gate *at the same time* :D | 23:02 |
| *** daemontool__ has joined #openstack-nova | 23:02 | |
| mriedem | it's like a double rainbow | 23:03 |
| *** eharney has quit IRC | 23:03 | |
| *** Sukhdev has joined #openstack-nova | 23:04 | |
| *** daemontool_ has quit IRC | 23:04 | |
| *** edtubill has quit IRC | 23:06 | |
| tonyb | :) | 23:07 |
| *** Sukhdev has quit IRC | 23:09 | |
| *** thorst has quit IRC | 23:09 | |
| *** Sukhdev has joined #openstack-nova | 23:09 | |
| *** thorst has joined #openstack-nova | 23:09 | |
| *** xyang1 has quit IRC | 23:09 | |
| *** Sukhdev has quit IRC | 23:10 | |
| *** Sukhdev has joined #openstack-nova | 23:10 | |
| *** Sukhdev has quit IRC | 23:12 | |
| *** Sukhdev has joined #openstack-nova | 23:13 | |
| *** aginwala has quit IRC | 23:14 | |
| *** thorst has quit IRC | 23:14 | |
| *** Jeffrey4l has joined #openstack-nova | 23:15 | |
| *** aginwala has joined #openstack-nova | 23:16 | |
| *** penick has quit IRC | 23:18 | |
| *** gokrokve has joined #openstack-nova | 23:19 | |
| *** daemontool_ has joined #openstack-nova | 23:20 | |
| *** baoli has quit IRC | 23:20 | |
| *** mriedem has quit IRC | 23:22 | |
| *** daemontool__ has quit IRC | 23:22 | |
| *** gokrokve has quit IRC | 23:23 | |
| *** zz_dimtruck is now known as dimtruck | 23:25 | |
| *** daemontool_ has quit IRC | 23:28 | |
| *** daemontool has joined #openstack-nova | 23:29 | |
| *** Sukhdev has quit IRC | 23:30 | |
| *** ssurana1 has joined #openstack-nova | 23:30 | |
| *** Sukhdev has joined #openstack-nova | 23:31 | |
| *** Sukhdev has quit IRC | 23:32 | |
| *** ssurana has quit IRC | 23:33 | |
| *** diana_clarke has quit IRC | 23:36 | |
| *** jwcroppe has joined #openstack-nova | 23:40 | |
| *** mrkz has joined #openstack-nova | 23:40 | |
| *** pushkaru has quit IRC | 23:40 | |
| *** pushkaru has joined #openstack-nova | 23:41 | |
| *** ssurana1 has quit IRC | 23:41 | |
| *** zenoway_ has quit IRC | 23:42 | |
| *** jwcroppe has quit IRC | 23:44 | |
| *** mrkz1 has joined #openstack-nova | 23:44 | |
| *** jwcroppe has joined #openstack-nova | 23:45 | |
| *** pushkaru has quit IRC | 23:45 | |
| *** Bjoern has quit IRC | 23:46 | |
| *** mrkz has quit IRC | 23:47 | |
| *** Sukhdev has joined #openstack-nova | 23:47 | |
| *** dedery has quit IRC | 23:47 | |
| *** wanghao has quit IRC | 23:47 | |
| *** su_zhang has quit IRC | 23:47 | |
| *** wanghao has joined #openstack-nova | 23:48 | |
| *** ndipanov has quit IRC | 23:49 | |
| *** jwcroppe has quit IRC | 23:49 | |
| *** eil397 has joined #openstack-nova | 23:51 | |
| *** Sukhdev has quit IRC | 23:54 | |
| *** mragupat_ has quit IRC | 23:57 | |
| *** su_zhang has joined #openstack-nova | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!