| *** haomaiwa_ has quit IRC | 00:01 | |
| *** haomaiwang has joined #openstack-nova | 00:01 | |
| *** irenab_ has joined #openstack-nova | 00:02 | |
| *** crinkle has quit IRC | 00:03 | |
| *** crinkle_ has joined #openstack-nova | 00:04 | |
| openstackgerrit | Ludovic Beliveau proposed openstack/nova: Update binding:profile for SR-IOV ports https://review.openstack.org/242573 | 00:04 |
|---|---|---|
| *** crinkle_ is now known as crinkle | 00:04 | |
| *** irenab has quit IRC | 00:04 | |
| *** irenab_ is now known as irenab | 00:04 | |
| *** mylu has joined #openstack-nova | 00:05 | |
| *** nelsnels_ has joined #openstack-nova | 00:07 | |
| *** nelsnelson has quit IRC | 00:08 | |
| *** ducttape_ has joined #openstack-nova | 00:09 | |
| *** armax has joined #openstack-nova | 00:10 | |
| *** nekrodesk has quit IRC | 00:11 | |
| *** thumpba has joined #openstack-nova | 00:13 | |
| *** Daisy has joined #openstack-nova | 00:14 | |
| *** hemna is now known as hemnafk | 00:16 | |
| *** kylek3h has quit IRC | 00:16 | |
| *** thumpba has quit IRC | 00:17 | |
| *** sb- is now known as sb | 00:17 | |
| *** Daisy has quit IRC | 00:18 | |
| *** mfedosin has quit IRC | 00:23 | |
| *** mylu has quit IRC | 00:26 | |
| *** kylek3h has joined #openstack-nova | 00:26 | |
| *** xyang1_ has joined #openstack-nova | 00:28 | |
| *** xyang1 has quit IRC | 00:28 | |
| *** xyang1_ is now known as xyang1 | 00:28 | |
| *** nic has quit IRC | 00:29 | |
| *** mylu has joined #openstack-nova | 00:29 | |
| *** ganso has quit IRC | 00:29 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes hex decoding related unit tests https://review.openstack.org/278808 | 00:31 |
| *** claudiub has joined #openstack-nova | 00:31 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes python 3 urllib quote / unquote usage https://review.openstack.org/278809 | 00:31 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes python 3 unit tests https://review.openstack.org/278810 | 00:31 |
| *** nekrodesk has joined #openstack-nova | 00:32 | |
| *** mingyu_ has joined #openstack-nova | 00:34 | |
| *** Dave_____ is now known as Dave | 00:34 | |
| *** mingyu_ has quit IRC | 00:34 | |
| *** Mingyu has joined #openstack-nova | 00:36 | |
| *** kylek3h has left #openstack-nova | 00:39 | |
| *** mylu has quit IRC | 00:40 | |
| *** mylu has joined #openstack-nova | 00:40 | |
| openstackgerrit | Corey Wright proposed openstack/nova: Implement images list default and maximum limit https://review.openstack.org/278815 | 00:42 |
| *** aginwala has quit IRC | 00:42 | |
| *** Swami has quit IRC | 00:44 | |
| *** aginwala has joined #openstack-nova | 00:45 | |
| *** mylu has quit IRC | 00:46 | |
| *** aginwala has quit IRC | 00:46 | |
| *** aginwala has joined #openstack-nova | 00:46 | |
| *** mylu has joined #openstack-nova | 00:47 | |
| *** kylek3h has joined #openstack-nova | 00:48 | |
| *** thorst_ has joined #openstack-nova | 00:49 | |
| *** mylu has quit IRC | 00:49 | |
| openstackgerrit | Corey Wright proposed openstack/nova: Implement images list default and maximum limit https://review.openstack.org/278815 | 00:49 |
| *** mylu has joined #openstack-nova | 00:49 | |
| *** Sukhdev has quit IRC | 00:51 | |
| *** kylek3h has quit IRC | 00:51 | |
| *** mylu has quit IRC | 00:52 | |
| openstackgerrit | Miguel Lavalle proposed openstack/nova: Use instance hostname for Neutron DNS unit tests https://review.openstack.org/277302 | 00:54 |
| *** ducttape_ has quit IRC | 00:55 | |
| *** mylu has joined #openstack-nova | 00:55 | |
| *** Jeffrey4l has joined #openstack-nova | 00:56 | |
| openstackgerrit | Miguel Lavalle proposed openstack/nova: Use instance hostname for Neutron DNS unit tests https://review.openstack.org/277302 | 00:57 |
| *** mylu has quit IRC | 00:59 | |
| *** haomaiwang has quit IRC | 01:01 | |
| *** mylu has joined #openstack-nova | 01:01 | |
| *** haomaiwa_ has joined #openstack-nova | 01:02 | |
| *** mylu has quit IRC | 01:03 | |
| *** eil397 has quit IRC | 01:03 | |
| *** thorst_ has quit IRC | 01:05 | |
| *** thorst_ has joined #openstack-nova | 01:06 | |
| *** mylu has joined #openstack-nova | 01:06 | |
| *** Jeffrey4l has quit IRC | 01:07 | |
| *** mylu has quit IRC | 01:08 | |
| *** mylu has joined #openstack-nova | 01:08 | |
| *** Sukhdev has joined #openstack-nova | 01:08 | |
| *** aginwala has quit IRC | 01:13 | |
| *** mylu has quit IRC | 01:13 | |
| *** thorst_ has quit IRC | 01:14 | |
| *** mylu has joined #openstack-nova | 01:15 | |
| *** vilobhmm11 has quit IRC | 01:17 | |
| *** harlowja has quit IRC | 01:18 | |
| *** mylu has quit IRC | 01:18 | |
| *** Jeffrey4l has joined #openstack-nova | 01:20 | |
| *** aginwala has joined #openstack-nova | 01:20 | |
| *** angdraug has quit IRC | 01:20 | |
| *** achanda has quit IRC | 01:21 | |
| *** achanda has joined #openstack-nova | 01:33 | |
| *** jasondotstar has quit IRC | 01:35 | |
| *** egarbade- has quit IRC | 01:35 | |
| *** raginbajin has quit IRC | 01:35 | |
| *** aginwala has quit IRC | 01:35 | |
| *** eliqiao_ has quit IRC | 01:36 | |
| *** egarbade- has joined #openstack-nova | 01:36 | |
| *** eliqiao_ has joined #openstack-nova | 01:36 | |
| *** dims_ has joined #openstack-nova | 01:37 | |
| *** jasondotstar has joined #openstack-nova | 01:37 | |
| *** ivase has joined #openstack-nova | 01:37 | |
| *** raginbajin has joined #openstack-nova | 01:37 | |
| *** dims has quit IRC | 01:38 | |
| *** dims_ has quit IRC | 01:41 | |
| *** apoorvad has quit IRC | 01:41 | |
| *** ivase has quit IRC | 01:41 | |
| *** zz_dimtruck is now known as dimtruck | 01:42 | |
| *** alex_xu has quit IRC | 01:43 | |
| *** ssurana has quit IRC | 01:44 | |
| *** alex_xu has joined #openstack-nova | 01:45 | |
| *** dims has joined #openstack-nova | 01:49 | |
| *** mylu has joined #openstack-nova | 01:51 | |
| *** baoli has joined #openstack-nova | 01:53 | |
| *** mylu has quit IRC | 01:54 | |
| *** ducttape_ has joined #openstack-nova | 01:59 | |
| *** eliqiao_ has quit IRC | 02:00 | |
| *** eliqiao_ has joined #openstack-nova | 02:00 | |
| *** baoli has quit IRC | 02:00 | |
| *** Daisy has joined #openstack-nova | 02:00 | |
| *** haomaiwa_ has quit IRC | 02:01 | |
| *** haomaiwang has joined #openstack-nova | 02:03 | |
| *** Daisy has quit IRC | 02:03 | |
| *** edmondsw has quit IRC | 02:03 | |
| *** mtanino has quit IRC | 02:04 | |
| *** shz has joined #openstack-nova | 02:06 | |
| *** rk4n_ has joined #openstack-nova | 02:06 | |
| *** Sukhdev_ has joined #openstack-nova | 02:06 | |
| *** browne has quit IRC | 02:07 | |
| *** mragupat has joined #openstack-nova | 02:08 | |
| *** mragupat has quit IRC | 02:08 | |
| *** alex_xu has quit IRC | 02:08 | |
| *** Jeffrey4l has quit IRC | 02:08 | |
| *** mragupat has joined #openstack-nova | 02:08 | |
| *** alex_xu has joined #openstack-nova | 02:10 | |
| *** daemontool has quit IRC | 02:11 | |
| *** Jeffrey4l has joined #openstack-nova | 02:11 | |
| *** thorst_ has joined #openstack-nova | 02:12 | |
| *** unicell has quit IRC | 02:14 | |
| *** ByPasS_ has joined #openstack-nova | 02:14 | |
| *** Sukhdev has quit IRC | 02:15 | |
| *** hanzhang has quit IRC | 02:15 | |
| *** rk4n has quit IRC | 02:15 | |
| *** hamzy has quit IRC | 02:15 | |
| *** openstackgerrit has quit IRC | 02:15 | |
| *** w_verdugo has quit IRC | 02:15 | |
| *** n0ano has quit IRC | 02:15 | |
| *** ByPasS has quit IRC | 02:15 | |
| *** ByPasS_ is now known as ByPasS | 02:15 | |
| *** penick has quit IRC | 02:16 | |
| *** baoli has joined #openstack-nova | 02:17 | |
| *** thorst_ has quit IRC | 02:19 | |
| *** hamzy has joined #openstack-nova | 02:21 | |
| *** baoli has quit IRC | 02:23 | |
| *** daemontool has joined #openstack-nova | 02:23 | |
| *** openstackgerrit has joined #openstack-nova | 02:24 | |
| *** mragupat has quit IRC | 02:24 | |
| *** w_verdugo has joined #openstack-nova | 02:24 | |
| *** baoli has joined #openstack-nova | 02:24 | |
| *** Jeffrey4l has quit IRC | 02:25 | |
| *** ducttape_ has quit IRC | 02:26 | |
| *** ducttape_ has joined #openstack-nova | 02:27 | |
| *** xyang1 has quit IRC | 02:29 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes python 3 unit tests https://review.openstack.org/278810 | 02:29 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: libvirt: fixes python 3 related unit tests https://review.openstack.org/278834 | 02:29 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: numa: Fixes NUMA topology related unit tests https://review.openstack.org/278835 | 02:29 |
| *** baoli has quit IRC | 02:29 | |
| *** nekrodesk has quit IRC | 02:31 | |
| *** vilobhmm11 has joined #openstack-nova | 02:34 | |
| *** ducttape_ has quit IRC | 02:35 | |
| *** Sukhdev_ has quit IRC | 02:41 | |
| *** armax has quit IRC | 02:43 | |
| *** su_zhang has quit IRC | 02:45 | |
| *** Jeffrey4l has joined #openstack-nova | 02:49 | |
| openstackgerrit | javeme proposed openstack/nova: raise exception ComputeHostNotFound if host is not found https://review.openstack.org/243105 | 02:52 |
| *** haomaiwang has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-nova | 03:01 | |
| *** vilobhmm11 has quit IRC | 03:04 | |
| *** browne has joined #openstack-nova | 03:06 | |
| *** mylu has joined #openstack-nova | 03:06 | |
| *** haplo37 has joined #openstack-nova | 03:06 | |
| *** daemontool has quit IRC | 03:08 | |
| *** links has joined #openstack-nova | 03:10 | |
| *** vilobhmm11 has joined #openstack-nova | 03:12 | |
| *** ljxiash has joined #openstack-nova | 03:12 | |
| *** fawadkhaliq has joined #openstack-nova | 03:16 | |
| *** yamahata has quit IRC | 03:17 | |
| *** thorst_ has joined #openstack-nova | 03:18 | |
| *** lucas-afk has quit IRC | 03:19 | |
| *** lucasagomes has joined #openstack-nova | 03:20 | |
| *** Marga_ has quit IRC | 03:21 | |
| *** sc68cal has quit IRC | 03:23 | |
| *** ljxiash has quit IRC | 03:23 | |
| *** ljxiash has joined #openstack-nova | 03:24 | |
| *** thorst_ has quit IRC | 03:24 | |
| *** haplo37 has quit IRC | 03:25 | |
| *** haplo37 has joined #openstack-nova | 03:26 | |
| *** baoli has joined #openstack-nova | 03:27 | |
| *** vilobhmm11 has quit IRC | 03:27 | |
| *** ljxiash has quit IRC | 03:28 | |
| *** vilobhmm11 has joined #openstack-nova | 03:31 | |
| *** mylu has quit IRC | 03:34 | |
| *** mylu has joined #openstack-nova | 03:35 | |
| *** su_zhang has joined #openstack-nova | 03:43 | |
| *** nekrodesk has joined #openstack-nova | 03:48 | |
| *** diana_clarke has joined #openstack-nova | 03:50 | |
| *** amotoki has joined #openstack-nova | 03:50 | |
| *** wyang has quit IRC | 03:53 | |
| *** wyang has joined #openstack-nova | 03:53 | |
| *** amotoki has quit IRC | 03:57 | |
| *** ducttape_ has joined #openstack-nova | 03:58 | |
| *** haomaiwang has quit IRC | 04:01 | |
| *** haomaiwang has joined #openstack-nova | 04:01 | |
| *** gjayavelu has joined #openstack-nova | 04:01 | |
| *** aginwala has joined #openstack-nova | 04:04 | |
| *** funeutron has joined #openstack-nova | 04:06 | |
| *** Sukhdev has joined #openstack-nova | 04:06 | |
| *** boris-42 has quit IRC | 04:14 | |
| *** dimtruck is now known as zz_dimtruck | 04:17 | |
| *** Daisy has joined #openstack-nova | 04:18 | |
| *** sc68cal has joined #openstack-nova | 04:20 | |
| *** mragupat has joined #openstack-nova | 04:21 | |
| *** thorst_ has joined #openstack-nova | 04:22 | |
| *** funeutron_ has joined #openstack-nova | 04:22 | |
| *** vilobhmm11 has quit IRC | 04:22 | |
| *** vilobhmm11 has joined #openstack-nova | 04:23 | |
| *** Daisy has quit IRC | 04:23 | |
| *** vilobhmm11 has quit IRC | 04:24 | |
| *** funeutron has quit IRC | 04:24 | |
| *** chhavi has joined #openstack-nova | 04:28 | |
| openstackgerrit | Augustina Ragwitz proposed openstack/nova: Fix _compare_result type handling comparison https://review.openstack.org/278191 | 04:28 |
| *** claudiub has quit IRC | 04:29 | |
| *** ducttape_ has quit IRC | 04:29 | |
| *** amotoki has joined #openstack-nova | 04:29 | |
| *** thorst_ has quit IRC | 04:30 | |
| *** Marga_ has joined #openstack-nova | 04:30 | |
| *** jgriffith is now known as jgriffith_away | 04:31 | |
| *** Marga_ has quit IRC | 04:32 | |
| *** PaulMurray has joined #openstack-nova | 04:32 | |
| *** Marga_ has joined #openstack-nova | 04:33 | |
| *** mylu has quit IRC | 04:34 | |
| *** armax has joined #openstack-nova | 04:34 | |
| *** aginwala has quit IRC | 04:34 | |
| *** mylu has joined #openstack-nova | 04:38 | |
| *** baoli has quit IRC | 04:39 | |
| *** salv-orl_ has joined #openstack-nova | 04:41 | |
| *** amotoki has quit IRC | 04:42 | |
| *** dims has quit IRC | 04:43 | |
| *** ljxiash has joined #openstack-nova | 04:43 | |
| *** salv-orlando has quit IRC | 04:44 | |
| *** haplo37 has quit IRC | 04:44 | |
| *** fawadkhaliq has quit IRC | 04:46 | |
| *** mfedosin has joined #openstack-nova | 04:47 | |
| *** ljxiash has quit IRC | 04:48 | |
| *** armax has quit IRC | 04:51 | |
| *** haomaiwang has quit IRC | 05:01 | |
| *** haomaiwa_ has joined #openstack-nova | 05:01 | |
| *** markvoelker has quit IRC | 05:03 | |
| *** armax has joined #openstack-nova | 05:05 | |
| *** jamielennox is now known as jamielennox|away | 05:06 | |
| openstackgerrit | Eric Brown proposed openstack/nova: Support multiple URLs for LDAP server https://review.openstack.org/277569 | 05:06 |
| *** aginwala has joined #openstack-nova | 05:08 | |
| *** zenoway has joined #openstack-nova | 05:08 | |
| *** fawadkhaliq has joined #openstack-nova | 05:11 | |
| *** aginwala has quit IRC | 05:12 | |
| *** zenoway has quit IRC | 05:12 | |
| *** funeutron_ has quit IRC | 05:14 | |
| *** dave-mcc_ has quit IRC | 05:15 | |
| *** shyama has joined #openstack-nova | 05:16 | |
| *** sudipto has joined #openstack-nova | 05:19 | |
| *** PaulMurray has quit IRC | 05:19 | |
| *** jgriffith_away is now known as jgriffith | 05:21 | |
| *** mylu has quit IRC | 05:24 | |
| *** mylu has joined #openstack-nova | 05:27 | |
| *** thorst_ has joined #openstack-nova | 05:27 | |
| *** jamielennox|away is now known as jamielennox | 05:31 | |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Remove version hardcoding for 'create' and 'update' methods https://review.openstack.org/259097 | 05:32 |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Make Xenplugin to work with glance v2 api https://review.openstack.org/266933 | 05:32 |
| *** thorst_ has quit IRC | 05:34 | |
| *** yamahata has joined #openstack-nova | 05:35 | |
| *** rk4n_ has quit IRC | 05:36 | |
| *** zenoway has joined #openstack-nova | 05:36 | |
| *** armax has quit IRC | 05:37 | |
| *** lpetrut has joined #openstack-nova | 05:39 | |
| *** zenoway has quit IRC | 05:40 | |
| *** tangchen has quit IRC | 05:40 | |
| *** ljxiash has joined #openstack-nova | 05:43 | |
| *** ljxiash has quit IRC | 05:48 | |
| *** vilobhmm11 has joined #openstack-nova | 05:48 | |
| *** SridharG has joined #openstack-nova | 05:49 | |
| *** ljxiash has joined #openstack-nova | 05:54 | |
| *** haomaiwa_ has quit IRC | 06:01 | |
| *** 18VAACNL5 has joined #openstack-nova | 06:01 | |
| *** lpetrut has quit IRC | 06:03 | |
| *** markvoelker has joined #openstack-nova | 06:04 | |
| *** wasmum has quit IRC | 06:05 | |
| *** jaosorior has joined #openstack-nova | 06:10 | |
| *** lpetrut has joined #openstack-nova | 06:19 | |
| *** SridharG has quit IRC | 06:24 | |
| *** thorst_ has joined #openstack-nova | 06:32 | |
| *** esracelik has joined #openstack-nova | 06:34 | |
| *** bkumar has joined #openstack-nova | 06:34 | |
| *** markvoelker has quit IRC | 06:34 | |
| *** moshele has joined #openstack-nova | 06:35 | |
| *** diana_clarke has quit IRC | 06:35 | |
| *** liverpooler has quit IRC | 06:37 | |
| *** unicell has joined #openstack-nova | 06:37 | |
| *** cfriesen has quit IRC | 06:39 | |
| *** thorst_ has quit IRC | 06:39 | |
| *** unicell has quit IRC | 06:42 | |
| *** unicell has joined #openstack-nova | 06:43 | |
| *** tyagiprince has joined #openstack-nova | 06:43 | |
| openstackgerrit | javeme proposed openstack/nova: raise exception ComputeHostNotFound if host is not found https://review.openstack.org/243105 | 06:44 |
| *** rotbeard has joined #openstack-nova | 06:48 | |
| *** garyk1 has joined #openstack-nova | 06:52 | |
| *** lpetrut has quit IRC | 06:53 | |
| *** ljxiash has quit IRC | 06:53 | |
| *** jaosorior has quit IRC | 06:56 | |
| *** mjura has joined #openstack-nova | 07:00 | |
| *** vishwanathj has quit IRC | 07:00 | |
| *** 18VAACNL5 has quit IRC | 07:01 | |
| *** haomaiwang has joined #openstack-nova | 07:01 | |
| openstackgerrit | esracelik proposed openstack/nova: Config options: centralize section "hyperv" https://review.openstack.org/255124 | 07:01 |
| *** aginwala has joined #openstack-nova | 07:02 | |
| *** mubirru has joined #openstack-nova | 07:02 | |
| *** tyagiprince1 has joined #openstack-nova | 07:03 | |
| *** sc68cal has quit IRC | 07:03 | |
| *** tyagiprince has quit IRC | 07:05 | |
| *** tyagiprince1 is now known as tyagiprince | 07:05 | |
| *** rk4n has joined #openstack-nova | 07:13 | |
| *** rdopiera has joined #openstack-nova | 07:14 | |
| *** paul-carlton1 has joined #openstack-nova | 07:15 | |
| *** Sukhdev has quit IRC | 07:15 | |
| *** SridharG has joined #openstack-nova | 07:19 | |
| *** jaosorior has joined #openstack-nova | 07:20 | |
| *** rdopiera has quit IRC | 07:24 | |
| *** rdopiera has joined #openstack-nova | 07:25 | |
| *** moshele has quit IRC | 07:25 | |
| openstackgerrit | Hans Lindgren proposed openstack/nova: Get instance security_groups from already fetched instance https://review.openstack.org/259813 | 07:30 |
| *** mpavone has joined #openstack-nova | 07:30 | |
| *** vilobhmm111 has joined #openstack-nova | 07:31 | |
| *** jaosorior has quit IRC | 07:32 | |
| *** jichen has joined #openstack-nova | 07:33 | |
| *** vilobhmm11 has quit IRC | 07:33 | |
| *** scheuran has joined #openstack-nova | 07:37 | |
| *** thorst_ has joined #openstack-nova | 07:37 | |
| *** amotoki has joined #openstack-nova | 07:38 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/278700 | 07:43 |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for hyperv options (1) https://review.openstack.org/268175 | 07:43 |
| *** thorst_ has quit IRC | 07:45 | |
| *** rk4n has quit IRC | 07:46 | |
| *** liusheng has quit IRC | 07:48 | |
| *** liusheng has joined #openstack-nova | 07:49 | |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for hyperv options (1) https://review.openstack.org/268175 | 07:49 |
| *** sahid has joined #openstack-nova | 07:50 | |
| *** zenoway has joined #openstack-nova | 07:54 | |
| *** ndipanov has quit IRC | 07:55 | |
| *** ljxiash has joined #openstack-nova | 07:56 | |
| *** vilobhmm11 has joined #openstack-nova | 07:57 | |
| *** vilobhmm111 has quit IRC | 07:57 | |
| *** nkrinner has joined #openstack-nova | 07:59 | |
| *** ljxiash has quit IRC | 08:00 | |
| *** haomaiwang has quit IRC | 08:01 | |
| *** rk4n has joined #openstack-nova | 08:01 | |
| *** haomaiwang has joined #openstack-nova | 08:01 | |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for hyperv options (2) https://review.openstack.org/278235 | 08:05 |
| *** jaosorior has joined #openstack-nova | 08:05 | |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for hyperv options (3) https://review.openstack.org/278236 | 08:09 |
| *** aginwala_ has joined #openstack-nova | 08:09 | |
| *** moshele has joined #openstack-nova | 08:09 | |
| *** sinese_ has joined #openstack-nova | 08:10 | |
| *** achanda has quit IRC | 08:11 | |
| *** aginwala has quit IRC | 08:11 | |
| *** zenoway has quit IRC | 08:11 | |
| *** zenoway has joined #openstack-nova | 08:12 | |
| *** zenoway has quit IRC | 08:12 | |
| *** mylu has quit IRC | 08:12 | |
| *** zenoway has joined #openstack-nova | 08:12 | |
| tyagiprince | Hey everyone.. I am facing a problem.. The thing is the hostname in the dns and the /etc/hostname file of the instance does not match.. dns finds the ip of the instance in this format : "host-10-X-X-X". I was trying to make it work. I just need the instance id to get the hostname (provided to instance in horizon during launching instance) from the metadata. | 08:13 |
| tyagiprince | Can someone please help me | 08:13 |
| *** achanda has joined #openstack-nova | 08:14 | |
| pkoniszewski | hello everyone, is it ok to add release notes post-merge? https://review.openstack.org/#/c/227278/ didn't contain release notes but I think it is something that is worth to mention in notes that it is possible now to block live migrate with a mix of shared and non-shared devices | 08:16 |
| *** paul-carlton1 has quit IRC | 08:21 | |
| bauzas | good morning Nova | 08:22 |
| bauzas | pkoniszewski: sure, we can discuss on the reno change | 08:22 |
| bauzas | and see if it's needed | 08:23 |
| *** jlanoux has joined #openstack-nova | 08:25 | |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Remove version hardcoding for 'create' and 'update' methods https://review.openstack.org/259097 | 08:25 |
| openstackgerrit | jichenjc proposed openstack/nova: relocate os_compute_api:servers:discoverable https://review.openstack.org/276740 | 08:26 |
| *** vilobhmm11 has quit IRC | 08:26 | |
| *** jed56 has joined #openstack-nova | 08:26 | |
| *** nkrinner has quit IRC | 08:26 | |
| *** zenoway has quit IRC | 08:28 | |
| pkoniszewski | bauzas: great | 08:28 |
| pkoniszewski | bauzas: so my change was part of a specless blueprint https://blueprints.launchpad.net/nova/+spec/block-live-migrate-with-attached-volumes | 08:29 |
| bauzas | pkoniszewski: in general, we try to provide a feature relnote only for the bps needing to modify a contractual API | 08:30 |
| bauzas | see http://docs.openstack.org/developer/nova/code-review.html#when-a-release-note-is-needed | 08:30 |
| bauzas | pkoniszewski: but like I said, we can see what you want to provide and discuss there | 08:30 |
| bauzas | I just dont want the relnotes to be a list of all our merged BPs | 08:31 |
| bauzas | :) | 08:31 |
| bauzas | but rather what the operators need to know about modifying | 08:31 |
| *** tyagiprince has quit IRC | 08:31 | |
| bauzas | that said - maybe johnthetubaguy and others could be having different opinions, like having a changelog getting all our merged BPs | 08:32 |
| *** nkrinner has joined #openstack-nova | 08:32 | |
| *** SridharG has quit IRC | 08:32 | |
| *** sridharg has joined #openstack-nova | 08:33 | |
| pkoniszewski | bauzas: well thats reasonable, im not sure we need a reno that mox has been removed from tests or something like that | 08:33 |
| openstackgerrit | esracelik proposed openstack/nova: Improve the help text for cells options (7) https://review.openstack.org/278927 | 08:33 |
| *** zakora has joined #openstack-nova | 08:34 | |
| *** liverpooler has joined #openstack-nova | 08:34 | |
| bauzas | yeah, that kind of BP is not needed as a note | 08:34 |
| *** su_zhang has quit IRC | 08:34 | |
| bauzas | all the refactoring bits too | 08:34 |
| pkoniszewski | bauzas: thing that i worked on was demanded by operators, thats why im thinking of providing reno for this | 08:34 |
| *** sinese_ has quit IRC | 08:34 | |
| pkoniszewski | bauzas: let me write reno and we will continue, ok? | 08:34 |
| *** su_zhang has joined #openstack-nova | 08:35 | |
| bauzas | hence me saying that in general things that are *real* features are either impacting a contractual API or adding a support to the hypervisor matrix | 08:35 |
| *** rcernin has joined #openstack-nova | 08:35 | |
| *** garyk2 has joined #openstack-nova | 08:36 | |
| *** garyk1 has quit IRC | 08:36 | |
| *** browne has quit IRC | 08:38 | |
| *** su_zhang has quit IRC | 08:39 | |
| *** paul-carlton1 has joined #openstack-nova | 08:40 | |
| *** mragupat has quit IRC | 08:40 | |
| openstackgerrit | Andrey Pavlov proposed openstack/nova: remove EC2 legacy code https://review.openstack.org/266425 | 08:41 |
| *** dedery has joined #openstack-nova | 08:41 | |
| *** e0ne has joined #openstack-nova | 08:42 | |
| *** thorst_ has joined #openstack-nova | 08:42 | |
| *** rk4n has quit IRC | 08:42 | |
| *** aginwala_ has quit IRC | 08:43 | |
| *** mylu has joined #openstack-nova | 08:43 | |
| *** achanda has quit IRC | 08:44 | |
| *** claudiub has joined #openstack-nova | 08:44 | |
| *** aginwala has joined #openstack-nova | 08:44 | |
| *** amotoki has quit IRC | 08:45 | |
| *** matrohon has joined #openstack-nova | 08:45 | |
| *** hemna_ has joined #openstack-nova | 08:45 | |
| *** irenab_ has joined #openstack-nova | 08:46 | |
| *** bkumar4 has joined #openstack-nova | 08:46 | |
| *** openstackgerrit has quit IRC | 08:47 | |
| *** openstackgerrit_ has joined #openstack-nova | 08:47 | |
| *** mylu has quit IRC | 08:47 | |
| openstackgerrit_ | Claudiu Belu proposed openstack/nova: Fixes python 3 unit tests https://review.openstack.org/278810 | 08:47 |
| *** openstackgerrit_ is now known as openstackgerrit | 08:48 | |
| *** gjayavelu has quit IRC | 08:49 | |
| *** aginwala has quit IRC | 08:49 | |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Added db API layer to add instance tag-list filtering support https://review.openstack.org/248805 | 08:49 |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Added server tags controller https://review.openstack.org/128940 | 08:49 |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Added server tags support in nova-api https://review.openstack.org/268932 | 08:49 |
| *** alex_xu_ has joined #openstack-nova | 08:49 | |
| openstackgerrit | javeme proposed openstack/nova: raise exception ComputeHostNotFound if host is not found https://review.openstack.org/243105 | 08:50 |
| *** thorst_ has quit IRC | 08:50 | |
| *** liverpooler has quit IRC | 08:51 | |
| *** ildikov has quit IRC | 08:51 | |
| *** liverpooler has joined #openstack-nova | 08:51 | |
| *** nkrinner has quit IRC | 08:52 | |
| *** liusheng has quit IRC | 08:52 | |
| *** bkumar has quit IRC | 08:52 | |
| *** shyama has quit IRC | 08:52 | |
| *** alex_xu has quit IRC | 08:52 | |
| *** irenab has quit IRC | 08:52 | |
| *** hemnafk has quit IRC | 08:52 | |
| *** irenab_ is now known as irenab | 08:52 | |
| *** liusheng has joined #openstack-nova | 08:52 | |
| *** shyama has joined #openstack-nova | 08:54 | |
| *** rmart04 has joined #openstack-nova | 08:54 | |
| *** omnipresent has joined #openstack-nova | 08:56 | |
| *** slunkad has quit IRC | 08:56 | |
| *** gszasz has joined #openstack-nova | 08:56 | |
| *** nkrinner has joined #openstack-nova | 08:57 | |
| *** rk4n has joined #openstack-nova | 08:57 | |
| *** fawadkhaliq has quit IRC | 08:59 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: libvirt: fixes python 3 related unit tests https://review.openstack.org/278834 | 08:59 |
| *** rk4n has quit IRC | 08:59 | |
| *** bsv has joined #openstack-nova | 08:59 | |
| *** paul-carlton1 has quit IRC | 09:00 | |
| *** bsv has left #openstack-nova | 09:00 | |
| *** derekh has joined #openstack-nova | 09:00 | |
| *** haomaiwang has quit IRC | 09:01 | |
| *** BobBall_AWOL is now known as Bobball | 09:01 | |
| *** Bobball is now known as BobBall | 09:01 | |
| *** haomaiwa_ has joined #openstack-nova | 09:01 | |
| *** rmart04 has left #openstack-nova | 09:05 | |
| *** tyagiprince has joined #openstack-nova | 09:05 | |
| tyagiprince | Hey everyone.. I am facing a problem.. The thing is the hostname in the dns and the /etc/hostname file of the instance does not match.. dns finds the ip of the instance in this format : "host-10-X-X-X". I was trying to make it work. I just need the instance id to get the hostname (provided to instance in horizon during launching instance) from the metadata. | 09:06 |
| *** rk4n has joined #openstack-nova | 09:07 | |
| *** ihrachys has joined #openstack-nova | 09:07 | |
| tyagiprince | I am not able to resolve the instance with the name given to the instance. It is because the dns entry is different. | 09:07 |
| *** slunkad has joined #openstack-nova | 09:08 | |
| *** omnipresent has quit IRC | 09:11 | |
| *** mylu has joined #openstack-nova | 09:13 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Update ComputeNode values with disk allocation ratios in the RT https://review.openstack.org/277953 | 09:15 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Move Disk allocation ratio to ResourceTracker https://review.openstack.org/245619 | 09:15 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Add disk_allocation_ratio to ComputeNode https://review.openstack.org/277934 | 09:15 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Update HostManager and DiskFilter to use ComputeNode disk ratio https://review.openstack.org/278112 | 09:15 |
| *** mylu has quit IRC | 09:18 | |
| *** rotbeard has quit IRC | 09:22 | |
| *** paul-carlton1 has joined #openstack-nova | 09:22 | |
| *** yassine has joined #openstack-nova | 09:25 | |
| *** efoley has joined #openstack-nova | 09:25 | |
| *** rk4n has quit IRC | 09:25 | |
| *** ildikov has joined #openstack-nova | 09:26 | |
| *** yassine has quit IRC | 09:26 | |
| *** yassine has joined #openstack-nova | 09:26 | |
| *** mvk has joined #openstack-nova | 09:27 | |
| *** eliqiao_ has quit IRC | 09:29 | |
| *** danpb has joined #openstack-nova | 09:31 | |
| *** markvoelker has joined #openstack-nova | 09:31 | |
| *** ildikov has quit IRC | 09:33 | |
| *** garyk has joined #openstack-nova | 09:35 | |
| *** garyk2 has quit IRC | 09:35 | |
| *** eliqiao_ has joined #openstack-nova | 09:35 | |
| *** markvoelker has quit IRC | 09:36 | |
| *** scheuran has quit IRC | 09:37 | |
| openstackgerrit | Stephen Finucane proposed openstack/nova: config options: Centralise 'virt.firewall' options https://review.openstack.org/257995 | 09:40 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: config options: Centralise 'virt.netutils' options https://review.openstack.org/258009 | 09:40 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: Improve 'virt.images' conf options documentation https://review.openstack.org/259553 | 09:40 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: config options: Centralise 'virt.imagecache' options https://review.openstack.org/258984 | 09:40 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: config options: Centralise 'virt.images' options https://review.openstack.org/258004 | 09:40 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: config options: Centralise 'virt.disk' options https://review.openstack.org/258912 | 09:40 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: config options: Centralise 'virt.driver' options https://review.openstack.org/257956 | 09:40 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: Improve 'virt.driver' conf options documentation https://review.openstack.org/266364 | 09:40 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: Improve 'virt.firewall' conf options documentation https://review.openstack.org/266365 | 09:41 |
| *** sfinucan has joined #openstack-nova | 09:41 | |
| *** shz has quit IRC | 09:41 | |
| johnthetubaguy | pkoniszewski: bauzas: so if there is something the reader of the release notes are interested in, it seems like a good add | 09:42 |
| *** shz has joined #openstack-nova | 09:42 | |
| *** omnipresent has joined #openstack-nova | 09:43 | |
| *** ociuhandu has quit IRC | 09:44 | |
| *** claudiub has quit IRC | 09:45 | |
| *** ildikov has joined #openstack-nova | 09:45 | |
| *** thorst_ has joined #openstack-nova | 09:48 | |
| *** omnipresent has quit IRC | 09:48 | |
| *** romainh has joined #openstack-nova | 09:48 | |
| *** ccesario has quit IRC | 09:50 | |
| *** andymaier has joined #openstack-nova | 09:50 | |
| *** ganso has joined #openstack-nova | 09:51 | |
| *** scheuran has joined #openstack-nova | 09:51 | |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: [microversions] Enable 2.20 https://review.openstack.org/278446 | 09:52 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: [microversions] Enable 2.21 https://review.openstack.org/278454 | 09:53 |
| *** kaisers has joined #openstack-nova | 09:54 | |
| *** thorst_ has quit IRC | 09:55 | |
| *** andymaier has quit IRC | 09:56 | |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Remove version hardcoding for 'create' and 'update' methods https://review.openstack.org/259097 | 09:59 |
| *** cgalan has joined #openstack-nova | 09:59 | |
| *** scheuran has quit IRC | 10:00 | |
| *** haomaiwa_ has quit IRC | 10:01 | |
| *** haomaiwang has joined #openstack-nova | 10:01 | |
| *** ccesario has joined #openstack-nova | 10:01 | |
| *** scheuran has joined #openstack-nova | 10:02 | |
| openstackgerrit | John Garbutt proposed openstack/nova: Update ComputeNode values with disk allocation ratios in the RT https://review.openstack.org/277953 | 10:02 |
| *** nekrodesk has quit IRC | 10:02 | |
| *** mingyu_ has joined #openstack-nova | 10:03 | |
| johnthetubaguy | bauzas: I attempted to reword that reno note. Does that look better ^ | 10:03 |
| *** mingyu_ has quit IRC | 10:03 | |
| bauzas | johnthetubaguy: looking :) | 10:03 |
| *** efoley_ has joined #openstack-nova | 10:04 | |
| *** omnipresent has joined #openstack-nova | 10:04 | |
| *** ljxiash has joined #openstack-nova | 10:04 | |
| bauzas | johnthetubaguy: I was basically lazy so I just copied my comments in the Liberty ReleaseNotes page | 10:05 |
| *** efoley has quit IRC | 10:05 | |
| bauzas | johnthetubaguy: but your comments look good | 10:05 |
| bauzas | johnthetubaguy: if you're okay, I'll just edit for a typo | 10:06 |
| johnthetubaguy | bauzas: totally, go for it | 10:06 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Update ComputeNode values with disk allocation ratios in the RT https://review.openstack.org/277953 | 10:07 |
| bauzas | johnthetubaguy: thanks for helping, I'm very bad at phrasing things :/ | 10:08 |
| *** ljxiash has quit IRC | 10:09 | |
| *** rk4n has joined #openstack-nova | 10:09 | |
| openstackgerrit | Moshe Levi proposed openstack/os-vif: Adding dev_type field to VIFHostDevice https://review.openstack.org/277871 | 10:10 |
| johnthetubaguy | bauzas: no problems, I just had an idea I wanted to try out really | 10:10 |
| *** tyagiprince1 has joined #openstack-nova | 10:11 | |
| johnthetubaguy | raito, oops | 10:11 |
| *** tyagiprince has quit IRC | 10:11 | |
| *** tyagiprince1 is now known as tyagiprince | 10:11 | |
| johnthetubaguy | bauzas: thanks for dropping that first reno note, I think thats what was confusing me about things yesterday, much easier read now | 10:12 |
| bauzas | johnthetubaguy: yeah it was confusing | 10:12 |
| bauzas | it was even confusing myself, oops | 10:12 |
| *** mylu has joined #openstack-nova | 10:14 | |
| *** rk4n_ has joined #openstack-nova | 10:15 | |
| *** rk4n has quit IRC | 10:15 | |
| *** ndipanov has joined #openstack-nova | 10:16 | |
| *** efoley_ is now known as efoley | 10:17 | |
| johnthetubaguy | bauzas: looks good though, +2 for all those | 10:19 |
| *** mylu has quit IRC | 10:19 | |
| bauzas | thanks | 10:19 |
| bauzas | I'm a bit torn, because I don't want to throw the resource-providers changes under the bus | 10:20 |
| *** rk4n_ has quit IRC | 10:20 | |
| *** rk4n has joined #openstack-nova | 10:20 | |
| bauzas | because I'm adding a new migration + a ComputeNode object version so it would wedge the UUID allocation ratio :) | 10:20 |
| johnthetubaguy | bauzas: do we have the patches up for those select_destinations bits, btw? looked like we got a good few of those | 10:20 |
| bauzas | and I think the resource-providers BP is by far more important :) | 10:21 |
| bauzas | johnthetubaguy: select_dests or check-dest ? | 10:21 |
| johnthetubaguy | bauzas: I mean check, oopps | 10:21 |
| bauzas | johnthetubaguy: so the one for evacuate merged | 10:21 |
| bauzas | I'm just atm working on the live-migrate one | 10:21 |
| bauzas | sec | 10:21 |
| bauzas | https://review.openstack.org/#/c/273104/ (merged) | 10:22 |
| bauzas | and https://review.openstack.org/277800 for live-migrate (WIP) | 10:22 |
| johnthetubaguy | bauzas: ah thats right, cools | 10:23 |
| bauzas | I will have to add for (cold) migrate and unshelve too | 10:23 |
| bauzas | once that will be merged, it could help a lot of operatoes | 10:23 |
| johnthetubaguy | bauzas: +1 | 10:23 |
| *** rk4n_ has joined #openstack-nova | 10:24 | |
| *** matrohon has quit IRC | 10:26 | |
| *** Daisy has joined #openstack-nova | 10:26 | |
| *** rk4n has quit IRC | 10:27 | |
| *** funeutron has joined #openstack-nova | 10:28 | |
| johnthetubaguy | bauzas: this pattern feels a bit odd, I wonder if we should just get hold of the request spec in the conductor, rather than create the request spec in one place and fetch it in the other? | 10:28 |
| bauzas | johnthetubaguy: not sure I got your idea | 10:29 |
| bauzas | johnthetubaguy: you don't like me hydrating a fresh one in the conductor ? | 10:29 |
| *** efoley has quit IRC | 10:30 | |
| bauzas | johnthetubaguy: since the RequestSpec object is created in the API DB, I really wanted to pass it down rather than getting it in the conductor | 10:30 |
| *** Daisy has quit IRC | 10:31 | |
| johnthetubaguy | bauzas: OK, but the conductor is also in the API layer, when calling the scheduler, I thought | 10:31 |
| bauzas | johnthetubaguy: that's atm not yet sure | 10:31 |
| johnthetubaguy | bauzas: yeah, true | 10:31 |
| bauzas | johnthetubaguy: we were discussing of an API-conductor stuff | 10:31 |
| bauzas | johnthetubaguy: that way, I was not wedging the cells v2 effort | 10:32 |
| johnthetubaguy | bauzas: so maybe switch this around, could we not create the missing request_spec in the API later, rather than pass None? | 10:32 |
| bauzas | because if I was calling the Spec object in the in-cell conductor, it would need some cell-to-api call | 10:32 |
| bauzas | johnthetubaguy: but what if the client is old ? | 10:32 |
| bauzas | I mean the RPC client | 10:33 |
| johnthetubaguy | the old API doesn't sent the request spec... true, I guess we need the logic there anyways, boo | 10:33 |
| bauzas | we will still get a nullable request_spec then ? | 10:33 |
| bauzas | yeah | 10:33 |
| bauzas | that's a bit unfortunate, I reckon | 10:33 |
| bauzas | also | 10:34 |
| bauzas | we somehow need to define when we want to stop supporting old running instances | 10:34 |
| johnthetubaguy | oh wait, the request_spec is only in the API? | 10:34 |
| bauzas | the persisted object, yes | 10:34 |
| bauzas | in the API DB tbc | 10:34 |
| johnthetubaguy | I thought that would move to the child cell DB, once the instance is created... hmm. | 10:34 |
| bauzas | no | 10:35 |
| bauzas | no AFAIK | 10:35 |
| bauzas | not* | 10:35 |
| bauzas | but we could check that with alaski | 10:35 |
| johnthetubaguy | OK, so I had some bad assumptions in my head here, but that works | 10:35 |
| bauzas | so we have the ReqSpec object and the BuildRequest object for getting the instance info | 10:35 |
| bauzas | in the API DB | 10:35 |
| bauzas | then, it's passed down thru the cell DB for creating the instance | 10:36 |
| johnthetubaguy | yeah, I guess I was having it attached to the instance, in my head, but that gets messy | 10:36 |
| bauzas | I know it's possible to communicate from the child cell to the API DB, but I would like to avoid that as much as possible | 10:36 |
| johnthetubaguy | totally agreed with you | 10:37 |
| johnthetubaguy | the cross talk needs to be at a minimum | 10:37 |
| johnthetubaguy | actually, I quite like not having a nova-conductor in the API cell, but the problem is build requests, and which conductor you send that to | 10:38 |
| bauzas | that's one of the reasons we have some race conditions when instance.save() for example :) | 10:38 |
| bauzas | in cells v1 | 10:38 |
| johnthetubaguy | bauzas: they plague our production system, I know them well | 10:38 |
| bauzas | CAP you said ? :p | 10:39 |
| * johnthetubaguy hmm, got disconnected there | 10:41 | |
| *** salv-orlando has joined #openstack-nova | 10:42 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/nova: Add support for using Ironic's node name as the hostname of the instance https://review.openstack.org/238593 | 10:42 |
| *** ankit_ag has joined #openstack-nova | 10:44 | |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: libvirt: Acquire TCP ports for console during live migration https://review.openstack.org/275801 | 10:44 |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: libvirt: Add serial ports to the migration data object https://review.openstack.org/215102 | 10:44 |
| openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: libvirt: move "_get_serial_ports_from_guest" to Guest class https://review.openstack.org/236051 | 10:44 |
| *** achanda has joined #openstack-nova | 10:44 | |
| *** salv-orl_ has quit IRC | 10:45 | |
| openstackgerrit | Moshe Levi proposed openstack/os-vif: Provide plugins an oslo_config group for their setup https://review.openstack.org/275281 | 10:45 |
| openstackgerrit | Moshe Levi proposed openstack/os-vif: import hw_veb plugin implementation https://review.openstack.org/278972 | 10:45 |
| *** dims has joined #openstack-nova | 10:48 | |
| *** scheuran has quit IRC | 10:50 | |
| *** lpetrut has joined #openstack-nova | 10:50 | |
| *** rfolco has joined #openstack-nova | 10:51 | |
| *** achanda has quit IRC | 10:51 | |
| *** thorst_ has joined #openstack-nova | 10:52 | |
| *** mvk has quit IRC | 10:54 | |
| dulek | I know dansmith isn't here now, but maybe someone else will be able to answer. | 10:55 |
| dulek | What is your policy of bumping the major version of the RPC API? | 10:55 |
| dulek | For rolling upgrades Nova supports N-1 backward compatibility. When to raise the RPC version then to be able not to break it? | 10:56 |
| *** cgalan has quit IRC | 10:57 | |
| *** cgalan has joined #openstack-nova | 10:57 | |
| *** thorst_ has quit IRC | 11:00 | |
| *** haomaiwang has quit IRC | 11:01 | |
| *** haomaiwa_ has joined #openstack-nova | 11:01 | |
| *** cdent has joined #openstack-nova | 11:01 | |
| *** ildikov has quit IRC | 11:03 | |
| *** aysyd has joined #openstack-nova | 11:03 | |
| *** scheuran has joined #openstack-nova | 11:05 | |
| ankit_ag | ndipanov: Hi, could you please add your opinion on a pending review https://review.openstack.org/#/c/221162 when you get time | 11:06 |
| *** FL1SK has quit IRC | 11:07 | |
| *** claudiub has joined #openstack-nova | 11:12 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes python 3 unit tests https://review.openstack.org/278810 | 11:14 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: libvirt: fixes python 3 related unit tests https://review.openstack.org/278834 | 11:14 |
| johnthetubaguy | dulek: I can answer some of that | 11:15 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: numa: Fixes NUMA topology related unit tests https://review.openstack.org/278835 | 11:15 |
| johnthetubaguy | dulek: we only ever support N to N+1 upgrades, and we make sure any changes respect that | 11:15 |
| johnthetubaguy | dulek: we generally bump the major versions at the end of a release when there is enough technical debt we can drop | 11:16 |
| *** fawadkhaliq has joined #openstack-nova | 11:16 | |
| openstackgerrit | Dmitry Stepanenko proposed openstack/nova: Added support_requests method to PciDevicePoolList https://review.openstack.org/242087 | 11:17 |
| johnthetubaguy | dulek: the online data migrations, and data schema contractions are the main limit on the N to N+1 rule at this point, rather than RPC or object versioning, although we make use of both when it makes sense | 11:17 |
| johnthetubaguy | dulek: is that what you were meaning? | 11:17 |
| *** romainh has quit IRC | 11:18 | |
| dulek | johnthetubaguy: In Cinder we would love to drop our RPC compatibility shims every release - that would require major version bump. | 11:18 |
| johnthetubaguy | dulek: when we bump the major version, we actually make the system support the new major version (n).0 and all the (n-1).x versions as well, for a short time, usually at the very end of a release | 11:18 |
| johnthetubaguy | dulek: let me find an example | 11:18 |
| dulek | johnthetubaguy: Ah, that's what I've wondered about! | 11:18 |
| johnthetubaguy | dulek: this is the supporting two major versions of an object: https://github.com/openstack/nova/blob/stable/liberty/nova/objects/instance.py#L980 | 11:19 |
| johnthetubaguy | dulek: this is the supporting of two major RPC versions: https://github.com/openstack/nova/blob/stable/liberty/nova/conductor/manager.py#L924 | 11:20 |
| bauzas | dulek: saw dan's blogposts ? | 11:20 |
| johnthetubaguy | dulek: basically a proxy that supports the latest version | 11:20 |
| dulek | johnthetubaguy: So as Cinder is running 1.x currently, we should bump to 2.0 just before M's release (but without dropping any compatibility) and in early N drop the support for 1.0? | 11:20 |
| johnthetubaguy | bauzas good point about this series: http://www.danplanet.com/blog/2015/10/05/upgrades-in-nova-rpc-apis/ | 11:21 |
| dulek | bauzas: Sure, I've read them multiple times. :) | 11:21 |
| johnthetubaguy | dulek: yeah, thats basically it, let me find our notes on that | 11:21 |
| bauzas | dulek: ack, so in general you bump a major when you want to stop supporting the old one, so what johnthetubaguy said | 11:21 |
| *** daemontool has joined #openstack-nova | 11:21 | |
| johnthetubaguy | dulek: https://wiki.openstack.org/wiki/Nova/ReleaseChecklist and https://wiki.openstack.org/wiki/RpcMajorVersionUpdates | 11:22 |
| dulek | bauzas: From the posts "I’ll focus on minor revisions below and save major bumps for a later post.". | 11:22 |
| dulek | bauzas: "Later" never happened I think. ;) | 11:22 |
| bauzas | dulek: but that's not really because of $release, rather because we want to change something | 11:22 |
| johnthetubaguy | dulek: yeah, I think thats the important one: https://wiki.openstack.org/wiki/RpcMajorVersionUpdates alongside some solid grenade tests to keep you honest | 11:23 |
| dulek | johnthetubaguy: Ah, these two wiki pages are awesome, thank you! | 11:23 |
| johnthetubaguy | dulek: we kept screwing it up, so we wrote it down! | 11:23 |
| * dulek is running back to the whiteboard. :) | 11:23 | |
| *** cgalan has quit IRC | 11:24 | |
| bauzas | dulek: see also https://review.openstack.org/#/c/230132/ | 11:25 |
| *** cgalan has joined #openstack-nova | 11:25 | |
| *** funeutron has quit IRC | 11:26 | |
| johnthetubaguy | bauzas: do we still need those now, with auto? I guess we only have auto for the compute nodes? | 11:26 |
| *** mvk has joined #openstack-nova | 11:26 | |
| bauzas | johnthetubaguy: yeah very good point | 11:26 |
| bauzas | johnthetubaguy: so | 11:27 |
| bauzas | I'm not dan | 11:27 |
| bauzas | but | 11:27 |
| bauzas | I would say the auto flag is not yet deprecating the old rpc pins | 11:27 |
| bauzas | lemme check the series | 11:27 |
| cdent | alaski, johnthetubaguy: your recent comments on the resource pools spec suggest we've still got a lot of ambiguity about in which database the resource-* tables are going to live. I think we're okay for the time being to carry on while we hash it out: it's easy to change (before things merge). | 11:28 |
| johnthetubaguy | cdent: yeah, I meant to ask about that, whats the current plan, compute cell DB? | 11:29 |
| bauzas | cdent: johnthetubaguy: so we had a brief discussion on that yesterday's evening for cells meeting | 11:29 |
| cdent | johnthetubaguy: that's where we left things the other evening, but I know alaski has some reservations | 11:29 |
| bauzas | tl;dr: doffm had an idea | 11:29 |
| bauzas | where the resource-* tables would stay in the cell DB while the aggregates table would be in the api db | 11:30 |
| bauzas | that would be the ideal | 11:30 |
| johnthetubaguy | so lets say scheduler is external, where does it live? | 11:30 |
| bauzas | we don't really want to have the API DB having the resource-* tables | 11:30 |
| bauzas | in a 3rd DB :) | 11:30 |
| bauzas | but that's a story for the ones who would split the scheduler | 11:31 |
| bauzas | neither for cellsv2 or the resouce-providers series IMHO | 11:31 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: [microversions] Enable 2.21 https://review.openstack.org/278454 | 11:31 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: [microversions] Enable 2.20 https://review.openstack.org/278446 | 11:31 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: [microversions] Add support for 2.19 https://review.openstack.org/278385 | 11:31 |
| bauzas | johnthetubaguy: if we would split the scheduler, it would require a 3rd DB anyway, right ? :) | 11:32 |
| * johnthetubaguy one sec on the phone | 11:32 | |
| bauzas | np | 11:32 |
| *** markvoelker has joined #openstack-nova | 11:32 | |
| johnthetubaguy | bauzas: cdent: so I was just thinking about this from the point of view of claiming a resource when you have two cells | 11:35 |
| johnthetubaguy | lets say an IP zone is shared between two cells | 11:35 |
| johnthetubaguy | we want atomic commits to a single DB table, so that resource table has to live somewhere, not two places | 11:36 |
| johnthetubaguy | that makes me think, put the resources in the API DB | 11:36 |
| johnthetubaguy | (or a scheduler DB) | 11:36 |
| johnthetubaguy | rather than the child cell DB | 11:36 |
| *** Marga_ has quit IRC | 11:36 | |
| *** markvoelker has quit IRC | 11:37 | |
| *** Marga_ has joined #openstack-nova | 11:37 | |
| johnthetubaguy | but host to host aggregate mapping, should live in the child DB, because when we add cells we add hosts (now the aggregate its-self probably should live in the API cell DB). | 11:37 |
| johnthetubaguy | bauzas: would that make sense? | 11:37 |
| * bauzas considering that | 11:38 | |
| johnthetubaguy | now the problem here, is that its a made up use case | 11:38 |
| cdent | johnthetubaguy: that presupposes resources which span cells, which is an unresolved question: one solution is for logical resources to be per cell, even if physical resources are not (slices of shared disk for example) | 11:38 |
| *** PaulMurray has joined #openstack-nova | 11:38 | |
| bauzas | cdent: johnthetubaguy: well, we also tackled that question yesterday | 11:38 |
| johnthetubaguy | OK, I should maybe read the log | 11:39 |
| cdent | however I don't want to be seen as promoting the "keep it in the cells db" case. I don't really have a strong position, because of ignorance. | 11:39 |
| bauzas | and for the moment, the point is that aggregates are spanning cells | 11:39 |
| johnthetubaguy | each cell having a different pool of CPU resources seems fine, and that probably should live in the cell db | 11:39 |
| johnthetubaguy | its when a single pool spans multiple cells | 11:39 |
| johnthetubaguy | maybe we just exclude that case | 11:40 |
| bauzas | so the point was that the operator could just reserve a chunk of the resources | 11:40 |
| bauzas | for each cell | 11:40 |
| johnthetubaguy | bauzas: thats not a terrible work around | 11:40 |
| *** amotoki has joined #openstack-nova | 11:41 | |
| johnthetubaguy | that is, for a case I don't really care about much | 11:41 |
| bauzas | so we would still have aggregates across cells, but resources being "sharded" (note the brackets) between cells | 11:41 |
| johnthetubaguy | cdent: bauzas: so I am good with plan A, i.e. keep it in the cell DB | 11:41 |
| johnthetubaguy | bauzas: aggregates crossing cells is fine | 11:41 |
| johnthetubaguy | well, thats almost certainly required | 11:42 |
| bauzas | there are many usecases indeed | 11:42 |
| johnthetubaguy | group all hosts with SSDs, etc | 11:42 |
| bauzas | heh, see the logs :p | 11:42 |
| cdent | I can predict with a fair degree of confidence that we'll cross this issue again when we start writing the resource-pools code, so we might just want to keep the top of change stack unmerged for a while? | 11:43 |
| cdent | at the moment everything is based off this: https://review.openstack.org/#/c/275432/ | 11:43 |
| bauzas | I dunno | 11:43 |
| cdent | the resource classes have merged, but that's fine because no db | 11:43 |
| johnthetubaguy | cdent: thats not a bad idea | 11:43 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: Provide plugins an oslo_config group for their setup https://review.openstack.org/275281 | 11:44 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: vif_plug_ovs: Disable IPv6 on bridge devices https://review.openstack.org/275293 | 11:44 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: import openvswitch plugin implementation https://review.openstack.org/273610 | 11:44 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: import linux bridge plugin implementation https://review.openstack.org/273609 | 11:44 |
| bauzas | cdent: johnthetubaguy: well tbh, I +2d it because I was agreeing the consensus to keep the resource-* tables in-tree | 11:44 |
| bauzas | oops | 11:44 |
| bauzas | in-cell | 11:44 |
| bauzas | but that could wait some cellsv2 folks to approve that | 11:45 |
| bauzas | like melwitt and alaski | 11:45 |
| johnthetubaguy | yeah, I like it, but I would love alaski to do the wambating | 11:45 |
| johnthetubaguy | wombat | 11:45 |
| cdent | I'm okay with anything, just want to avoid as much dancing or duplication as possible when doing the work. | 11:45 |
| bauzas | johnthetubaguy: so, let's wait for his axe then :) | 11:46 |
| bauzas | merging the compute.uuid is certainly not a dependency tho, so +1ing your idea to unblock it, cdent | 11:46 |
| bauzas | even if that would trample my own series :p | 11:47 |
| *** achanda has joined #openstack-nova | 11:48 | |
| *** achanda has quit IRC | 11:52 | |
| *** tyagiprince1 has joined #openstack-nova | 11:53 | |
| johnthetubaguy | bauzas: cdent: wondering if we can use cell0 for any shared pools... but probably best not to overthink it | 11:54 |
| cdent | johnthetubaguy: that idea is somewhere in the logs of the few other times we've had this discussion and the reaction was the same "best not to overthink it" | 11:55 |
| *** tyagiprince has quit IRC | 11:55 | |
| *** tyagiprince1 is now known as tyagiprince | 11:55 | |
| johnthetubaguy | cdent: ah, cool | 11:55 |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Remove version hardcoding for 'create' and 'update' methods https://review.openstack.org/259097 | 11:57 |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Make Xenplugin to work with glance v2 api https://review.openstack.org/266933 | 11:57 |
| *** thorst_ has joined #openstack-nova | 11:58 | |
| *** moorryan has joined #openstack-nova | 11:58 | |
| *** garyk has quit IRC | 11:59 | |
| *** garyk has joined #openstack-nova | 11:59 | |
| *** amotoki has quit IRC | 12:00 | |
| *** haomaiwa_ has quit IRC | 12:01 | |
| *** haomaiwang has joined #openstack-nova | 12:01 | |
| *** thorst_ has quit IRC | 12:02 | |
| sdague | BobBall: is it important that the xen plugins don't end in .py? | 12:03 |
| BobBall | Nope | 12:03 |
| sdague | because, that's making flake8 ignore them in tree | 12:03 |
| *** amotoki has joined #openstack-nova | 12:03 | |
| BobBall | xapi plugins don't have to be python either - just anything that can interpret xmlrpm | 12:03 |
| BobBall | xmlrpc of course | 12:03 |
| openstackgerrit | Alvaro Lopez Garcia proposed openstack/nova: libvirt: make live_migration_uri flag dependent on virt_type https://review.openstack.org/175780 | 12:03 |
| sdague | ok, what would be the impact if we renamed them to have a .py? | 12:03 |
| BobBall | incompatible change - so I'd say a major version number increment. johnthetubaguy might have a few words to say about that as it would require upgrading plugins + nova in sync rather than upgrading the plugins on the host in advance | 12:04 |
| tdurakov_ | johnthetubaguy, hey, got question about upgrades, migrate_data object was introduced during mitaka, and already got 2 versions: 1.0 and 1.1 should I worry about compatibility of its versions, as I'm going to add 1.2 to migrate_data | 12:06 |
| BobBall | Can we do symlinks? or are they verboten in the nova code? | 12:07 |
| sdague | honestly, I don't know | 12:07 |
| BobBall | If we can symlink, we could rename files + symlink from the original to ease the upgrade impact | 12:07 |
| tdurakov_ | so nova guarantee compatibility between releases or during one release too? | 12:07 |
| sdague | the current state is just kind of sad | 12:07 |
| johnthetubaguy | we have symlinks in the specs repo | 12:07 |
| *** tdurakov_ is now known as tdurakov | 12:07 | |
| sdague | there is no testing on those files at all | 12:07 |
| johnthetubaguy | I think the original idea for .py and not .py is that the .py were not executable | 12:07 |
| sdague | and they don't even syntax check | 12:08 |
| BobBall | xenstore.py is executable though :) The only plugin that has the .py extension | 12:08 |
| johnthetubaguy | sdague: BobBall: I have a patch to add unit testing, I dug that out for BobBall a week ago or so | 12:08 |
| sdague | johnthetubaguy: ok, well can we prioritize that? | 12:08 |
| BobBall | testing is on our backlog - but my team is in PRC and all on vacation :) | 12:08 |
| BobBall | The joys of chinese new year. | 12:09 |
| *** garyk has quit IRC | 12:09 | |
| *** garyk has joined #openstack-nova | 12:09 | |
| sdague | I kind of feel like this is a giant hole with the xen code, to the point of demoting it's support class. | 12:09 |
| sdague | and it makes any of these other changes to do things like the glance support so much harder to review | 12:09 |
| johnthetubaguy | I am good with renaming them all to be .py, and add symlinks for backwards compat | 12:10 |
| BobBall | It is a big hole indeed - and while I don't expect unit tests to get in Mitaka, I'm sure they will be up for review early in Newton. | 12:10 |
| johnthetubaguy | BobBall: why not Mitaka? just the basic bits, not fully tested | 12:11 |
| *** garyk has quit IRC | 12:11 | |
| *** garyk has joined #openstack-nova | 12:11 | |
| johnthetubaguy | we need to get to a point where all new stuff can add tests, ASAP, and add more coverage will take time, and be netwon | 12:11 |
| BobBall | Basic coverage may be possible; I was meaning proper coverage. let's discuss tomorrow. | 12:11 |
| openstackgerrit | Andrey Pavlov proposed openstack/nova: remove EC2 legacy code https://review.openstack.org/266425 | 12:11 |
| *** tyagiprince has quit IRC | 12:11 | |
| sdague | BobBall: I'm happy for *anything* at this point | 12:11 |
| sdague | because right now we have nothing | 12:12 |
| johnthetubaguy | right | 12:12 |
| sdague | like we don't even know if those files are python when we approve them | 12:12 |
| *** ig0r_ has joined #openstack-nova | 12:12 | |
| johnthetubaguy | tdurakov: you have to worry about compatibility, until the major version is bumped | 12:14 |
| *** mylu has joined #openstack-nova | 12:14 | |
| johnthetubaguy | tdurakov: its a little tricky, but the unit test usually tells you if you changed something that matters | 12:14 |
| johnthetubaguy | sdague: BobBall: to the rename to .py seems like a good idea, so we get the pep8 stuff back | 12:14 |
| sdague | how are these installed? | 12:15 |
| johnthetubaguy | sdague: for us, using the RPM that we wripped out of tree | 12:15 |
| johnthetubaguy | as in it used to be in tree, but its not now | 12:15 |
| sdague | ok, but if the name is important it can be converted during that process right? | 12:15 |
| johnthetubaguy | but its basically xcopy install | 12:15 |
| johnthetubaguy | yeah, I mean we can fix it, I was worried about the other users really | 12:16 |
| tdurakov | johnthetubaguy, so we maintain only major release compatibility, say if someone have mitaka-1 compute and mitaka-2 compute I should not carry about its compatibility, right? | 12:16 |
| sdague | ok, I'm going to be mean for a second, are there really all that many other users? | 12:16 |
| *** drankis has joined #openstack-nova | 12:16 | |
| BobBall | the user survey says "not many" - but there are several | 12:17 |
| johnthetubaguy | tdurakov: oh dear, I was very unclear, we maintain compatibility between major releases (I was talking object versions). any commit in mitaka should upgrade to a later commit in mitaka, in a rolling upgrade way, and also any commit should be OK from the last major release (or at least latest stable) | 12:17 |
| sdague | it's kind of unfortunate that there isn't a real install tool for this in the nova tree | 12:18 |
| sdague | the devstack in line stuff is kind of gorpy | 12:18 |
| BobBall | The RPM was pulled out of tree a while ago because it was deemed the wrong place for it iirc | 12:18 |
| johnthetubaguy | yeah, we took it out, because we didn't like the RPM, I don't remember why | 12:18 |
| sdague | https://github.com/openstack-dev/devstack/blob/d3be1581814a77c7d9b0d423f0c5c38673f28b9d/lib/nova_plugins/hypervisor-xenserver#L73 | 12:18 |
| *** mylu has quit IRC | 12:18 | |
| sdague | right, I don't think an rpm like that should be in tree | 12:18 |
| BobBall | sdague: Indeed... | 12:19 |
| sdague | I'm just saying that the install process here is currently: bubble gum, duck tape, bailing twine, and prayers | 12:19 |
| pkoniszewski | does novaclient support microversions? | 12:20 |
| sdague | and given that, I'd honestly say that changing it in tree, pushing a reno that warns of the change should be fine | 12:20 |
| *** smatzek has joined #openstack-nova | 12:20 | |
| sdague | and, better, would be if nova had "install_xen_dom0" which is something everyone could call, and then we'd give a real interface | 12:21 |
| BobBall | Perhaps. I will look into that for sure. | 12:23 |
| tdurakov | andreykurilin, ^^^ | 12:23 |
| *** pradk_ has quit IRC | 12:24 | |
| BobBall | Got a meeting now. | 12:24 |
| *** cgalan has quit IRC | 12:26 | |
| johnthetubaguy | pkoniszewski: yes, some of them | 12:26 |
| *** tyagiprince has joined #openstack-nova | 12:27 | |
| johnthetubaguy | pkoniszewski: and example is here: https://github.com/openstack/python-novaclient/commit/6cbb22583b94660cfd78d8ee0068778d5279ceca | 12:28 |
| * johnthetubaguy goes for food | 12:28 | |
| pkoniszewski | johnthetubaguy: thanks! :) | 12:28 |
| *** doug-fish has joined #openstack-nova | 12:29 | |
| *** root has joined #openstack-nova | 12:30 | |
| *** bkumar4 has quit IRC | 12:30 | |
| *** root is now known as Guest63254 | 12:30 | |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Remove version hardcoding for 'create' and 'update' methods https://review.openstack.org/259097 | 12:32 |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Make Xenplugin to work with glance v2 api https://review.openstack.org/266933 | 12:32 |
| *** markvoelker has joined #openstack-nova | 12:33 | |
| *** ljxiash has joined #openstack-nova | 12:34 | |
| *** vladikr has joined #openstack-nova | 12:36 | |
| *** markvoelker has quit IRC | 12:37 | |
| *** pradk_ has joined #openstack-nova | 12:38 | |
| *** thorst_ has joined #openstack-nova | 12:38 | |
| *** ljxiash has quit IRC | 12:38 | |
| *** daemontool has quit IRC | 12:39 | |
| *** thorst_ has quit IRC | 12:39 | |
| *** daemontool has joined #openstack-nova | 12:39 | |
| *** dims has quit IRC | 12:40 | |
| openstackgerrit | sahid proposed openstack/nova: virt: reserved hugepages on compute host https://review.openstack.org/277422 | 12:40 |
| *** thorst_ has joined #openstack-nova | 12:41 | |
| *** amotoki has quit IRC | 12:41 | |
| *** ivase has joined #openstack-nova | 12:41 | |
| *** baoli has joined #openstack-nova | 12:47 | |
| *** sudipto has quit IRC | 12:48 | |
| *** baoli_ has joined #openstack-nova | 12:48 | |
| *** daemontool_ has joined #openstack-nova | 12:49 | |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Remove version hardcoding for 'create' and 'update' methods https://review.openstack.org/259097 | 12:49 |
| openstackgerrit | Mike Fedosin proposed openstack/nova: Make Xenplugin to work with glance v2 api https://review.openstack.org/266933 | 12:49 |
| *** daemontool__ has joined #openstack-nova | 12:49 | |
| *** daemontool has quit IRC | 12:51 | |
| *** baoli has quit IRC | 12:51 | |
| *** raildo-afk is now known as raildo | 12:53 | |
| *** daemontool_ has quit IRC | 12:54 | |
| *** jaypipes has joined #openstack-nova | 12:56 | |
| *** kylek3h has joined #openstack-nova | 12:57 | |
| *** su_zhang has joined #openstack-nova | 12:59 | |
| *** mfedosin has quit IRC | 13:00 | |
| *** haomaiwang has quit IRC | 13:01 | |
| *** bkumar4 has joined #openstack-nova | 13:01 | |
| *** haomaiwang has joined #openstack-nova | 13:01 | |
| *** bkumar4 has quit IRC | 13:01 | |
| *** bkumar4 has joined #openstack-nova | 13:02 | |
| *** su_zhang has quit IRC | 13:03 | |
| *** FL1SK has joined #openstack-nova | 13:05 | |
| tobasco | does anybody have a good insight into the live migration process? | 13:07 |
| tobasco | im using cinder root volumes with multipath and when i live migrate the multipath removal process fails on the source host causing a kpartx and blkid process to spawn and get stuck causing the next live migration to happen on that host to hang | 13:09 |
| *** amotoki has joined #openstack-nova | 13:09 | |
| *** delattec has joined #openstack-nova | 13:10 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova-specs: Moves Hyper-V Fibre Channel support spec to Newton https://review.openstack.org/279041 | 13:10 |
| openstackgerrit | Claudiu Belu proposed openstack/nova-specs: Moves Hyper-V Storage QOS spec to Newton https://review.openstack.org/279042 | 13:10 |
| openstackgerrit | Claudiu Belu proposed openstack/nova-specs: Moves the Hyper-V Cluster spec to Newton https://review.openstack.org/279043 | 13:10 |
| openstackgerrit | Claudiu Belu proposed openstack/nova-specs: Moves the Hyper-V NUMA instance spec to Newton https://review.openstack.org/279044 | 13:10 |
| openstackgerrit | Claudiu Belu proposed openstack/nova-specs: Moves the Hyper-V UEFI Secure Boot spec to Newton https://review.openstack.org/279045 | 13:10 |
| *** andreykurilin__ has joined #openstack-nova | 13:10 | |
| *** romainh has joined #openstack-nova | 13:10 | |
| *** bkumar4 has quit IRC | 13:10 | |
| *** ducttape_ has joined #openstack-nova | 13:11 | |
| *** nekrodesk has joined #openstack-nova | 13:11 | |
| andreykurilin__ | pkoniszewski: hi! novaclient supports most of the existing microversions(2.1-2.18 now) | 13:12 |
| andreykurilin__ | pkoniszewski: also, 2.19-2.21 are on review. | 13:12 |
| *** delatte has quit IRC | 13:13 | |
| *** romainh1 has joined #openstack-nova | 13:14 | |
| *** romainh has quit IRC | 13:14 | |
| *** mylu has joined #openstack-nova | 13:14 | |
| *** nekrodesk has joined #openstack-nova | 13:16 | |
| *** nekrodesk has joined #openstack-nova | 13:18 | |
| *** nekrodesk has quit IRC | 13:18 | |
| *** veena has joined #openstack-nova | 13:18 | |
| *** eharney has joined #openstack-nova | 13:18 | |
| *** tcammann has joined #openstack-nova | 13:18 | |
| *** mylu has quit IRC | 13:19 | |
| *** romainh1 has left #openstack-nova | 13:20 | |
| *** dims has joined #openstack-nova | 13:20 | |
| *** romainh1 has joined #openstack-nova | 13:20 | |
| andreykurilin__ | phoniszewski: If you want to write support for your 2.22 microversion, you should depends on the previous one(2.21 - https://review.openstack.org/#/c/278454/). info about implementation microversions in novaclient , you can found here - https://specs.openstack.org/openstack/nova-specs/specs/liberty/approved/novaclient-api-microversions.html + | 13:20 |
| andreykurilin__ | https://review.openstack.org/#/c/211206/ | 13:20 |
| *** nekrodesk has joined #openstack-nova | 13:20 | |
| *** romainh1 has quit IRC | 13:21 | |
| *** nekrodesk has joined #openstack-nova | 13:22 | |
| *** daemontool__ has quit IRC | 13:23 | |
| *** smatzek has quit IRC | 13:23 | |
| *** markvoelker has joined #openstack-nova | 13:23 | |
| andreykurilin__ | johnthetubaguy, sdague, alex_xu_ : can you look at the small change for novaclient's spec? https://review.openstack.org/#/c/211206/ | 13:23 |
| andreykurilin__ | please | 13:23 |
| *** tcammann has quit IRC | 13:23 | |
| *** Daisy has joined #openstack-nova | 13:24 | |
| *** xyang1 has joined #openstack-nova | 13:24 | |
| *** nekrodesk has joined #openstack-nova | 13:26 | |
| *** nekrodesk has quit IRC | 13:26 | |
| andrei__ | Hi all, Do someone know why VIR_MIGRATE_PERSIST_DEST is missing in default flags for live migration in nova.conf ? | 13:27 |
| *** ducttape_ has quit IRC | 13:27 | |
| *** brault has quit IRC | 13:27 | |
| *** nekrodesk has joined #openstack-nova | 13:30 | |
| johnthetubaguy | andrei__ unsure, danpb might know | 13:30 |
| johnthetubaguy | tobasco: PaulMurray is leading up the team currently focusing on live-migrate, there are lots of active bugs around volume issues, but not sure multipath has come up before | 13:31 |
| *** delatte has joined #openstack-nova | 13:32 | |
| *** nekrodesk has joined #openstack-nova | 13:33 | |
| *** tyagiprince has quit IRC | 13:33 | |
| *** delattec has quit IRC | 13:35 | |
| *** daemontool__ has joined #openstack-nova | 13:35 | |
| *** mubirru has quit IRC | 13:36 | |
| *** cdent has quit IRC | 13:37 | |
| pkoniszewski | andrei__: we reorganized live migration options a bit | 13:37 |
| *** burgerk has joined #openstack-nova | 13:37 | |
| pkoniszewski | andrei__: right now this flag is managed by nova automatically | 13:38 |
| pkoniszewski | andrei__: https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L702 | 13:38 |
| *** funeutron has joined #openstack-nova | 13:38 | |
| *** edmondsw has joined #openstack-nova | 13:38 | |
| *** lpetrut has quit IRC | 13:39 | |
| pkoniszewski | andrei__: you can also refer to reno: https://github.com/openstack/nova/blob/master/releasenotes/notes/libvirt-live-migration-flags-mangling-a2407a31ddf17427.yaml | 13:42 |
| *** daemontool_ has joined #openstack-nova | 13:42 | |
| andrei__ | Thank you! | 13:43 |
| *** brault has joined #openstack-nova | 13:44 | |
| danpb | andrei__: iirc nova will write out hte config itself, so its redundnat | 13:44 |
| *** ankit_ag has quit IRC | 13:45 | |
| *** daemontool__ has quit IRC | 13:46 | |
| *** pradk_ is now known as pradk | 13:46 | |
| *** smatzek has joined #openstack-nova | 13:47 | |
| *** diana_clarke has joined #openstack-nova | 13:48 | |
| *** links has quit IRC | 13:48 | |
| *** cgalan has joined #openstack-nova | 13:50 | |
| *** achanda has joined #openstack-nova | 13:50 | |
| *** mfedosin has joined #openstack-nova | 13:51 | |
| jaypipes | johnthetubaguy, sdague: thx for the review on generic-resource-pools. I'll push up a new rev shortgly. | 13:51 |
| openstackgerrit | Dongcan Ye proposed openstack/nova: VMware: Make http connection pool configurable https://review.openstack.org/265346 | 13:52 |
| *** cgalan has quit IRC | 13:53 | |
| tobasco | johnthetubaguy: thanks, i'm looking into this issue i have right now i also "signed up" for a ping when the next live-migration meeting is taking place | 13:55 |
| tobasco | hopefully i can get somewhere | 13:55 |
| *** achanda has quit IRC | 13:55 | |
| *** daemontool_ is now known as daemontool | 13:56 | |
| johnthetubaguy | tobasco: cool, it would be interesting if you get the same issues with cold migrate, its a similar pattern of calls, if not that might help narrow it down | 13:56 |
| *** jfding has quit IRC | 13:56 | |
| *** rlrossit has joined #openstack-nova | 13:57 | |
| alaski | johnthetubaguy: bauzas late start this morning, a bit under the weather, but I read scrollback and it seems there was some discussion and then some agreement? | 13:57 |
| bauzas | alaski: so we were discussing about the relationship between resource-* and cells v2 | 13:57 |
| alaski | also doffm has got a plan in his head on how this might all fit together and offerred to write it up | 13:57 |
| bauzas | and yeah I just commented that | 13:58 |
| bauzas | alaski: but maybe could you please review https://review.openstack.org/275432 ? | 13:58 |
| bauzas | I +2'd it because I was okay adding them in the cells db | 13:58 |
| alaski | okay, will do | 13:58 |
| bauzas | but your thoughts would be nivce | 13:59 |
| bauzas | nice* | 13:59 |
| *** garyk has quit IRC | 13:59 | |
| *** cdent has joined #openstack-nova | 13:59 | |
| bauzas | (and it's not late IIRC , just 9am ! ) | 13:59 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: vif_plug_ovs: Disable IPv6 on bridge devices https://review.openstack.org/275293 | 13:59 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: import openvswitch plugin implementation https://review.openstack.org/273610 | 13:59 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: import linux bridge plugin implementation https://review.openstack.org/273609 | 13:59 |
| bauzas | ah no | 13:59 |
| *** edmondsw has quit IRC | 13:59 | |
| johnthetubaguy | alaski: cool, sounds good. I can see how the current plan could work, and actually be a bit better that what I was first thinking we were going to do, so all good. Will look out for doffm's write up | 13:59 |
| alaski | bauzas: heh, I'm not usually as early as dansmith or sdague, but usually in earlier than this | 14:00 |
| bauzas | don't ask me when I'm coming :) | 14:00 |
| flip214 | meeting? | 14:00 |
| bauzas | right | 14:00 |
| johnthetubaguy | Nova meeting time in #openstack-meeting | 14:00 |
| johnthetubaguy | mikal tjones cburgess jgrimm adrian_otto funzo mjturek jcookekhugen irina_pov krtaylor danpb alexpilotti flip214 jaypipes garyk edleafe dims moshele anteaya Nisha sileht claudiub lxsli neiljerram markus_z swamireddy alevine tonyb andreykurilin ndipanov sc68cal akuriata artom jlvillal mnestratov kashyap aloga rgeragnov bauzas xyang tpatil med_ nic scottda nagyz belliott dguitarbite sdague jroll diana_clarke | 14:00 |
| johnthetubaguy | _diana_ raildo jichen gjayavelu mdorman klindgren sorrison belmoreira mrda auggy PaulMurray gcb thorst duncant macsz thomasem tpatzig bobball | 14:00 |
| johnthetubaguy | doing my cutting and pasting | 14:00 |
| *** haomaiwang has quit IRC | 14:01 | |
| *** haomaiwang has joined #openstack-nova | 14:01 | |
| alaski | johnthetubaguy: I'm not convinced it's better in the long run, but certainly better know for getting it done, and I don't think it blocks the work we may want to do later | 14:01 |
| *** ildikov has joined #openstack-nova | 14:02 | |
| *** jaosorior has quit IRC | 14:05 | |
| *** jaosorior has joined #openstack-nova | 14:05 | |
| thomasem | hello | 14:06 |
| *** dave-mccowan has joined #openstack-nova | 14:08 | |
| *** ljxiash has joined #openstack-nova | 14:09 | |
| PaulMurray | pkoniszewski, remember to update the implementation status of your blueprints | 14:10 |
| PaulMurray | pkoniszewski, eg: https://blueprints.launchpad.net/nova/+spec/block-live-migrate-with-attached-volumes | 14:10 |
| pkoniszewski | PaulMurray: sure, I will, still need python-novaclient change for pausing | 14:10 |
| pkoniszewski | PaulMurray: thanks for reminder! | 14:10 |
| PaulMurray | pkoniszewski, of course - I forgot about that | 14:11 |
| *** dave-mcc_ has joined #openstack-nova | 14:11 | |
| *** Daisy has quit IRC | 14:11 | |
| *** zz_dimtruck is now known as dimtruck | 14:11 | |
| openstackgerrit | Andrey Pavlov proposed openstack/nova: remove EC2 legacy code https://review.openstack.org/266425 | 14:11 |
| *** ctrath has joined #openstack-nova | 14:12 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova-specs: Fix bauzas Launchpad link in M priorities https://review.openstack.org/279091 | 14:12 |
| *** mdrabe has joined #openstack-nova | 14:13 | |
| *** ljxiash has quit IRC | 14:14 | |
| *** dave-mccowan has quit IRC | 14:14 | |
| openstackgerrit | Ludovic Beliveau proposed openstack/nova: Update binding:profile for SR-IOV ports https://review.openstack.org/242573 | 14:14 |
| flip214 | johnthetubaguy: thanks for getting around to me. | 14:15 |
| *** krotscheck_dcm is now known as krotscheck | 14:15 | |
| flip214 | There was a blueprint: https://review.openstack.org/#/c/134153/ | 14:15 |
| flip214 | but when os-brick got split out, and all that was left was the bits to connect Nova to os-brick, I thought (or got told? Can't remember) that no blueprint is necessary for that anymore. | 14:16 |
| *** mriedem has joined #openstack-nova | 14:17 | |
| *** ljxiash has joined #openstack-nova | 14:20 | |
| *** ccarmack has joined #openstack-nova | 14:22 | |
| *** Daisy has joined #openstack-nova | 14:23 | |
| *** su_zhang has joined #openstack-nova | 14:23 | |
| mdbooth | Are there any folks about who understand ploop? | 14:23 |
| *** ccarmack has quit IRC | 14:26 | |
| *** Daisy has quit IRC | 14:26 | |
| *** links has joined #openstack-nova | 14:27 | |
| johnthetubaguy | flip214: no spec is needed, you still need a blueprint | 14:28 |
| PaulMurray | mdbooth, try pinging mnestratov | 14:28 |
| johnthetubaguy | flip214: we changed that in kilo I think, relaxed the hard everything needs a spec rule | 14:28 |
| johnthetubaguy | flip214: its written up, sort of, here: http://docs.openstack.org/developer/nova/blueprints.html#specs | 14:29 |
| flip214 | johnthetubaguy: https://blueprints.launchpad.net/nova/+spec/drbd-driver << that one okay? | 14:29 |
| sdague | mriedem: ok, the 4 byte stuff is in my head. We once said it was ok for names, no one seems to have implemented it | 14:30 |
| johnthetubaguy | flip214: yes, thats fine, just needs submitting to be approved for a specific release, once we are open for new blueprints for newton | 14:30 |
| mriedem | sdague: ok | 14:30 |
| flip214 | johnthetubaguy: thank you very much. | 14:30 |
| flip214 | to make sure I understand: | 14:30 |
| johnthetubaguy | flip214: once we release RC1, basically bug me about what needs to be done | 14:31 |
| flip214 | * when newton is open, submit blueprint; submit code. | 14:31 |
| HenryG | johnthetubaguy: jaypipes: ping regarding get-me-a-network | 14:31 |
| johnthetubaguy | flip214: yeah, thats it | 14:31 |
| flip214 | johnthetubaguy: will do. thanks for the explanation. | 14:31 |
| johnthetubaguy | HenryG: ah, yeah | 14:32 |
| johnthetubaguy | HenryG: sorry, armax said you would be around, I totally forgot | 14:32 |
| HenryG | johnthetubaguy: I was in the meeting and saw the call for help fly by :) | 14:32 |
| johnthetubaguy | HenryG: ah, no worries, it looks like everything is ready for the Nova modification now | 14:33 |
| sdague | markus_z / auggy on - https://wiki.openstack.org/wiki/Nova/BugTriage#Weekly_bug_skimming_duty for the tag owners, can I suggest some changes? | 14:33 |
| HenryG | johnthetubaguy: yes | 14:33 |
| markus_z | sdague: sure, please | 14:33 |
| auggy | sdague: yes please, that list looks out of date to me | 14:33 |
| sdague | we should delete ec2 and docker, they aren't in tree any more and their owners aren't really around | 14:34 |
| sdague | baremetal as an official tag should go away | 14:34 |
| sdague | you can list me as api bug tag owner | 14:34 |
| HenryG | johnthetubaguy: please let whoever works on it know they can ping me for help from the neutron side | 14:35 |
| sdague | I'm not sure that wendar is really triaging at this point, I would probably remove her from the table | 14:35 |
| markus_z | sdague: ok, cool, I'll do that. Do you maybe know a contact for "network"? | 14:35 |
| *** scheuran has quit IRC | 14:35 | |
| mriedem | sdague: the postgres job was moved to experimental? | 14:36 |
| sdague | mriedem: no | 14:36 |
| mriedem | just v2.0? | 14:36 |
| sdague | yes | 14:36 |
| mriedem | k | 14:36 |
| *** liverpooler has quit IRC | 14:36 | |
| mriedem | markus_z: we don't really have any dedicated network people | 14:37 |
| sdague | markus_z: honestly, maybe dansmith | 14:37 |
| mriedem | for nova-net, dan probably knows the most | 14:37 |
| mriedem | neutronv2 api is no mans land | 14:37 |
| markus_z | ah, ok, I hoped I could spread that a little and pester different people. | 14:38 |
| johnthetubaguy | HenryG: cool, thank you | 14:38 |
| auggy | I thought it was the "postres" job.. had me wanting churros ;) | 14:39 |
| *** ociuhandu has joined #openstack-nova | 14:39 | |
| *** lpetrut has joined #openstack-nova | 14:41 | |
| * johnthetubaguy mmm, churros | 14:41 | |
| mriedem | HenryG: so i quickly skimmed through http://specs.openstack.org/openstack/neutron-specs/specs/liberty/get-me-a-network.html | 14:43 |
| mriedem | HenryG: but is there something that describes what the changes are in nova for get me a network? | 14:43 |
| HenryG | mriedem: I would talk to jaypipes - he went over it with armax at the mid-cycle | 14:44 |
| mriedem | GET jaypipes returns a 409 | 14:44 |
| jaypipes | HenryG: that work is currently 4th on my priority list, unfortunately, behind sending emails to the ML with updates on resource-providers framework, email to ML on os-vif progress, and code review of resource-providers patch series. :( | 14:44 |
| mriedem | this isn't going to happen for mitaka anyway (in nova) | 14:45 |
| mriedem | but i'm at least interested | 14:45 |
| *** knikolla has joined #openstack-nova | 14:48 | |
| mriedem | looks like basically if a specific network isn't provided at nova boot time, and the auto_allocate extension is enabled in neutron, nova can call that to auto-allocate a network/subnet/router and then nova uses that per-tenant network to boot into | 14:48 |
| mriedem | and nova will create the port | 14:48 |
| mriedem | HenryG: jaypipes: ^ is that the gist? | 14:48 |
| jaypipes | mriedem: yes | 14:49 |
| HenryG | mriedem: you got it | 14:49 |
| mriedem | cool, easy peasy | 14:49 |
| * mriedem was just looking at the API tests in https://review.openstack.org/#/c/277683/2/neutron/tests/api/admin/test_auto_allocated_topology.py | 14:49 | |
| HenryG | mriedem: the operator needs to provision a couple of things first, too | 14:49 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/278700 | 14:50 |
| jaypipes | HenryG: right, just like how nova-network works with pre-created networks | 14:50 |
| mriedem | HenryG: which is what's happening in resource_setup for that test right? | 14:50 |
| mriedem | # Ensure the public external network is the default external network | 14:50 |
| mriedem | # Ensure a shared default subnetpool for auto-allocated subnets | 14:50 |
| HenryG | mriedem: yep | 14:50 |
| mriedem | still, that's stuff that would need to be done for the neutron side to work, if that wasn't done, the call from nova to the auto-allocate API would just fail and nova would have to raise that up | 14:51 |
| HenryG | right | 14:51 |
| mriedem | i guess it would be nice if nova could tell that early in the boot process before we cast off to the compute service | 14:52 |
| mriedem | nova could at least validate that the auto_allocate extension is enabled from nova-api (if a network isn't provided) | 14:52 |
| mriedem | and fail fast | 14:52 |
| wendar | markus_z: confirmed, I haven't had time for triaging in a long while | 14:53 |
| mriedem | HenryG: how fast is client.get_auto_allocated_topology(tenant_id) ? | 14:53 |
| HenryG | mriedem: yes, and although neutron doesn't provide a "will-get-me-a-network-succeed", you can check for the default external net and default subnetpool. | 14:54 |
| markus_z | wendar: too bad. let me know when you find time again or have any questions. | 14:54 |
| *** mrkz has joined #openstack-nova | 14:54 | |
| wendar | markus_z: thanks :) | 14:54 |
| mriedem | HenryG: would be nice if neutron provided a readiness api of sorts for that so nova wouldn't need to know those details | 14:54 |
| HenryG | mriedem: I have not measured the timing. What is the concern? | 14:54 |
| *** erlon has joined #openstack-nova | 14:55 | |
| mriedem | HenryG: i was just thinking if nova-api could allocate the network before we cast off to the compute to build the vm | 14:55 |
| mriedem | or at least call that api and get a reservation id, something we can track, but would also let nova-api know early if auto-allocate is going to fail because the prereqs aren't setup | 14:56 |
| sdague | mriedem: I really don't like that idea, nova api doing a synchronous block on another API is a good way to tie up a ton of resources | 14:56 |
| mriedem | sdague: yeah i know | 14:56 |
| mriedem | sdague: but i'd like a way to know if things are at least setup so when we do call the neutron api, it's not going to just fail b/c the admin didn't setup the pre-req stuff | 14:57 |
| *** cfriesen has joined #openstack-nova | 14:57 | |
| *** esracelik has quit IRC | 14:57 | |
| *** diana_clarke has joined #openstack-nova | 14:57 | |
| mriedem | b/c if that's in the compute then networking fails and i think it triggers a rebuild retry right? | 14:57 |
| *** ccarmack has joined #openstack-nova | 14:57 | |
| mriedem | but we'll just exhaust the rebuild retries since neutron isn't setup properly yet and end up with NoValidHost | 14:57 |
| *** ducttape_ has joined #openstack-nova | 14:59 | |
| *** haomaiwang has quit IRC | 15:01 | |
| mriedem | anyway, i'm interested, but it's a newton thing and probably needs a nova spec anyway | 15:01 |
| *** haomaiwang has joined #openstack-nova | 15:01 | |
| ccesario | hello.... could someone please tell what is the best way to provide a complete instance backup (with all volumes attached) ? | 15:01 |
| *** burgerk has quit IRC | 15:01 | |
| *** eharney has quit IRC | 15:02 | |
| mriedem | sdague: you should probably set a priority on the bug for this if you're proposing it for stable/kilo https://review.openstack.org/#/c/278992/ | 15:04 |
| jaypipes | ccesario: please see /topic. That is best asked on openstack@ mailing list. thanks. | 15:04 |
| *** efoley has joined #openstack-nova | 15:04 | |
| efoley | nstack-meeting | 15:04 |
| *** jaosorior has quit IRC | 15:06 | |
| *** nekrodesk has quit IRC | 15:06 | |
| *** sc68cal has joined #openstack-nova | 15:07 | |
| *** brault has quit IRC | 15:07 | |
| alaski | mriedem: a compromise option would be to have the conductor check the network allocation before scheduling and error there | 15:08 |
| *** jwang_ has joined #openstack-nova | 15:08 | |
| *** gjayavelu has joined #openstack-nova | 15:09 | |
| ccesario | jaypipes, oops... sorry... whrong channel... it was to be on #openstack :) | 15:09 |
| jaypipes | ccesario: no worries :) | 15:11 |
| *** jwang has quit IRC | 15:12 | |
| *** Daisy has joined #openstack-nova | 15:16 | |
| *** Daisy has quit IRC | 15:17 | |
| *** mragupat has joined #openstack-nova | 15:18 | |
| *** omnipresent has quit IRC | 15:18 | |
| *** omnipresent has joined #openstack-nova | 15:19 | |
| *** ig0r_ has quit IRC | 15:21 | |
| anteaya | johnthetubaguy: if you get any volunteers for the nova side of get-me-a-network, I'm interested in knowing so I can help if need be | 15:21 |
| *** eharney has joined #openstack-nova | 15:21 | |
| *** pushkaru has joined #openstack-nova | 15:21 | |
| *** mfedosin has quit IRC | 15:24 | |
| *** omnipresent has quit IRC | 15:24 | |
| *** burgerk has joined #openstack-nova | 15:26 | |
| openstackgerrit | Andrea Rosa proposed openstack/nova: WIP: Abort an ongoing live migration https://review.openstack.org/277971 | 15:26 |
| *** veena has quit IRC | 15:27 | |
| dansmith | johnthetubaguy: alaski: did I miss some irc discussion about the resource providers db patch | 15:28 |
| dansmith | ? | 15:28 |
| alaski | dansmith: it was mostly johnthetubaguy getting caught up with previous discussions | 15:29 |
| dansmith | alaski: okay, meaning we're good? or ... at least, you're good? | 15:29 |
| johnthetubaguy | yeah, it was totally me catch up | 15:29 |
| *** angdraug has joined #openstack-nova | 15:29 | |
| alaski | dansmith: yep, nothing has changed | 15:29 |
| openstackgerrit | Martins Jakubovics proposed openstack/nova: Fix error message in imagebackend https://review.openstack.org/279140 | 15:29 |
| johnthetubaguy | yeah, that was my outstanding action | 15:30 |
| *** daemontool_ has joined #openstack-nova | 15:32 | |
| *** omnipresent has joined #openstack-nova | 15:32 | |
| johnthetubaguy | alaski: is that, you are happy, or equally unhappy? | 15:32 |
| dansmith | johnthetubaguy: glad you asked, I was confused :) | 15:33 |
| dansmith | figured it was the lack of caffeine | 15:33 |
| *** brault has joined #openstack-nova | 15:33 | |
| alaski | johnthetubaguy: heh, I am fine with the current direction. what would make me happy would totally derail things so I'm good with starting simple | 15:33 |
| alaski | I'm not unhappy | 15:34 |
| *** tyagiprince has joined #openstack-nova | 15:34 | |
| tyagiprince | Hii Everyone.. | 15:34 |
| tyagiprince | How do I pass the neutron metadata proxy? | 15:34 |
| *** daemontool has quit IRC | 15:34 | |
| *** achanda has joined #openstack-nova | 15:34 | |
| dansmith | alaski: this is a thing we've been talking about for like two cycles, I feel like we have to make some incremental progress or we'll never get it going | 15:34 |
| alaski | dansmith: right, I'm completely on board with that | 15:35 |
| dansmith | alaski: speaking of lots of talking and incremental progress, | 15:35 |
| *** zul_ has joined #openstack-nova | 15:36 | |
| dansmith | alaski: what two cells patches should I look at right this moment while I'm in a good mood? | 15:36 |
| *** zul has quit IRC | 15:36 | |
| openstackgerrit | Martins Jakubovics proposed openstack/nova: Fix error message in imagebackend https://review.openstack.org/279140 | 15:36 |
| *** omnipresent has quit IRC | 15:36 | |
| mriedem | alaski: dansmith: jaypipes: the resource-classes bp is a dependency for the resource-providers bp right? | 15:37 |
| alaski | dansmith: heh, you can pick any two out of my series at https://review.openstack.org/#/c/273641/, or melwitt/doffms work at https://review.openstack.org/#/c/270565/ | 15:37 |
| jaypipes | mriedem: yes. | 15:37 |
| *** dave-mcc_ has quit IRC | 15:37 | |
| dansmith | mriedem: jaypipes refuses to put them into a line so I don't know what depends on what :) | 15:37 |
| jaypipes | mriedem: resource-classes bp and code is already merged now, though. | 15:37 |
| tyagiprince | I am trying to set the correct entry in the openstack dns which should be the instance name. For that I made some changes to the code where dns entry is given. For that I made the hostname lookup for the instance through it ip. But my instance is not able to boot up. It is still booting with the 169.254.169.254 in the routing table | 15:37 |
| *** Daisy has joined #openstack-nova | 15:37 | |
| jaypipes | dansmith: sorry man :( | 15:37 |
| mriedem | jaypipes: yeah i was just going to make that relationship in launchpad | 15:37 |
| dansmith | jaypipes: *shakes fist* | 15:37 |
| *** dave-mccowan has joined #openstack-nova | 15:37 | |
| mriedem | jaypipes: since there are like 3 specs and i was trying to sort out their order | 15:37 |
| jaypipes | dansmith: I'm still working on this email to the ML about our progress and decisions around scheduler and RT | 15:38 |
| jaypipes | mriedem: in order: | 15:38 |
| jaypipes | * resource-classes (bp MERGED, code MERGED) | 15:38 |
| jaypipes | * pci-generate-stats (bp MERGED, code IN REVIEW) | 15:38 |
| jaypipes | * resource-providers (bp MERGED, code IN REVIEW) | 15:38 |
| jaypipes | * generic-resource-pools (bp IN REVIEW, code TODO) | 15:38 |
| jaypipes | * compute-node-inventory (bp IN REVIEW, code TODO) | 15:38 |
| jaypipes | * resource-providers-allocations (bp IN REVIEW, code TODO) | 15:38 |
| jaypipes | * resource-providers-scheduler (bp IN REVIEW, code TODO) | 15:38 |
| mriedem | jaypipes: your email isn't about suggesting we schedule things once per year in a different location is it? :P | 15:38 |
| jaypipes | mriedem: heh | 15:38 |
| dansmith | alaski: so I can choose the top patch on either of those sets? that sounds unhelpful :D | 15:39 |
| tyagiprince | help people :p | 15:39 |
| mriedem | tyagiprince: see the channel topic | 15:39 |
| doffm | dansmith: I'd like you to look at melwitt's db connection switching. https://review.openstack.org/#/c/161906/ | 15:40 |
| alaski | dansmith: I leave it up to your good mood to decide how helpful you would like to be :) I would recommend starting at the bottom though | 15:40 |
| *** mtanino has joined #openstack-nova | 15:40 | |
| dansmith | alaski: fine, fine | 15:40 |
| doffm | Although happy to have you look at my cell0 stuff also, the connection switching seems pretty core, and i'm going to depend on it for a couple of cell0 patches. | 15:40 |
| mriedem | jaypipes: cdent: if the resource-classes bp is complete, it should probably be marked that way in LP https://blueprints.launchpad.net/nova/+spec/resource-classes | 15:40 |
| dansmith | alaski: no tests? https://review.openstack.org/#/c/273641/3 | 15:41 |
| dansmith | alaski: there appear to be no tests in the tree either | 15:41 |
| jaypipes | mriedem: it was just merged yesterday... | 15:41 |
| jaypipes | mriedem: feel free to mark it. | 15:41 |
| *** Daisy has quit IRC | 15:41 | |
| mriedem | jaypipes: there are 3 patches in the bp whiteboard, they are all abandoned | 15:41 |
| mriedem | so i'm not sure which patch(es) it was | 15:41 |
| mriedem | https://review.openstack.org/#/c/273100/ ? | 15:41 |
| jaypipes | mriedem: ok, sorry, those patches are all the poc work. | 15:41 |
| alaski | dansmith: nope, probably should add those. I'll pretend I left it that way so you can -1 and improve your mood | 15:42 |
| jaypipes | mriedem: yup, that's it. | 15:42 |
| *** tyagiprince has left #openstack-nova | 15:42 | |
| mriedem | jaypipes: ok, i'll mark it complete | 15:42 |
| jaypipes | cheers, sorry about that | 15:42 |
| *** achanda has quit IRC | 15:42 | |
| *** fawadkhaliq has quit IRC | 15:42 | |
| *** fawadkhaliq has joined #openstack-nova | 15:43 | |
| dansmith | alaski: oooo, that does improve my mood | 15:43 |
| *** edmondsw has joined #openstack-nova | 15:43 | |
| *** mylu has joined #openstack-nova | 15:44 | |
| dansmith | alaski: I should probably hit that connection switching one | 15:46 |
| dansmith | I was kinda hoping the sun would burn out before I had to come back to it, but.. | 15:46 |
| mriedem | jaypipes: want to create a stub in LP for this? https://blueprints.launchpad.net/nova/+spec/compute-node-inventory | 15:46 |
| alaski | dansmith: heh, go for it. whenever your whimsy swings back around this way my series will be waiting | 15:46 |
| dansmith | alaski: poke me when you add a trivial test so I can keep my dignity and I'll hit that one again | 15:46 |
| jaypipes | mriedem: ya, sec | 15:46 |
| alaski | dansmith: will do | 15:47 |
| johnthetubaguy | dansmith: I was OK ish on the connection switching, my suggest might be a bit nit-picky, let me know | 15:47 |
| dansmith | johnthetubaguy: okay | 15:48 |
| jaypipes | mriedem: done: https://blueprints.launchpad.net/nova/+spec/compute-node-inventory | 15:48 |
| mriedem | where does https://blueprints.launchpad.net/nova/+spec/resource-objects fit into the dep tree? | 15:48 |
| mriedem | lp has resource-objects depending on pci-stats-generate | 15:48 |
| jaypipes | mriedem: that's essentially deprecated and made redundant by the new blueprints. | 15:49 |
| mriedem | blarg | 15:49 |
| *** jlanoux has quit IRC | 15:49 | |
| jaypipes | mriedem: it's marked as Superseded | 15:49 |
| bauzas | yeah, old story | 15:49 |
| mriedem | i'm sure it is | 15:49 |
| mriedem | an old story | 15:50 |
| knikolla | Hi guys, quick question related to nova/keystone interaction. In the attach_volume flow, does nova-api use the user credentials? | 15:50 |
| bauzas | mriedem: nah, it was just the old liberty (and mitaka) BP which was referring to all that - and what we asked to split | 15:51 |
| bauzas | so, the spec is still in-tree | 15:51 |
| *** andrei__ has quit IRC | 15:52 | |
| bauzas | http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/resource-objects.html | 15:52 |
| bauzas | maybe we should git rm it ? | 15:52 |
| bauzas | that said, it was approved | 15:52 |
| *** CustosLimen has quit IRC | 15:52 | |
| *** _gryf has quit IRC | 15:52 | |
| *** amotoki has quit IRC | 15:52 | |
| *** gjayavelu has quit IRC | 15:52 | |
| *** inara has quit IRC | 15:52 | |
| johnthetubaguy | bauzas: we haven't done that in the past, be we could I guess | 15:52 |
| dansmith | johnthetubaguy: let's go for it | 15:52 |
| dansmith | johnthetubaguy: I +2d | 15:52 |
| *** timello has quit IRC | 15:53 | |
| *** efoley has quit IRC | 15:53 | |
| mriedem | https://blueprints.launchpad.net/nova/+spec/compute-node-allocations doesn't exist either | 15:53 |
| *** EmilienM has quit IRC | 15:53 | |
| *** q3k has quit IRC | 15:53 | |
| *** NightKhaos has quit IRC | 15:54 | |
| mriedem | bauzas: i'd at least amend the history section of the spec saying it was completed in mitaka but also superseded by * other blueprints | 15:54 |
| *** pav0 has quit IRC | 15:54 | |
| bauzas | mriedem: fine by me | 15:54 |
| dansmith | alaski: do you have time to do that trivial test? if not, I will do it for you | 15:54 |
| mriedem | where * is resource-providers, resource-classes, whatever else | 15:54 |
| bauzas | mriedem: that said, we only partially approved | 15:54 |
| alaski | dansmith: I do, but not at this moment. about to run out for an appointment | 15:54 |
| alaski | but feel free | 15:54 |
| mriedem | bauzas: i wouldn't remove the spec since the bp was marked complete and there were patches merged against it | 15:55 |
| dansmith | alaski: I'll do it | 15:55 |
| dansmith | johnthetubaguy: are we waiting for melwitt on https://review.openstack.org/#/c/275432/ ? | 15:55 |
| bauzas | mriedem: mmm, really ? some patches landed ? | 15:55 |
| *** HenryG has quit IRC | 15:55 | |
| mriedem | bauzas: yeah a lot https://review.openstack.org/#/q/topic:bp/resource-objects | 15:55 |
| alaski | dansmith: I just +W'ed that | 15:55 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: Provide plugins an oslo_config group for their setup https://review.openstack.org/275281 | 15:55 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: vif_plug_ovs: Disable IPv6 on bridge devices https://review.openstack.org/275293 | 15:55 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: import openvswitch plugin implementation https://review.openstack.org/273610 | 15:55 |
| openstackgerrit | Daniel Berrange proposed openstack/os-vif: import linux bridge plugin implementation https://review.openstack.org/273609 | 15:55 |
| dansmith | alaski: I'm not sure where the top of that set is.. hint? | 15:55 |
| dansmith | alaski: woot, thanks | 15:56 |
| dansmith | alaski: is it https://review.openstack.org/#/c/263925/5 ? | 15:56 |
| bauzas | WAAAAT | 15:56 |
| johnthetubaguy | dansmith: yeah, I was just wanting to make sure the cells folks were happy, sounds like we are good | 15:56 |
| alaski | dansmith: https://review.openstack.org/#/c/263925/5 | 15:56 |
| dansmith | alaski: got it | 15:56 |
| alaski | dansmith: everything after that is just noise atm | 15:57 |
| dansmith | roger | 15:57 |
| bauzas | mriedem: okay, https://review.openstack.org/#/q/topic:bp/resource-objects+is:merged+project:openstack/nova shows me that it was unrelated to the BP | 15:57 |
| bauzas | for some reason, dan, paul and I were jinxed by jaypipes | 15:57 |
| dansmith | bauzas: eh? :) | 15:58 |
| bauzas | dansmith: see the url :) | 15:58 |
| bauzas | but I understand now | 15:58 |
| bauzas | all of these patches were easing how we model our resources | 15:59 |
| bauzas | so it was loosely related to the resource-objects BP | 15:59 |
| alaski | biab | 15:59 |
| *** achanda has joined #openstack-nova | 15:59 | |
| *** mriedem is now known as mriedem_meeting | 16:00 | |
| *** sridharg has quit IRC | 16:00 | |
| *** haomaiwang has quit IRC | 16:01 | |
| *** liusheng has quit IRC | 16:01 | |
| *** haomaiwa_ has joined #openstack-nova | 16:01 | |
| *** _gryf has joined #openstack-nova | 16:01 | |
| openstackgerrit | Matthew Booth proposed openstack/nova: Implement import_file for LVM backend https://review.openstack.org/278011 | 16:01 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Update import_file to provide a general-purpose interface https://review.openstack.org/267662 | 16:01 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Implement import_file for Qcow2 backend https://review.openstack.org/278010 | 16:01 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Remove fake_imagebackend.Raw and cleanup dependent tests https://review.openstack.org/267661 | 16:01 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Rename Image.check_image_exists to Image.exists() https://review.openstack.org/270998 | 16:01 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Make import_file abstract https://review.openstack.org/278013 | 16:01 |
| *** timello has joined #openstack-nova | 16:01 | |
| *** HenryG has joined #openstack-nova | 16:02 | |
| openstackgerrit | Matthew Booth proposed openstack/nova: Remove deprecated option libvirt.remove_unused_kernels https://review.openstack.org/265886 | 16:02 |
| *** vstinner has joined #openstack-nova | 16:02 | |
| openstackgerrit | Matthew Booth proposed openstack/nova: Only attempt to inject files if the injection disk exists https://review.openstack.org/250872 | 16:02 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Implement import_file for ploop backend https://review.openstack.org/278012 | 16:02 |
| *** andrei__ has joined #openstack-nova | 16:02 | |
| *** vstinner has left #openstack-nova | 16:02 | |
| *** achanda has quit IRC | 16:02 | |
| *** NightKhaos has joined #openstack-nova | 16:02 | |
| *** inara has joined #openstack-nova | 16:02 | |
| *** CustosLimen has joined #openstack-nova | 16:03 | |
| *** funeutron has quit IRC | 16:03 | |
| *** links has quit IRC | 16:04 | |
| *** pav0 has joined #openstack-nova | 16:04 | |
| *** efoley has joined #openstack-nova | 16:06 | |
| *** EmilienM has joined #openstack-nova | 16:06 | |
| openstackgerrit | Ludovic Beliveau proposed openstack/nova: Allow saving empty pci_device_pools in ComputeNode object https://review.openstack.org/216049 | 16:10 |
| doffm | alaski: Just writing up the cells resource pools proposal now. (discussed in yesterdays meeting) Will put out a link later today. | 16:10 |
| *** mjura has quit IRC | 16:11 | |
| *** mylu has quit IRC | 16:12 | |
| *** mylu has joined #openstack-nova | 16:12 | |
| *** haplo37 has joined #openstack-nova | 16:13 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Create BuildRequest object during boot process https://review.openstack.org/278125 | 16:14 |
| openstackgerrit | Dan Smith proposed openstack/nova: Add build_requests database table and model https://review.openstack.org/263926 | 16:14 |
| openstackgerrit | Dan Smith proposed openstack/nova: Add BuildRequest object https://review.openstack.org/278124 | 16:14 |
| openstackgerrit | Dan Smith proposed openstack/nova: Make InstanceMappings.cell_id nullable https://review.openstack.org/278598 | 16:14 |
| openstackgerrit | Dan Smith proposed openstack/nova: WIP Populate instance_mappings during boot. https://review.openstack.org/263925 | 16:14 |
| openstackgerrit | Dan Smith proposed openstack/nova: Change populate_security_groups to return a SecurityGroupList https://review.openstack.org/273641 | 16:14 |
| openstackgerrit | Dan Smith proposed openstack/nova: Improve unit tests for instance multiple create https://review.openstack.org/276882 | 16:14 |
| tdurakov | danpb, hi, are you around? | 16:15 |
| *** mylu_ has joined #openstack-nova | 16:17 | |
| *** mylu has quit IRC | 16:17 | |
| tdurakov | dansmith hey, want to discuss cancelation of stuck live-migrations | 16:18 |
| openstackgerrit | Roey Chen proposed openstack/nova: Don't specify security-groups on an instance when using Neutron https://review.openstack.org/278369 | 16:19 |
| dansmith | tdurakov: are you asking _if_ I want to? :D | 16:20 |
| *** q3k has joined #openstack-nova | 16:20 | |
| tdurakov | dansmith, :) I'm proposing, if you have no time, could talk later | 16:21 |
| *** bnemec has quit IRC | 16:21 | |
| *** browne has joined #openstack-nova | 16:21 | |
| *** drankis has quit IRC | 16:21 | |
| dansmith | tdurakov: meh, I'm probably not the best person to talk about it with.. what do you want to discuss specifically? | 16:22 |
| *** burt has joined #openstack-nova | 16:22 | |
| dansmith | tdurakov: danpb was around earlier, so I expect he'll be back | 16:22 |
| danpb | i'm around but fairly busy | 16:23 |
| tdurakov | here is the problem, in case of block-migration it's possible to end up with canceled live-migration, because as I understood disk is transfered first | 16:23 |
| tdurakov | and it looks like live-migration is stucked | 16:23 |
| tdurakov | dansmith^ | 16:23 |
| danpb | hmm, don't we account for disk + memory when looking if it is stuck | 16:24 |
| *** mvk has quit IRC | 16:24 | |
| tdurakov | danpb, data_total is memory + disk? | 16:25 |
| *** jlanoux has joined #openstack-nova | 16:25 | |
| danpb | yes, data rfers tothe combination of both | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Implement import_file for LVM backend https://review.openstack.org/278011 | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Update import_file to provide a general-purpose interface https://review.openstack.org/267662 | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Implement import_file for Qcow2 backend https://review.openstack.org/278010 | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Remove fake_imagebackend.Raw and cleanup dependent tests https://review.openstack.org/267661 | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Rename Image.check_image_exists to Image.exists() https://review.openstack.org/270998 | 16:26 |
| tdurakov | danpb, well, than code is valid | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Make import_file abstract https://review.openstack.org/278013 | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Remove deprecated option libvirt.remove_unused_kernels https://review.openstack.org/265886 | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Only attempt to inject files if the injection disk exists https://review.openstack.org/250872 | 16:26 |
| openstackgerrit | Matthew Booth proposed openstack/nova: Implement import_file for ploop backend https://review.openstack.org/278012 | 16:26 |
| *** thumpba has joined #openstack-nova | 16:26 | |
| tdurakov | but at the same time I got block-migration canceled, while there was data transfer | 16:27 |
| tdurakov | danpb, ^ | 16:27 |
| danpb | well the logs should show all the data values indicating why it cancelled | 16:28 |
| danpb | can you pastebin the nova compute log for the migration | 16:28 |
| tobasco | johnthetubaguy: seems like the live-migration issue did not happen when i use cold migrations ("Migrate") or atleast not after testing like 10-15 migrations | 16:28 |
| tobasco | johnthetubaguy: updated the launchpad bug with some more info | 16:29 |
| tdurakov | danpb, sure | 16:29 |
| johnthetubaguy | tobasco: OK, good info, certainly looks like its worth comparing the Nova -> Cinder API call to work out whats causing your problem | 16:29 |
| tdurakov | there were periodic messages like: Migration running for 135 secs, memory 100% remaining; (bytes processed=0, remaining=0, total=0) | 16:29 |
| *** jdurgin1 has joined #openstack-nova | 16:30 | |
| danpb | oh, that's misleading of us - we should be printing the full data size, not just memory | 16:30 |
| danpb | since we're doing calculations based on full data | 16:31 |
| openstackgerrit | John Garbutt proposed openstack/nova: Improve 'virt.driver' conf options documentation https://review.openstack.org/266364 | 16:31 |
| tobasco | tdurakov: you should get an exception when the migration reaches the max timeout | 16:31 |
| tdurakov | danpb, tobasco http://paste.openstack.org/show/486725/ | 16:31 |
| tdurakov | btw, nova instance-action-list doesn't mark migration as failed | 16:32 |
| *** Swami has joined #openstack-nova | 16:32 | |
| danpb | the "stuck" messages indicate that it isn't making forwards progress overall | 16:32 |
| tdurakov | in this case | 16:32 |
| danpb | unfortunately we'll need to fix the log message to print out full data rather than just memory, to give you a clearer idea of what happened | 16:33 |
| tobasco | johnthetubaguy: don't think it's related to the cinder api afaik multipathing is handled by nova and the os-brick iscsi integration? not sure tho | 16:33 |
| *** browne has quit IRC | 16:33 | |
| danpb | certainyl seems like it could not keep up with dirtying of disk blocks | 16:33 |
| tdurakov | danpb, on the same env, I managed to block-migrate instance with smaller flavor | 16:33 |
| tdurakov | danpb it was kind of synthetic test, so no ops on instance, or smth that could dirty disky blocks | 16:35 |
| tdurakov | s/disky/disk | 16:35 |
| *** rk4n_ has quit IRC | 16:35 | |
| tdurakov | I'll edit log message and try again, to get more info | 16:36 |
| *** mylu has joined #openstack-nova | 16:36 | |
| tobasco | kashyap: can you have a look at https://bugs.launchpad.net/nova/+bug/1525802 when you have time, i know you have a lot of insight into libvirt and i'm pretty curious on my libvirt migration flags as i posted in my latest comment, but the two most recent includes a lot of info moving the bug forward | 16:36 |
| openstack | Launchpad bug 1525802 in OpenStack Compute (nova) "live migration with multipath cinder volumes crashes node" [Undecided,New] | 16:36 |
| kashyap | tobasco: Heh, "lot of insight", don't get tricked too easily :-) | 16:37 |
| kashyap | tobasco: Will look in 10 mins, in the middle of something | 16:37 |
| *** Vek has quit IRC | 16:37 | |
| bauzas | just in case a spec-core could want to hit me https://review.openstack.org/#/c/279091/ | 16:37 |
| *** nikhil_k has joined #openstack-nova | 16:38 | |
| *** ganso has quit IRC | 16:38 | |
| tobasco | kashyap: hehe, well you beat me by some lengths atleast ;) thanks | 16:38 |
| *** drankis has joined #openstack-nova | 16:38 | |
| *** eliqiao has quit IRC | 16:38 | |
| *** wendar_ has joined #openstack-nova | 16:38 | |
| *** mylu_ has quit IRC | 16:38 | |
| *** zhenguo_ has quit IRC | 16:38 | |
| *** andrewbogott has quit IRC | 16:38 | |
| *** junjie has quit IRC | 16:38 | |
| *** tpeoples has quit IRC | 16:38 | |
| *** comstud has quit IRC | 16:38 | |
| tobasco | kashyap: might be leaving in a while so if gone just shoot a comment and i'll check it out asap and get back | 16:39 |
| *** wasmum has joined #openstack-nova | 16:39 | |
| *** rook has quit IRC | 16:39 | |
| *** betherly has quit IRC | 16:39 | |
| *** Mingyu has quit IRC | 16:39 | |
| *** toan has quit IRC | 16:39 | |
| *** wendar has quit IRC | 16:39 | |
| *** nikhil has quit IRC | 16:39 | |
| kashyap | tobasco: Okay, quick note - remove this VIR_MIGRATE_TUNNELLED when doing live block migration | 16:39 |
| *** toan has joined #openstack-nova | 16:40 | |
| *** ganso has joined #openstack-nova | 16:40 | |
| *** krtaylor has quit IRC | 16:40 | |
| kashyap | tobasco: Also a note for future releases of Nova: Both the flags 'live_migration_flag' and 'block_migration_flag' are now deprecated | 16:40 |
| *** armax has joined #openstack-nova | 16:41 | |
| *** salv-orl_ has joined #openstack-nova | 16:41 | |
| kashyap | Mandatory flags are handled by default - https://review.openstack.org/#/c/263433/ | 16:41 |
| *** betherly has joined #openstack-nova | 16:41 | |
| *** unicell1 has joined #openstack-nova | 16:42 | |
| *** junjie has joined #openstack-nova | 16:42 | |
| kashyap | Instead the tunnelling will be handled (by upcoming versions) with a single flag. (See here for more info: https://review.openstack.org/#/c/263434/) | 16:42 |
| *** scheuran has joined #openstack-nova | 16:42 | |
| *** eliqiao has joined #openstack-nova | 16:43 | |
| *** unicell has quit IRC | 16:43 | |
| *** tpeoples has joined #openstack-nova | 16:43 | |
| *** andrewbogott has joined #openstack-nova | 16:44 | |
| *** Mingyu has joined #openstack-nova | 16:44 | |
| *** jichen has quit IRC | 16:44 | |
| *** comstud has joined #openstack-nova | 16:44 | |
| *** zhenguo_ has joined #openstack-nova | 16:44 | |
| *** salv-orlando has quit IRC | 16:44 | |
| tobasco | kashyap: ok thanks, i will try the changes, seems like the issue with the bug is because of multipath cleanup so was not really related but wanted some good input on that config :) | 16:44 |
| *** browne has joined #openstack-nova | 16:44 | |
| tobasco | kashyap: so if i understand it correctly in the future all needed flags would be determined by nova(?) and since I don't need tunneling I would just remove the live_migration_flag config | 16:45 |
| *** whenry__ has joined #openstack-nova | 16:45 | |
| *** whenry__ has quit IRC | 16:45 | |
| *** Marga_ has quit IRC | 16:46 | |
| *** mgoddard_ has joined #openstack-nova | 16:46 | |
| *** Marga_ has joined #openstack-nova | 16:46 | |
| kashyap | tobasco: Just remove the VIR_MIGRATE_TUNNELLED option from block_migration_flag. (There are no multiple changes (you say "the changes"), as you're using Liberty or below & not Git (I assume), just that one flag I mentioned.) | 16:46 |
| kashyap | tobasco: Yes, on your last point. | 16:47 |
| *** scheuran has quit IRC | 16:47 | |
| tobasco | kashyap: Cool thanks | 16:48 |
| *** mgoddard has quit IRC | 16:49 | |
| *** nikhil_k is now known as nikhil | 16:50 | |
| *** r-daneel has joined #openstack-nova | 16:52 | |
| hogepodge | jaypipes: HenryG: wait, I heard it was third, now fourth? | 16:52 |
| *** zakora has quit IRC | 16:52 | |
| *** amotoki has joined #openstack-nova | 16:52 | |
| *** mriedem_meeting is now known as mriedem | 16:52 | |
| *** paul-carlton1 has quit IRC | 16:53 | |
| *** mdrabe has quit IRC | 16:54 | |
| *** rcernin has quit IRC | 16:56 | |
| *** bnemec has joined #openstack-nova | 16:56 | |
| *** nekrodesk has joined #openstack-nova | 16:57 | |
| *** krtaylor has joined #openstack-nova | 16:57 | |
| *** vishwanathj has joined #openstack-nova | 16:57 | |
| *** daemontool_ has quit IRC | 16:57 | |
| *** amotoki has quit IRC | 16:58 | |
| *** nelsnels_ has quit IRC | 16:59 | |
| *** nelsnelson has joined #openstack-nova | 17:00 | |
| *** nekrodesk has joined #openstack-nova | 17:00 | |
| *** nekrodesk has quit IRC | 17:00 | |
| *** haomaiwa_ has quit IRC | 17:01 | |
| *** haomaiwang has joined #openstack-nova | 17:01 | |
| tdurakov | danpb, something strange, http://xsnippet.org/361415/ need to go now | 17:02 |
| *** whenry has joined #openstack-nova | 17:04 | |
| *** jwcroppe has joined #openstack-nova | 17:06 | |
| *** pushkaru has quit IRC | 17:06 | |
| *** ndipanov has quit IRC | 17:06 | |
| *** mdrabe has joined #openstack-nova | 17:09 | |
| *** lucasagomes is now known as lucas-dinner | 17:10 | |
| *** derekh has quit IRC | 17:12 | |
| *** ociuhandu has quit IRC | 17:12 | |
| *** paul-carlton1 has joined #openstack-nova | 17:13 | |
| *** sfinucan has quit IRC | 17:13 | |
| *** e0ne has quit IRC | 17:14 | |
| *** penick has joined #openstack-nova | 17:15 | |
| *** nekrodesk has joined #openstack-nova | 17:15 | |
| *** apoorvad has joined #openstack-nova | 17:17 | |
| tobasco | johnthetubaguy: i think there is a race condition between the scsi devices being removed and the multipath being flushed but i cannot confirm it yet, just been scrolling through some code and compare to logs | 17:19 |
| *** nekrodesk has joined #openstack-nova | 17:19 | |
| *** nkrinner has quit IRC | 17:19 | |
| *** rk4n has joined #openstack-nova | 17:21 | |
| *** pushkaru has joined #openstack-nova | 17:24 | |
| *** mpavone has quit IRC | 17:27 | |
| *** ociuhandu has joined #openstack-nova | 17:33 | |
| *** fawadkhaliq has quit IRC | 17:33 | |
| *** unicell1 has quit IRC | 17:33 | |
| *** paul-carlton1 has quit IRC | 17:34 | |
| *** nekrodesk has joined #openstack-nova | 17:37 | |
| *** jlanoux has quit IRC | 17:38 | |
| *** jdurgin1 has quit IRC | 17:43 | |
| *** rdopiera has quit IRC | 17:45 | |
| *** danpb has quit IRC | 17:47 | |
| *** paul-carlton1 has joined #openstack-nova | 17:48 | |
| *** nic has joined #openstack-nova | 17:49 | |
| *** penick has quit IRC | 17:50 | |
| *** moshele has quit IRC | 17:50 | |
| *** Guest15678 is now known as mariusv | 17:51 | |
| *** mariusv has quit IRC | 17:51 | |
| *** mariusv has joined #openstack-nova | 17:51 | |
| *** mgoddard_ has quit IRC | 17:51 | |
| *** mgoddard has joined #openstack-nova | 17:51 | |
| *** sahid has quit IRC | 17:52 | |
| dansmith | ugh | 17:53 |
| dansmith | tests failed at position 2 in the gate | 17:53 |
| * cdent hands dansmith a tissue | 17:53 | |
| *** knikolla has quit IRC | 17:54 | |
| *** ssurana has joined #openstack-nova | 17:55 | |
| cdent | dansmith: in the world of nova.objects do we usually call object.delete() or Class.delete(some identifier) if we want to destroy something? | 17:56 |
| *** fawadkhaliq has joined #openstack-nova | 17:56 | |
| dansmith | cdent: it's conventionally destroy, and conventionally on an object not a class | 17:58 |
| dansmith | cdent: if you have the identifier, it's trivial: | 17:58 |
| cdent | cool, thanks | 17:58 |
| dansmith | instance = objects.Instance(id=1) | 17:58 |
| dansmith | instance.destroy() | 17:58 |
| dansmith | you don't have to look it up first | 17:58 |
| * cdent nods | 17:58 | |
| *** ljxiash has quit IRC | 18:00 | |
| *** haomaiwang has quit IRC | 18:01 | |
| *** haomaiwang has joined #openstack-nova | 18:01 | |
| mriedem | dansmith: notice a resemblance? https://review.openstack.org/#/c/252059/ | 18:03 |
| sdague | oh, hey, on those engine facade patches - https://bugs.launchpad.net/nova/+bug/1531930 | 18:03 |
| openstack | Launchpad bug 1531930 in OpenStack Compute (nova) "SQLalchemy API crashes executing migration_get_unconfirmed_by_dest_compute" [Undecided,New] | 18:04 |
| sdague | apparently unreleased oslo.db code was referenced | 18:04 |
| dansmith | mriedem: doh, sorry! | 18:04 |
| mriedem | np | 18:04 |
| mriedem | hans pointed it out | 18:04 |
| mriedem | i mean on mine before i wasted time rebasing | 18:04 |
| dims | sdague : mriedem : need a oslo.db release or a g-r bump? | 18:05 |
| sdague | dims: I think both | 18:05 |
| melwitt | sdague: ah man | 18:05 |
| sdague | how did that get past | 18:06 |
| *** unicell has joined #openstack-nova | 18:06 | |
| mriedem | sdague: depends-on? | 18:06 |
| *** browne has quit IRC | 18:07 | |
| sdague | no | 18:07 |
| sdague | it doesn't work that way | 18:07 |
| dims | y, depends-on does not kick in unless LIBS_FROM_GIT is also there | 18:08 |
| mriedem | if it's unreleased how are we not completely blocked... | 18:08 |
| dims | mriedem : want to log a release request? | 18:08 |
| sdague | mriedem: right, that's what I'm trying to figure out | 18:08 |
| melwitt | I guess that code isn't hit during testing? I don't know | 18:08 |
| dims | melwitt : y, possible | 18:08 |
| mriedem | sdague: periodic task | 18:08 |
| sdague | mriedem: ah... | 18:09 |
| melwitt | I've seen that allow_async around for a long time and had no idea it was unreleased, just assumed it was because it's been used a lot already | 18:09 |
| sdague | right, those are basically completely untested | 18:09 |
| mriedem | @periodic_task.periodic_task | 18:09 |
| mriedem | def _poll_unconfirmed_resizes(self, context): | 18:09 |
| mriedem | yar | 18:09 |
| dansmith | :( | 18:10 |
| melwitt | usually when I see a commit patch that says "use this thing from oslo.*" I do a manual check on requirements range and feature release point | 18:10 |
| mriedem | and that runs like every 10 minutes or something | 18:10 |
| mriedem | dims: i can put up the release request quick yeah | 18:10 |
| melwitt | *commit message | 18:10 |
| sdague | hmmm this landed https://review.openstack.org/#/c/249235/ | 18:10 |
| sdague | that was a while ago | 18:10 |
| mriedem | in since 4.1.0 | 18:10 |
| dims | mriedem : y, i want to avoid log a review and then work on it as well :) | 18:10 |
| sdague | ok, so this person has just a totally lying environment | 18:11 |
| dims | i am generating what's not released yet | 18:11 |
| mriedem | sdague: so the bug reporter probably doesn't have new enough oslo.db | 18:11 |
| *** ljxiash has joined #openstack-nova | 18:11 | |
| dims | whew | 18:11 |
| sdague | the bug reporter specificially said they had 4.2.0 | 18:11 |
| melwitt | oh | 18:11 |
| mriedem | g-r reqiures at least 4.1.0 https://github.com/openstack/requirements/blob/master/global-requirements.txt#L98 | 18:11 |
| mriedem | so that's why i'm not finding hits on that in logstash | 18:12 |
| dims | mriedem : sdague : there's nothing left to release in oslo.db (one requirements change) | 18:12 |
| sdague | mriedem: yeh, ok, I'll mark it up to bad bug reporter | 18:12 |
| mriedem | i invalidated the bug | 18:12 |
| sdague | it was detailed enough that I assumed they knew what they were talking about | 18:12 |
| sdague | mriedem: I did too | 18:13 |
| * mriedem goes back to cafeteria salad | 18:13 | |
| *** aginwala has joined #openstack-nova | 18:14 | |
| dansmith | heh | 18:14 |
| *** ljxiash has quit IRC | 18:15 | |
| *** e0ne has joined #openstack-nova | 18:15 | |
| *** rcernin has joined #openstack-nova | 18:17 | |
| mriedem | doffm: this curly wurly is not what i expected | 18:17 |
| mriedem | and now i've bitten my tongue. damn you hilariously named british candy! | 18:18 |
| *** baoli_ has quit IRC | 18:20 | |
| *** jwcroppe_ has joined #openstack-nova | 18:21 | |
| *** jwcroppe_ has quit IRC | 18:22 | |
| *** rmart04 has joined #openstack-nova | 18:22 | |
| *** jwcroppe_ has joined #openstack-nova | 18:22 | |
| *** mriedem has quit IRC | 18:22 | |
| *** jwcroppe has quit IRC | 18:23 | |
| sdague | oh, right, testr py34 db format bug | 18:24 |
| sdague | that's 30 minutes I'm not getting back trying to figure out why this test didn't run | 18:24 |
| *** mriedem has joined #openstack-nova | 18:24 | |
| mriedem | reminds me of the time that manfred mann got his anus curly wurly'ed and asked me if i needed a ride | 18:25 |
| cburgess_ | What? | 18:25 |
| dansmith | mriedem: +1 | 18:25 |
| mriedem | cburgess_: go listen to any classic rock FM station | 18:26 |
| dansmith | for at least two hours and you'll hear it | 18:26 |
| openstackgerrit | Tracy Jones proposed openstack/nova: config options: centralize section "volume" https://review.openstack.org/253747 | 18:26 |
| *** jwcroppe_ has quit IRC | 18:26 | |
| cburgess_ | I'm so lost | 18:26 |
| dansmith | http://www.amiright.com/misheard/artist/manfredmannsearthband.shtml | 18:27 |
| *** rmart04 has quit IRC | 18:27 | |
| mriedem | https://www.youtube.com/watch?v=lcWVL4B-4pI | 18:27 |
| cburgess_ | Yeah OK I've heard that song and I've never heard that.. wow... | 18:27 |
| *** eil397 has joined #openstack-nova | 18:28 | |
| cburgess_ | Y'all are messed up | 18:28 |
| mriedem | i don't say "y'all" so who's messed up now? | 18:28 |
| *** ihrachys has quit IRC | 18:28 | |
| alaski | that's a perfectly cromulent word | 18:29 |
| mriedem | don't go out tonight, there is a bathroom on the right | 18:29 |
| cburgess_ | I will give you that y'all is one of the bad things I still have from growing up in the south. | 18:29 |
| auggy | y'all is a perfectly valid plural you form | 18:30 |
| mriedem | i need an all white pant suit with no sleeves | 18:30 |
| cburgess_ | I would pay to see that. | 18:30 |
| *** cdent has quit IRC | 18:30 | |
| *** Jeffrey4l has quit IRC | 18:30 | |
| auggy | the fact that it is one syllable makes it more efficient than "you all" | 18:30 |
| *** nkrinner has joined #openstack-nova | 18:31 | |
| auggy | the one that drives me crazy is "yunz" | 18:31 |
| mriedem | cburgess_: i have a red leisure suit hand-me-down, i could wear that in austin, but the hipsters would probably not even notice it's out of place | 18:31 |
| auggy | i never heard that until i went to a friends wedding in new hampshire | 18:31 |
| cburgess_ | Hmm... | 18:31 |
| cburgess_ | That might be worth a fiver. | 18:32 |
| auggy | mriedem: and roller skates | 18:32 |
| *** PaulMurray has quit IRC | 18:32 | |
| mriedem | no way, never trusted those death traps | 18:32 |
| sdague | oh, that's neat | 18:32 |
| sdague | virtualenv-14.0.6 breaks being able to run tox locally | 18:33 |
| *** ctrath has quit IRC | 18:33 | |
| *** zhenguo_ has quit IRC | 18:33 | |
| edleafe | auggy: in Texas, y'all is singular. They use "all y'all" as the plural. | 18:34 |
| mriedem | as in, f all y'all | 18:34 |
| mriedem | the redundancy is endearing | 18:34 |
| dansmith | edleafe: y'all is usable in the singular, just like "they" | 18:34 |
| JayF | I always used all y'all as plural plural | 18:34 |
| *** mdrabe has quit IRC | 18:34 | |
| JayF | like y'all can be multiple people | 18:34 |
| JayF | all y'all is multiple groups of people | 18:35 |
| dansmith | JayF: yeah, all y'all is usually lots of people | 18:35 |
| *** mtanino__ has quit IRC | 18:35 | |
| edleafe | I have no idea what youse are talking about | 18:35 |
| *** vladikr has quit IRC | 18:35 | |
| JayF | that may be NC-dialect of y'all though :P | 18:35 |
| *** Kevin_Zheng has quit IRC | 18:35 | |
| *** mtanino__ has joined #openstack-nova | 18:35 | |
| *** zhenguo_ has joined #openstack-nova | 18:35 | |
| *** ctrath has joined #openstack-nova | 18:36 | |
| alaski | JayF: that was my understanding of all y'all before I came to NC, so not specific to here | 18:36 |
| *** su_zhang has quit IRC | 18:36 | |
| dansmith | alaski: VA is just NNC | 18:36 |
| alaski | dansmith: hah | 18:36 |
| JayF | dansmith: only parts of VA that are country. Not NoVA | 18:36 |
| *** baoli has joined #openstack-nova | 18:37 | |
| sc68cal | dansmith: hey, dumb question - but for live upgrades tag, did Nova move to doing multinode for grenade, and remove the single node job? | 18:37 |
| dansmith | JayF: right, part of VA is actually DC I think.. it's a messed up state, basically | 18:37 |
| dansmith | sc68cal: well, not for the tag, but yeah we did | 18:37 |
| *** Kevin_Zheng has joined #openstack-nova | 18:37 | |
| JayF | dansmith: heh, there's a reason I live on the west coast now :P | 18:37 |
| dansmith | JayF: +1 | 18:37 |
| auggy | JayF: +1 | 18:37 |
| sc68cal | dansmith: cool, that's what I thought. armax and I were talking about multinode grenade for neutron and plans for the future | 18:37 |
| dansmith | sc68cal: cool | 18:38 |
| *** jlvillal has quit IRC | 18:38 | |
| *** gszasz has quit IRC | 18:38 | |
| *** moshele has joined #openstack-nova | 18:39 | |
| *** angdraug has quit IRC | 18:39 | |
| *** jlvillal has joined #openstack-nova | 18:39 | |
| openstackgerrit | Sean Dague proposed openstack/nova: Fix reported ppc64le bug on video selection https://review.openstack.org/279227 | 18:40 |
| mtreinish | sdague: ^^^ that'll take weeks to wash off :) | 18:41 |
| sdague | mtreinish: yeh, well, it means one less bug in the tracker | 18:41 |
| dansmith | GROSS | 18:41 |
| *** Vek has joined #openstack-nova | 18:41 | |
| edleafe | dansmith: show us on this doll where the power chip touched you | 18:42 |
| dansmith | lol | 18:42 |
| * dansmith rips open the doll's head and points to the insides | 18:43 | |
| edleafe | dansmith: lol | 18:43 |
| *** browne has joined #openstack-nova | 18:44 | |
| *** ccarmack has quit IRC | 18:45 | |
| *** penick has joined #openstack-nova | 18:46 | |
| *** knikolla has joined #openstack-nova | 18:46 | |
| lbeliveau | looking for reviews on this quick bug fix (migration state): https://review.openstack.org/#/c/277451 | 18:46 |
| *** kbringard has joined #openstack-nova | 18:47 | |
| *** ccarmack has joined #openstack-nova | 18:48 | |
| *** mdrabe has joined #openstack-nova | 18:48 | |
| *** vladikr has joined #openstack-nova | 18:49 | |
| *** gtt116__ has joined #openstack-nova | 18:49 | |
| *** eliqiao_ has quit IRC | 18:50 | |
| *** gjayavelu has joined #openstack-nova | 18:50 | |
| doffm | mriedem: You prefered the toffee crisp? Good to know. | 18:50 |
| *** paul-carlton1 has quit IRC | 18:50 | |
| *** gtt116_ has quit IRC | 18:51 | |
| *** eliqiao_ has joined #openstack-nova | 18:52 | |
| *** daemontool has joined #openstack-nova | 18:54 | |
| *** gjayavelu has quit IRC | 18:54 | |
| *** amotoki has joined #openstack-nova | 18:55 | |
| *** vilobhmm11 has joined #openstack-nova | 18:55 | |
| mriedem | sdague: co-authored-by to xiaojinwei001@163.com ? | 18:56 |
| *** mpavone has joined #openstack-nova | 18:58 | |
| *** ig0r_ has joined #openstack-nova | 18:58 | |
| *** amotoki has quit IRC | 18:59 | |
| *** mpavone has quit IRC | 19:00 | |
| sdague | mriedem: for 1 line? | 19:00 |
| *** vishwana_ has joined #openstack-nova | 19:00 | |
| sdague | if you really want it, sure | 19:00 |
| dansmith | we get a CAB for one line? | 19:00 |
| *** haomaiwang has quit IRC | 19:01 | |
| dansmith | I should go claim a bunch of credit... | 19:01 |
| *** haomaiwang has joined #openstack-nova | 19:01 | |
| *** lpetrut has quit IRC | 19:01 | |
| mriedem | i just figured it it's a patch in a bug report we were supposed to do that, | 19:02 |
| mriedem | i didn't -1 for it though | 19:02 |
| mriedem | more of a question | 19:02 |
| openstackgerrit | Sean Dague proposed openstack/nova: Fix reported ppc64le bug on video selection https://review.openstack.org/279227 | 19:02 |
| *** moshele has quit IRC | 19:02 | |
| sdague | mriedem: there you go | 19:02 |
| dansmith | is there something going on with the py34 job? | 19:03 |
| *** ig0r_ has quit IRC | 19:03 | |
| mriedem | it's racy | 19:03 |
| sdague | the mox bits, or something else? | 19:03 |
| *** vishwanathj has quit IRC | 19:03 | |
| *** harlowja has joined #openstack-nova | 19:04 | |
| dansmith | this is dict/set/object ordering that I just hit | 19:04 |
| dansmith | there are a lot of failing nova patches in the gate | 19:04 |
| dansmith | but not all are py34 | 19:04 |
| mriedem | is py34 in the gate? | 19:05 |
| mriedem | i thought it was moved to the check queue | 19:05 |
| sdague | yes | 19:05 |
| sdague | we never did that change | 19:05 |
| sdague | new upstream releasE? | 19:06 |
| dansmith | of whaT? | 19:06 |
| *** su_zhang has joined #openstack-nova | 19:07 | |
| sdague | dansmith: it's all object representation issues? | 19:08 |
| dansmith | sdague: it's just a buggy object comparison utility I guess | 19:08 |
| dansmith | haven't seen it fail in like forever | 19:08 |
| *** yamahata has quit IRC | 19:08 | |
| *** ig0r_ has joined #openstack-nova | 19:08 | |
| sdague | {'nova_object.changes': ['cpuset', 'memory', 'id'] vs. {'nova_object.changes': ['cpuset', 'id', 'memory' | 19:08 |
| dansmith | actually, | 19:08 |
| dansmith | that's just checking the json'd versions to see if they're the same | 19:09 |
| dansmith | which won't be reliable | 19:09 |
| sdague | all the other nova patches failing in gate are different testing | 19:10 |
| sdague | that's just a bad test | 19:10 |
| dansmith | it's been in there since 2014 | 19:10 |
| sdague | which is hash seed randomizing to fail from time to time | 19:10 |
| dansmith | can't say I've ever noticed it fail before | 19:10 |
| dansmith | right | 19:10 |
| sdague | right, well, still a bad test :) | 19:10 |
| dansmith | yep, yep | 19:10 |
| sdague | might be nice if objects sorted that field | 19:11 |
| sdague | just to make everyone's life a little simpler | 19:11 |
| dansmith | it's a set | 19:11 |
| dansmith | which json doesn't have | 19:11 |
| dansmith | sorting every message we send over the wire so our unit tests pass would be pretty crazy :) | 19:12 |
| sdague | right, I get that, but keeping the json representation stable seems worthwhile | 19:12 |
| dansmith | I'll fix up the test | 19:12 |
| *** fawadkhaliq has quit IRC | 19:12 | |
| *** nelsnels_ has joined #openstack-nova | 19:12 | |
| mriedem | i feel like this conversation has happened before | 19:12 |
| *** w_verdugo has quit IRC | 19:14 | |
| mriedem | timello: fyi https://review.openstack.org/#/c/279227/ | 19:14 |
| *** w_verdugo has joined #openstack-nova | 19:15 | |
| *** nelsnelson has quit IRC | 19:15 | |
| *** spandhe has joined #openstack-nova | 19:16 | |
| sdague | glance went kind of drunk - http://logs.openstack.org/56/257956/23/gate/gate-tempest-dsvm-full/fcf0b7f//logs/screen-g-api.txt.gz#_2016-02-11_18_34_22_007 | 19:16 |
| sdague | that's why things are failing in the gate for one of those | 19:17 |
| *** cdent has joined #openstack-nova | 19:17 | |
| *** yassine has quit IRC | 19:18 | |
| mriedem | sdague: weird, that doesn't show up in logstash | 19:20 |
| *** ccarmack has quit IRC | 19:20 | |
| *** nekrodesk has quit IRC | 19:22 | |
| *** Sukhdev has joined #openstack-nova | 19:23 | |
| *** ccarmack has joined #openstack-nova | 19:24 | |
| *** rlrossit has quit IRC | 19:24 | |
| *** ijw has joined #openstack-nova | 19:24 | |
| *** ccarmack has quit IRC | 19:26 | |
| *** penick has quit IRC | 19:26 | |
| *** boris-42 has joined #openstack-nova | 19:26 | |
| *** rlrossit_ has joined #openstack-nova | 19:28 | |
| *** funeutron has joined #openstack-nova | 19:29 | |
| *** lpetrut has joined #openstack-nova | 19:30 | |
| timello | mriedem, thanks! | 19:30 |
| *** ccarmack has joined #openstack-nova | 19:31 | |
| *** markus_z has quit IRC | 19:32 | |
| *** ctrath has quit IRC | 19:32 | |
| *** toure is now known as toure|biab | 19:32 | |
| *** ctrath has joined #openstack-nova | 19:33 | |
| *** ctrath has quit IRC | 19:33 | |
| *** rook has joined #openstack-nova | 19:35 | |
| *** penick has joined #openstack-nova | 19:36 | |
| *** markusz has joined #openstack-nova | 19:38 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Add StableObjectJsonFixture and use it in our base test class https://review.openstack.org/279242 | 19:40 |
| dansmith | mriedem: sdague ^ | 19:41 |
| dansmith | if you like that I can sync to o.vo and we can start using it in 1.5 constraint-years or whatever | 19:41 |
| *** mylu has quit IRC | 19:42 | |
| ivase | hi all looking at https://blueprints.launchpad.net/nova/liberty wondering is there blueprint describing overall changes in 2.1 i see some blueprint talk for the api relax validation or for some policy enforcement in 2.1 but could not find general blueprint/doc describing delta changes between v2 and v2.1 | 19:43 |
| doffm | mriedem: You know the issue you found with the Flavor API db move? (link to the flavor through the instance.... can never delete flavor record) | 19:44 |
| *** mylu has joined #openstack-nova | 19:44 | |
| doffm | mriedem: Do you think the same issue might be found elsewhere? I'm thinking specifically of os-keypairs and perhaps security groups. | 19:44 |
| mriedem | the spec? | 19:44 |
| openstackgerrit | Ludovic Beliveau proposed openstack/nova: Update binding:profile for SR-IOV ports https://review.openstack.org/242573 | 19:44 |
| doffm | mriedem: This spec. https://review.openstack.org/#/c/265282/ | 19:45 |
| doffm | I just wonder if the same problem could be found for some other objects not just flavors. | 19:45 |
| *** drankis has quit IRC | 19:46 | |
| *** nekrodesk has joined #openstack-nova | 19:46 | |
| mriedem | doffm: what do you mean by "can never delete flavor record" | 19:47 |
| mriedem | ? | 19:47 |
| * dansmith resists the urge to be sarcastic | 19:47 | |
| doffm | dansmith: Go right ahead. Don't leave us hanging. | 19:49 |
| mriedem | doffm: you mean if that spec is implemented to not return the flavor id as part of the server response body, and the subresource doesn't have the flavor id, we can't then delete the flavor | 19:49 |
| mriedem | the admin can always delete a flavor record... | 19:50 |
| mriedem | i think the point was, instance show has a flavor 99 which has been deleted, and we can do a flavor show on 99 b/c it reads deleted flavors | 19:51 |
| mriedem | until they are hard-deleted | 19:51 |
| doffm | Yeah, that. | 19:51 |
| *** moshele has joined #openstack-nova | 19:51 | |
| *** gjayavelu has joined #openstack-nova | 19:51 | |
| mriedem | yeah, so the issue is if you can read deleted on any other instance-related things via their apis, not the server api | 19:52 |
| doffm | Ok, I guess I'll look out for that. | 19:53 |
| *** Sukhdev has quit IRC | 19:54 | |
| mriedem | so i guess you'd look for things that the server response shows (id's) to things that can be shown in the api even though they are deleted | 19:54 |
| *** Sukhdev has joined #openstack-nova | 19:54 | |
| mriedem | doffm: i think image is the other obvious one | 19:55 |
| *** nekrodesk has joined #openstack-nova | 19:55 | |
| *** dimtruck is now known as zz_dimtruck | 19:55 | |
| *** zz_dimtruck is now known as dimtruck | 19:55 | |
| mriedem | doffm: but images are a bit difference since those are in glance, | 19:55 |
| *** eliqiao_ has quit IRC | 19:56 | |
| *** amotoki has joined #openstack-nova | 19:56 | |
| mriedem | we don't 'read' deleted images from the nova db since they aren't in the nova db, we'd just get a 404 from glance | 19:56 |
| mriedem | and pass that buck | 19:56 |
| *** aginwala has quit IRC | 19:56 | |
| *** Swami has quit IRC | 19:57 | |
| *** eliqiao_ has joined #openstack-nova | 19:58 | |
| *** mylu has quit IRC | 19:59 | |
| *** penick has quit IRC | 19:59 | |
| *** ociuhandu has quit IRC | 19:59 | |
| *** moshele has quit IRC | 19:59 | |
| *** drankis has joined #openstack-nova | 20:00 | |
| erlon | guys, quick question, Im booting some VMs from images/creating volumes, but my hypervisor is limiting the number of VMs based on the local disk consumed. Is there a way/config to change that? | 20:00 |
| *** amotoki has quit IRC | 20:00 | |
| *** haomaiwang has quit IRC | 20:01 | |
| *** haomaiwa_ has joined #openstack-nova | 20:01 | |
| erlon | I can boot 21VMs of 10GB disk, the compute host has 211GB, but after the creation it still has +-200 free space | 20:01 |
| erlon | as expected | 20:02 |
| *** fawadkhaliq has joined #openstack-nova | 20:02 | |
| *** ljxiash has joined #openstack-nova | 20:03 | |
| *** baoli has quit IRC | 20:03 | |
| *** chhavi has quit IRC | 20:04 | |
| *** unicell has quit IRC | 20:04 | |
| *** aginwala has joined #openstack-nova | 20:04 | |
| *** unicell has joined #openstack-nova | 20:04 | |
| mriedem | erlon: this isn't really the right channel, see topic, but i think you're looking for disk_allocation_ratio in http://docs.openstack.org/liberty/config-reference/content/section_compute-scheduler.html | 20:04 |
| *** ccarmack has quit IRC | 20:04 | |
| erlon | mriedem: sorry, the errors are happening in nova logs, but thank you!! | 20:05 |
| *** aysyd has quit IRC | 20:06 | |
| sdague | file-injection is a thing we want to deprecate right - https://bugs.launchpad.net/nova/+bug/1491216 ? | 20:08 |
| openstack | Launchpad bug 1491216 in OpenStack Compute (nova) "Inject an ssh key failed when booting a instance by using direct image injection." [Undecided,New] - Assigned to Chung Chih, Hung (lyanchih) | 20:08 |
| sdague | given the security issues | 20:08 |
| *** Bjoern has joined #openstack-nova | 20:08 | |
| *** kbringard has quit IRC | 20:08 | |
| alaski | it would be deprecate it, but it's still used quite a bit from what I've seen | 20:09 |
| alaski | woops, it would be great to deprecate it | 20:09 |
| *** ctrath has joined #openstack-nova | 20:10 | |
| sdague | yeh, it seems like it's also got bugs and might not be working atm | 20:10 |
| alaski | I can't vouch for it working for libvirt, if it stops working for xen I think we'll hear about it | 20:13 |
| *** toure|biab is now known as toure | 20:16 | |
| *** kbringard has joined #openstack-nova | 20:16 | |
| alaski | doffm: the eager load on the db relationship didn't work, from https://review.openstack.org/#/c/278124/2/nova/objects/build_request.py | 20:16 |
| doffm | alaski: :( Hmm. I'll take a look. | 20:18 |
| *** zenoway has joined #openstack-nova | 20:19 | |
| *** su_zhang has quit IRC | 20:19 | |
| alaski | I pinged zzzeek in -dev to see if he has insight as well | 20:19 |
| doffm | alaski: Thats a good idea :) | 20:20 |
| mriedem | sdague: surprise, you modifed the db schema in https://review.openstack.org/#/c/279227/ :) | 20:21 |
| *** gjayavelu has quit IRC | 20:21 | |
| *** andreykurilin__ has quit IRC | 20:22 | |
| sdague | mriedem: there must be an australian around somewhere we can lock up for that | 20:23 |
| mriedem | ha | 20:23 |
| sdague | turbo hipster go home, you are stoned | 20:23 |
| mriedem | they are all enjoying clotted cream atm and reading about the queen, try later | 20:23 |
| *** ctrath has quit IRC | 20:24 | |
| cdent | wow | 20:24 |
| *** rcernin has quit IRC | 20:24 | |
| *** penick has joined #openstack-nova | 20:24 | |
| sdague | also, the logs for that test run don't say anything about such a fail | 20:25 |
| sdague | mikal: you must be waking up shortly | 20:27 |
| *** rmart04 has joined #openstack-nova | 20:28 | |
| jaypipes | cdent: what? | 20:28 |
| openstackgerrit | Dan Smith proposed openstack/nova: Add StableObjectJsonFixture and use it in our base test class https://review.openstack.org/279242 | 20:28 |
| *** slagle has quit IRC | 20:28 | |
| cdent | jaypipes: I was wowing at mriedem | 20:28 |
| jaypipes | oh :) | 20:28 |
| cdent | now I'm wowing that your scheduler email just now showed up | 20:29 |
| jaypipes | cdent: heh | 20:29 |
| *** rdopiera has joined #openstack-nova | 20:29 | |
| cdent | I'm going to wait to read it though so that I am not distracted from glorious Allocation objects | 20:29 |
| cdent | or rather, not _more_ distracted | 20:29 |
| *** slagle has joined #openstack-nova | 20:29 | |
| *** ctrath has joined #openstack-nova | 20:30 | |
| *** fawadkhaliq has quit IRC | 20:30 | |
| *** rmart04 has quit IRC | 20:32 | |
| *** funeutron has quit IRC | 20:32 | |
| *** Swami has joined #openstack-nova | 20:33 | |
| *** omnipresent has joined #openstack-nova | 20:35 | |
| *** zenoway has quit IRC | 20:38 | |
| dansmith | cdent: allocation is newton stuff right? | 20:39 |
| cdent | dansmith: I'm working to the spec. The spec says make Allocation objects :) | 20:40 |
| dansmith | ah | 20:40 |
| *** aginwala has quit IRC | 20:40 | |
| dansmith | how's the object patch atop the inventory bit going? should I bang on some more testing for that? | 20:40 |
| *** aginwala has joined #openstack-nova | 20:40 | |
| cdent | that's where I am write now | 20:41 |
| cdent | I'll be committing soonish and releasing the lock for the night if you wanna have a gander | 20:41 |
| dansmith | okay | 20:41 |
| *** efoley has quit IRC | 20:43 | |
| anteaya | mriedem: I looked at the clotted cream and couldn't get any further | 20:44 |
| anteaya | dansmith's rant on it was enough to convince me I'd made a good choice | 20:44 |
| tonyb | mmmm clotted cream | 20:44 |
| anteaya | ewwwww | 20:44 |
| dansmith | I still have naughty dreams | 20:44 |
| anteaya | I don't want to know | 20:44 |
| *** omnipresent has quit IRC | 20:44 | |
| dansmith | anteaya: wait, I ranted about its goodness, right? | 20:45 |
| anteaya | mriedem: can I trouble you for a requirements patch review for tomorrow's gerrit downtime? https://review.openstack.org/#/c/279257/1 | 20:45 |
| anteaya | dansmith: I don't know the point of your rant, but it involved clotted cream | 20:45 |
| anteaya | I'm learning to avoid things you have strong opinions about | 20:45 |
| dansmith | anteaya: it was probably "omg this is so good" or "omg I want to smear this all over my body" or ... related things | 20:46 |
| anteaya | that could be | 20:46 |
| sdague | dansmith: tmi tmi TMI | 20:46 |
| anteaya | large gestures and pointing at clotted cream was my takeaway | 20:46 |
| jaypipes | alaski: is there a "default cell" in cellsv2? | 20:46 |
| anteaya | sdague: thank you | 20:46 |
| * cdent lives in the land of clotted cream | 20:46 | |
| cdent | Sadly I hate the stuff. | 20:46 |
| dansmith | jaypipes: default cell? what would that mean? | 20:47 |
| alaski | jaypipes: not sure what that would mean either | 20:47 |
| dansmith | alaski: maybe he's asking for "if tenant_id == "jaypipes": cell = cell0" for all public clouds | 20:47 |
| jaypipes | dansmith, alaski: if there is only a single cell in a deployment... | 20:47 |
| dansmith | jaypipes: well, then everything goes there, | 20:48 |
| dansmith | but I wouldn't call it a default | 20:48 |
| *** zenoway has joined #openstack-nova | 20:48 | |
| dansmith | it's like having one compute node | 20:48 |
| alaski | right, the only special cell is cell0 | 20:48 |
| dansmith | exactly like that. | 20:48 |
| dansmith | and it's not really a cell, | 20:48 |
| alaski | yeah | 20:48 |
| dansmith | it's a wasteland of unstarted instances | 20:48 |
| * dansmith considers making a rude analogy | 20:48 | |
| sdague | i thought it was going to be called the zombie cell | 20:48 |
| jaypipes | dansmith, alaski: what I'm really asking is can I make the --cell-uuid parameter to nova resource-pool-create an optional parameter if there is only a single cell in the deployment. | 20:48 |
| dansmith | no | 20:49 |
| jaypipes | ok, nuf said :) | 20:49 |
| dansmith | that would be like saying "can I forget which compute node I booted a thing on, if I only have one" | 20:49 |
| alaski | jaypipes: as dansmith said. there's no way to pick a cell without something to key off of | 20:49 |
| *** su_zhang has joined #openstack-nova | 20:50 | |
| anteaya | mriedem: thank you | 20:50 |
| mriedem | yw | 20:50 |
| mikal | sdague: turbo hispter is angry? | 20:50 |
| dansmith | angry .. DRUNK | 20:50 |
| mikal | Oh nice | 20:50 |
| openstackgerrit | Augustina Ragwitz proposed openstack/nova: Fix _compare_result type handling comparison https://review.openstack.org/278191 | 20:50 |
| mikal | How come its drunk and not me? | 20:51 |
| mikal | jhesketh: you around yet? | 20:51 |
| dansmith | because it's morning? | 20:51 |
| sdague | it believes adding a constant related to ppc vga consoles has changed the mysql schema | 20:51 |
| bauzas | jaypipes: thanks for your email | 20:51 |
| mikal | IT HAS | 20:51 |
| sdague | https://review.openstack.org/#/c/279227/ | 20:51 |
| mikal | Review link? | 20:51 |
| mikal | Ta | 20:51 |
| bauzas | I just feel we need somehow to discuss about the real not yet consensus on the scheduler claims | 20:51 |
| bauzas | not now of course | 20:52 |
| *** rlrossit_ has quit IRC | 20:52 | |
| mikal | sdague: is this the only th wierdness? Or is it being weird to all reviews? | 20:52 |
| *** ctrath has quit IRC | 20:52 | |
| jaypipes | bauzas: I thought we already agreed to just agree with me and that will make everyone happy. | 20:52 |
| bauzas | :D | 20:52 |
| *** daemontool_ has joined #openstack-nova | 20:52 | |
| sdague | mikal: I do not know, mriedem pointed this one out | 20:52 |
| jaypipes | :P | 20:52 |
| dansmith | jaypipes: only if your agreements are underwritten my my agreement | 20:53 |
| mriedem | mikal: that's the only one i've seen | 20:53 |
| *** e0ne has quit IRC | 20:53 | |
| * mikal reads logs | 20:53 | |
| bauzas | jaypipes: so maybe an hangout could be nice tomorrow ? | 20:53 |
| bauzas | it's just awfully late for me tonight :( | 20:54 |
| jaypipes | bauzas: I can do after 10:30am EST | 20:54 |
| cdent | can you invite me too, for lurking? | 20:54 |
| jaypipes | of course | 20:54 |
| *** smatzek has quit IRC | 20:54 | |
| bauzas | jaypipes: yeah 11am EST is okay | 20:55 |
| bauzas | I just have a 10-11am meeting | 20:55 |
| mikal | sdague: I don't see any real errors in the logs, I might just do a recheck and see what happens | 20:55 |
| *** daemontool has quit IRC | 20:55 | |
| *** Sukhdev has quit IRC | 20:56 | |
| *** ctrath has joined #openstack-nova | 20:56 | |
| *** ctrath has quit IRC | 20:56 | |
| *** su_zhang has quit IRC | 20:56 | |
| jaypipes | alaski, dansmith: I guess I'm a little uncomfortable having --cell-id in the API here. My reasoning is that I always assumed cells were an implementation detail and not to be exposed in a REST API. Remember that the reosurce-pool-create API is going to be a REST API in the broken out scheduler. Adding a cell ID to the request payload will mean Nova's scaling implementation details are leaking out of the public REST API | 20:56 |
| jaypipes | for the scheduler. | 20:56 |
| bauzas | tbc, I'm not sure we could yet have an agreement, but I'd like to reviezw your opinions for the week-edn | 20:56 |
| bauzas | :) | 20:56 |
| dansmith | jaypipes: I'm probably missing context, | 20:56 |
| dansmith | but I agree it shouldn't leak out the user side of the api | 20:57 |
| *** ociuhandu has joined #openstack-nova | 20:57 | |
| alaski | jaypipes: I agree, but by putting those tables into the cell db this is a necessary evil | 20:57 |
| alaski | dansmith: my comment on https://review.openstack.org/#/c/253187/ | 20:57 |
| jaypipes | alaski: and what about when we have generic resource pools for things that are in Neutron (for example the IPv4 subnet pools)? Neutron doesn't have cells, so how do you envision reconciling this? | 20:58 |
| dansmith | ah, this is why I said "user side of the api" | 20:58 |
| bauzas | well, I'm seeing the cell as an admin API resource | 20:58 |
| dansmith | bauzas: exactly | 20:58 |
| alaski | jaypipes: the conversation we had a few days ago went in the direction of needing to slice subnet pools by cell | 20:59 |
| *** ccarmack has joined #openstack-nova | 20:59 | |
| dansmith | jaypipes: configuring neutron to report this will require neutron to have some idea of a lot of nova-ness, | 20:59 |
| dansmith | jaypipes: but you'll configure a segment to report to a given cell | 20:59 |
| alaski | jaypipes: deployers, or some tool, will need to decide how many to put in each cell and rebalance as necessary | 20:59 |
| dansmith | jaypipes: which I would expect to be a metadata-like key on a segment or something | 20:59 |
| jaypipes | dansmith, alaski: one option we could consider is the following: create a UUID column on the aggregates table. Create a cell_to_aggregate mapping table in the API database. Then, when a resource pool is associated with an aggregate, the cell ID can be determined from the cell_to_aggregate mapping table. | 21:00 |
| *** baoli has joined #openstack-nova | 21:00 | |
| dansmith | I think that probably makes sense | 21:00 |
| dansmith | although, | 21:00 |
| *** haomaiwa_ has quit IRC | 21:01 | |
| dansmith | we can just do it with metadata on an aggregate I think, | 21:01 |
| bauzas | isn't what doffm has to write an ascii diagram ? :p | 21:01 |
| dansmith | because you'd want an aggregate to be able to span cells I think, right? | 21:01 |
| jaypipes | dansmith: as opposed to a uuid? | 21:01 |
| alaski | I think that's what doffm had in mind as well | 21:01 |
| dansmith | jaypipes: meaning not make it a primary property of an aggregate | 21:01 |
| bauzas | heh, jinxing alaski | 21:01 |
| *** haomaiwang has joined #openstack-nova | 21:01 | |
| *** notmorgan is now known as morganfainberg | 21:02 | |
| *** aginwala has quit IRC | 21:02 | |
| *** drankis has quit IRC | 21:02 | |
| bauzas | so, having an aggregate per resource pool (and then per cell) is okay by me | 21:02 |
| bauzas | that doesn't mean it stops the aggregates spanning cells | 21:03 |
| *** aginwala has joined #openstack-nova | 21:03 | |
| dansmith | well, | 21:03 |
| dansmith | if you have a cell-uuid property of aggregate, | 21:03 |
| mriedem | jaypipes honey, i have a question about the image api stuff in nova | 21:03 |
| bauzas | ah that no | 21:04 |
| dansmith | then it does mean an aggregate is pinned to one, unless you make it nullable in which case it means any cell | 21:04 |
| bauzas | sorry I was not clear | 21:04 |
| mikal | Have I failed at calendars, or are we meant to be having a nova meeting now? | 21:04 |
| bauzas | I'm okay with a 1:N relationship, not a 1:1 | 21:04 |
| dansmith | bauzas: yeah, same | 21:04 |
| bauzas | so a indirection table looks fine by me | 21:04 |
| alaski | mikal: you have failed | 21:04 |
| mikal | alaski: please explain the nature of my failure? | 21:05 |
| bauzas | 1400UTC today | 21:05 |
| dansmith | mikal: total | 21:05 |
| alaski | mikal: early meeting today | 21:05 |
| bauzas | well, yesterday for mikal | 21:05 |
| mikal | Herm, that's not what the repeating event in calendar says. Did we swap weeks somewhere or something? | 21:05 |
| bauzas | yeah | 21:05 |
| alaski | there were two odd weeks in a row at the end of the year | 21:06 |
| mikal | Oh, that will have messed me up | 21:06 |
| bauzas | mikal: http://lists.openstack.org/pipermail/openstack-dev/2016-January/083216.html | 21:06 |
| *** yamahata has joined #openstack-nova | 21:06 | |
| * mikal wonders if that will have broken tonyb's calendaring thingie | 21:06 | |
| auggy | mikal: not sure what calendar client you use, but I just set the meeting location to Reykjavik so it converts to whatever timezone I'm in | 21:06 |
| mikal | Given he's calendar-core | 21:06 |
| bauzas | auggy: or use a correct calendar which knows UTC | 21:07 |
| mikal | auggy: no, the problem is we have alternating times, but tweaked where we were in the alternation sequence | 21:07 |
| mriedem | where/how does nova.image.s3 ever get used? | 21:07 |
| bauzas | auggy: but TZs are unrelated to mikal's problem | 21:07 |
| auggy | Ha yeah just realized that after I posted;) | 21:07 |
| tonyb | mikal: yes and no. I knew it was a thing. I haven't propgated the change out to all the private calendars which is probably an issue .. meh | 21:08 |
| *** zenoway has quit IRC | 21:08 | |
| mikal | tonyb: please fix my calendar | 21:08 |
| mikal | tonyb: cause its performance review season | 21:08 |
| * auggy is on phone so typing slowlyyy | 21:08 | |
| tonyb | mikal: "other duties as required" *sigh* | 21:08 |
| tonyb | I wont point out that the change happened 6 weeks ago and mikal is only just workign it out ;P | 21:09 |
| *** raildo is now known as raildo-afk | 21:09 | |
| mriedem | andreykurilin: do you know anything about nova.image.s3? | 21:09 |
| mikal | tonyb: yes well | 21:09 |
| mikal | tonyb: I blame Christmas / mid-cycles / lca | 21:09 |
| *** su_zhang has joined #openstack-nova | 21:09 | |
| mikal | tonyb: or worse yet I might lash out in a fit of management rage | 21:09 |
| mikal | tonyb: so careful now | 21:09 |
| mikal | tonyb: :P | 21:09 |
| jaypipes | mriedem: what is your ? on the image API? | 21:10 |
| mriedem | jaypipes: yeah | 21:10 |
| tonyb | mikal: I can only tink of one response and that's not appropriate for a public channel ;D | 21:10 |
| mriedem | jaypipes: like if image api get_all calls _get_session https://github.com/openstack/nova/blob/master/nova/image/api.py#L54 | 21:10 |
| mriedem | and that always gives us the GlanceImageService object, | 21:10 |
| mriedem | i'm trying to figure out where in nova we'd ever create and use the S3ImageService object | 21:11 |
| jaypipes | alaski, dansmith: and ok, sorry, do we have an agreement on my solution for not having the cell ID be something we pass in the API? | 21:11 |
| *** ihrachys has joined #openstack-nova | 21:11 | |
| jaypipes | mriedem: we don't any longer, AFAIK. That was for the nova-objectstore daemon that has been removed now that ec2-api is separate, right? | 21:11 |
| *** aginwala has quit IRC | 21:11 | |
| dansmith | jaypipes: personally I don't really think it's a huge problem, especially since this is an api for admins, but I'm fine with your aggregate approach too, as long as it's not a primary/required property for aggregate | 21:12 |
| *** morganfainberg is now known as notmorgan | 21:12 | |
| mriedem | jaypipes: that's the only thing i can think of | 21:12 |
| alaski | jaypipes: I think there's agreement that we may be able to make it work. Honestly I would like to see this all drawn out somehow as there's a lot to keep straight here | 21:12 |
| jaypipes | alaski: k, I will map it out in the next rev of generic-resource-pools spec. | 21:13 |
| mriedem | jaypipes: ah, removed here https://review.openstack.org/#/c/266425/ | 21:13 |
| *** vishwanathj has joined #openstack-nova | 21:13 | |
| mriedem | well, will be removed there | 21:13 |
| *** zenoway has joined #openstack-nova | 21:14 | |
| alaski | jaypipes: doffm is writing up a proposal on this as well, which I think matches yours. But it would be good to verify we're all actually on the same page here | 21:14 |
| jaypipes | k | 21:14 |
| *** vishwana_ has quit IRC | 21:15 | |
| dansmith | alaski: I suppose the other option would just be a mapping table in the api of resource-pool uuids to cells, like we have for instances | 21:16 |
| dansmith | purely for routing | 21:16 |
| *** aginwala has joined #openstack-nova | 21:16 | |
| dansmith | given the relationship between aggregates and pools, it probably makes sense to go that route though | 21:16 |
| *** papodaca_ has joined #openstack-nova | 21:16 | |
| *** penick has quit IRC | 21:17 | |
| papodaca_ | anyone know how `virsh capabilities` decides what host types are supported? mine only has a host section for some reason | 21:17 |
| alaski | right. And we're pretty set on aggregates being a thing in the api db | 21:17 |
| dansmith | aye | 21:18 |
| *** NightKhaos has quit IRC | 21:19 | |
| *** nekrodesk has quit IRC | 21:19 | |
| *** edtubill has joined #openstack-nova | 21:19 | |
| *** NightKhaos has joined #openstack-nova | 21:21 | |
| openstackgerrit | Chris Dent proposed openstack/nova: WIP resource-provider versioned objects https://review.openstack.org/277466 | 21:22 |
| *** ijw has quit IRC | 21:23 | |
| dansmith | cdent: what else do you think needs to be done there to s/WIP// | 21:23 |
| dansmith | ? | 21:23 |
| cdent | dansmith: that's an excellent question | 21:23 |
| *** flip214 has quit IRC | 21:23 | |
| * cdent thinks for a moment | 21:23 | |
| cdent | a) I'm unclear that is has the requisite methods on the objects to do their proper jobs | 21:24 |
| cdent | b) I'm not cognizant of the level of robustness/defensiveness these things are supposed to have | 21:24 |
| *** flip214 has joined #openstack-nova | 21:24 | |
| *** flip214 has joined #openstack-nova | 21:24 | |
| dansmith | well, I guess we won't fully know the answer to (a) until we have the compute node object patch above it to create/migrate those inventories, eh? | 21:24 |
| cdent | c) I've committed what I've got cuz I'm tired and haven't read through it all to see if I think it is somewhere close to done but I kinda think it probably is, but was going to do that tomorrow morning and give it a proper commit message then | 21:25 |
| dansmith | maybe I should work on that | 21:25 |
| cdent | if you've got the cycles go for it | 21:25 |
| cdent | I kinda think this stuff has been done in a weird order. | 21:26 |
| *** ihrachys has quit IRC | 21:26 | |
| dansmith | really? it's being done in the only order I think works at all :) | 21:26 |
| *** ihrachys has joined #openstack-nova | 21:27 | |
| dansmith | I hope I haven't stepped on your toes with it | 21:27 |
| dansmith | well, *too much*... a little toe assaulting isn't so bad | 21:27 |
| *** pushkaru has quit IRC | 21:27 | |
| cdent | I'd prefer to work backwards from a use case, building the objects and models as required, rather than starting from the models->objects->use case | 21:28 |
| cdent | Your help has been _very_ welcome | 21:28 |
| *** pushkaru has joined #openstack-nova | 21:28 | |
| dansmith | well, doing this in the reverse order means we don't get to land any of it until it's all done, | 21:28 |
| dansmith | which means newton or later | 21:29 |
| dansmith | and then either some big-bang aspect of it, or more complexity | 21:29 |
| cdent | yes, I'm learning to adapt to those constraints | 21:29 |
| *** zenoway has quit IRC | 21:35 | |
| *** rlrossit has joined #openstack-nova | 21:38 | |
| *** mylu has joined #openstack-nova | 21:39 | |
| mriedem | sdague: i haven't followed the status of the ec2 removal stuff, but is https://review.openstack.org/#/c/266425/ something we want to get done in mitaka? | 21:41 |
| armax | mriedem: ping | 21:42 |
| mriedem | pong | 21:42 |
| *** mylu has quit IRC | 21:42 | |
| *** rdopiera has quit IRC | 21:42 | |
| armax | mriedem: hi, I was reading the exchange between you and HenryG on this channel regarding get-me-a-network | 21:42 |
| armax | that was probably too early for me to jump in. | 21:43 |
| armax | you make a sensible point on lack of proper setup, but how you suggest we tackle it? | 21:43 |
| armax | I am not sure there’s really much it can be done…failures due to deployment misconfiguration can occur in any random circumstance | 21:44 |
| cfriesen | does anyone know where "xsd:dict" in our API guides is actually documented? | 21:44 |
| mriedem | armax: i don't have the code in front of me, but the functional test setup is checking 2 things (subnet and something else) | 21:44 |
| armax | mriedem: perhaps we can pick up the conversation in rochester, as I saw you signed up, but if you wanted to kick the tires sooner, feel free to give me a shout | 21:44 |
| armax | mriedem: I need to see the new patchset | 21:44 |
| mriedem | armax: HenryG said nova could validate those also, but that's leaking neutron details that nova shouldn't need to know about imo | 21:44 |
| armax | mriedem: and yes, you need certain deployment things to be there | 21:45 |
| armax | mriedem: those things could be done in devstack, the same way we create a public and a private tenant network | 21:45 |
| mriedem | armax: sure, but sans devstack, | 21:45 |
| armax | mriedem: sans devstack you don’t even have keystone catalog | 21:45 |
| armax | mriedem: so :) | 21:45 |
| mriedem | i'm looking for a way to fail fast in the nova-api before we cast to the compute node to actually try creating the network, just to find out things aren't configured and neutron fails the request b/c of that | 21:46 |
| mriedem | armax: well, sans devstack meaning joe schmo operator setting up his cloud w/o devstack | 21:46 |
| mriedem | for sure we'll need to do that setup in devstack to test this in the gate | 21:46 |
| armax | mriedem: my understanding is that hte network validation is done on the api side | 21:46 |
| mriedem | armax: neutron api? | 21:46 |
| armax | I mean, checking that the tenant has networks | 21:47 |
| armax | no in the nova-api | 21:47 |
| armax | so we would fail fast | 21:47 |
| mriedem | well, if we're doing get-me-a-network, the tenant doesn't have a network, and wants that auto-created for them, | 21:47 |
| mriedem | but nova-api doesn't want to actually do the network create at that time | 21:47 |
| mriedem | we do that in the compute | 21:47 |
| armax | correct, today we fail because we have no networks | 21:48 |
| mriedem | sort of like quota checking in the nova-api for network resources | 21:48 |
| mriedem | so what i was trying to get at, was doing some early sniff validation in nova-api such that the neutron network auto-create won't fail simply because the operator doesn't have the subnet setup or whatever | 21:48 |
| armax | so right now, but my understanding may be wrong, if we boot without requested networks | 21:48 |
| armax | where is the actual validation that ultimately leads a user error that there are no networks to be connected? | 21:49 |
| armax | api or compute? | 21:49 |
| mriedem | let me look | 21:49 |
| *** zenoway has joined #openstack-nova | 21:49 | |
| armax | I thought taht was on the api side | 21:49 |
| *** papodaca_ has left #openstack-nova | 21:50 | |
| *** rk4n has quit IRC | 21:50 | |
| *** ctrath has joined #openstack-nova | 21:51 | |
| *** eharney has quit IRC | 21:52 | |
| *** ociuhandu has quit IRC | 21:52 | |
| armax | mriedem: I agree with the fail fast principle, we just have to figure out what’s the best way of doing so and whether we check at every tenant request | 21:53 |
| *** ociuhandu has joined #openstack-nova | 21:53 | |
| *** cdent has quit IRC | 21:53 | |
| mriedem | armax: yeah i think we get here https://github.com/openstack/nova/blob/master/nova/network/neutronv2/api.py#L202-L207 | 21:53 |
| *** aginwala has quit IRC | 21:54 | |
| armax | mriedem: that happens on nova-api does it not? | 21:54 |
| armax | it’s hard to tell | 21:54 |
| armax | I have a faint recollocation of my old nova days though | 21:54 |
| *** thumpba has quit IRC | 21:54 | |
| mriedem | armax: yeah, that's from nova.compute.api_validate_networks -> neutronv2.api.validate_networks | 21:55 |
| mriedem | eventually gets down in there | 21:55 |
| mriedem | interestingly, | 21:55 |
| mriedem | i don't see anything that checks if there are 0 networks to use and would fail that validation in the api side | 21:56 |
| armax | mriedem: go on… | 21:56 |
| armax | mriedem: um that can’t be right | 21:56 |
| mriedem | can this result in an empty list? https://github.com/openstack/nova/blob/master/nova/network/neutronv2/api.py#L202-L207 | 21:56 |
| *** daemontool_ has quit IRC | 21:56 | |
| *** amotoki has joined #openstack-nova | 21:56 | |
| *** smatzek has joined #openstack-nova | 21:57 | |
| *** aginwala has joined #openstack-nova | 21:57 | |
| mriedem | and if that is the case at allocate time in the compute, we don't seem to care much https://github.com/openstack/nova/blob/master/nova/network/neutronv2/api.py#L568 | 21:57 |
| armax | mriedem: I had this at one point, I am gonna have to dig deeper, now I have to run the neutron drivers meeting…I’ll get back to this | 21:58 |
| armax | I wodner if it’s something in the base api layer for network | 21:59 |
| armax | but it doesn’t look like it | 21:59 |
| *** nekrodesk has joined #openstack-nova | 22:00 | |
| *** nkrinner has quit IRC | 22:00 | |
| *** delatte has quit IRC | 22:00 | |
| *** _ducttape_ has joined #openstack-nova | 22:00 | |
| *** ijw has joined #openstack-nova | 22:00 | |
| mriedem | i don't think so | 22:00 |
| *** haomaiwang has quit IRC | 22:01 | |
| mriedem | i think people intentionally just put their heads in the sand here https://bugs.launchpad.net/nova/+bug/1191044 | 22:01 |
| openstack | Launchpad bug 1191044 in OpenStack Compute (nova) grizzly "Create VM with no networks with Quantum Gives SecurityGroups Exception" [Medium,Fix released] - Assigned to Aaron Rosen (arosen) | 22:01 |
| mriedem | 'trying to boot a vm with no networks results in a security group issue, so let's just not run that code...' | 22:01 |
| *** amotoki has quit IRC | 22:01 | |
| *** haomaiwang has joined #openstack-nova | 22:01 | |
| mriedem | no one ever notices with devstack b/c devstack sets up the network for you | 22:01 |
| armax | mriedem: let’s clean this up! | 22:02 |
| mriedem | heh, there was even a -1 from a young and starry eyed mriedem | 22:02 |
| *** burgerk has quit IRC | 22:03 | |
| mriedem | unless....you can later attach a port | 22:03 |
| *** ducttape_ has quit IRC | 22:03 | |
| mriedem | but that kind of craps all over our vif plugging callback code i think | 22:03 |
| *** penick has joined #openstack-nova | 22:05 | |
| *** bwallis has joined #openstack-nova | 22:06 | |
| mriedem | looks like we're ok on the callback code, network_info vif list will be empty and we won't wait for any events | 22:08 |
| *** Sukhdev has joined #openstack-nova | 22:08 | |
| *** nekrodesk has joined #openstack-nova | 22:09 | |
| *** aginwala_ has joined #openstack-nova | 22:10 | |
| *** aginwala has quit IRC | 22:10 | |
| *** thorst_ has quit IRC | 22:11 | |
| *** thorst_ has joined #openstack-nova | 22:11 | |
| *** dimtruck is now known as zz_dimtruck | 22:11 | |
| mtreinish | mriedem: you can easily test that in devstack. Devstack only setups up networks for the tenants it creates, if you create any additional ones it doesn't have anything | 22:12 |
| mriedem | and there isn't a public shared network? | 22:13 |
| *** ijw has quit IRC | 22:13 | |
| *** ijw has joined #openstack-nova | 22:13 | |
| armax | mriedem: yes there | 22:14 |
| armax | is | 22:14 |
| mriedem | yeah, so from what i can tell, nova allows you to create a vm without any networking (with neutron), | 22:15 |
| mriedem | and i think, if you try attaching an interface to it without specifying a network, it'll puke (as it should) | 22:16 |
| *** thorst_ has quit IRC | 22:16 | |
| armax | mriedem: yeah…I am starting to remember that the behavior changed at one point | 22:16 |
| openstackgerrit | Augustina Ragwitz proposed openstack/nova: Fix _compare_result type handling comparison https://review.openstack.org/278191 | 22:17 |
| mriedem | just no one would hit this with devstack if devstack is creating a shared public network | 22:17 |
| mriedem | anyway, this makes the nova stuff for get me a network at least a bit more complicated since it's possible to boot without a network today and it's not an error | 22:17 |
| mriedem | i really need to get a current devstack setup with neutron, haven't done that in awhile | 22:18 |
| *** knikolla has quit IRC | 22:19 | |
| *** ganso has quit IRC | 22:19 | |
| *** zz_dimtruck is now known as dimtruck | 22:20 | |
| openstackgerrit | Andrew Laski proposed openstack/nova: Create BuildRequest object during boot process https://review.openstack.org/278125 | 22:20 |
| openstackgerrit | Andrew Laski proposed openstack/nova: Add build_requests database table and model https://review.openstack.org/263926 | 22:20 |
| openstackgerrit | Andrew Laski proposed openstack/nova: Add BuildRequest object https://review.openstack.org/278124 | 22:20 |
| openstackgerrit | Andrew Laski proposed openstack/nova: Make InstanceMappings.cell_id nullable https://review.openstack.org/278598 | 22:20 |
| openstackgerrit | Andrew Laski proposed openstack/nova: WIP Populate instance_mappings during boot. https://review.openstack.org/263925 | 22:20 |
| mtreinish | mriedem: it probably sets up an external network, but IIRC you can't boot with just that | 22:21 |
| *** rk4n has joined #openstack-nova | 22:21 | |
| mriedem | mtreinish: i seem to recall that check in nova too, sec | 22:21 |
| *** mragupat has quit IRC | 22:23 | |
| *** dims has quit IRC | 22:23 | |
| armax | mriedem: I am not sure the code will be more complicated | 22:23 |
| mriedem | armax: it would be hard to make nova.network.neutronv2.api more complicated :) | 22:23 |
| armax | mriedem: fair enough :) | 22:23 |
| mtreinish | mriedem: I have faith that you can do it | 22:24 |
| mriedem | that makes one of us | 22:24 |
| mtreinish | haha | 22:24 |
| armax | mriedem: in the end, nova checks for what networks the tenant has access to | 22:24 |
| armax | mriedem: if there are, fine, if there aren’t it can provision one and get on with it | 22:25 |
| SpamapS | ctrath: hey, we were talking offline, let's chat a bit in here.. | 22:25 |
| mriedem | armax: sure | 22:25 |
| SpamapS | ctrath: can you point me at any docs about how the scheduler will work with cells v2? | 22:25 |
| armax | mriedem: now that raises the question as two what we do with the code path where we boot a vm without networks | 22:25 |
| mriedem | armax: yeah, see, that's what i'm thinking about | 22:26 |
| *** Swami has quit IRC | 22:26 | |
| armax | but we’d need to ensure consistency between the nova-network codepath and the neutron one | 22:26 |
| mriedem | like, do you just let that continue, or do you actually try creating the network (how do you determine what the user's intent was) | 22:26 |
| *** dims has joined #openstack-nova | 22:26 | |
| mtreinish | armax: while I agree that's a good direction, it change a lot of the current workflow with neutron | 22:26 |
| mriedem | armax: i think john was also expecting a nova microversion of this, i have to think through how that is going to be checked, because as of this morning i thought a network was required | 22:27 |
| mtreinish | would that have negative effects for people who have code out there today to do the network allocation outside of nova | 22:27 |
| mriedem | mtreinish: if you create the network outside of nova aren't you expected to provide that info at boot time of the vm? | 22:28 |
| armax | mriedem, mtreinish: agreed. | 22:28 |
| mriedem | well, unless nova is looking up the networks for your tenant (like it does today) | 22:28 |
| *** bwallis has quit IRC | 22:28 | |
| *** pradk has quit IRC | 22:28 | |
| *** mragupat has joined #openstack-nova | 22:29 | |
| *** smatzek has quit IRC | 22:29 | |
| mtreinish | mriedem: right that's my concern. The only time you have to specify is if there is > 1 network (because nova doesn't know how to boot with 2 networks) | 22:30 |
| mriedem | mtreinish: this is the thing i was looking for https://review.openstack.org/#/c/85189/ | 22:30 |
| mtreinish | we have a lot of logic in tempest around doing all the neutron magic to boot a guest with the right network | 22:30 |
| mriedem | mtreinish: well, i'd think that if you boot a vm w/o specifying a network and there are no networks available, then nova would request one | 22:30 |
| mriedem | especially if you can't boot a vm w/o a network and then later attach one | 22:31 |
| mriedem | but i think you might be able to do the latter if you're providing a port/fixed_ip | 22:31 |
| *** thorst_ has joined #openstack-nova | 22:31 | |
| mriedem | idk, maybe for the first round we don't want to even deal with attach_interface | 22:32 |
| *** thorst_ has quit IRC | 22:33 | |
| *** thorst_ has joined #openstack-nova | 22:33 | |
| *** daemontool has joined #openstack-nova | 22:34 | |
| *** sshen has joined #openstack-nova | 22:35 | |
| *** _ducttape_ has quit IRC | 22:37 | |
| *** ducttape_ has joined #openstack-nova | 22:38 | |
| *** thorst_ has quit IRC | 22:39 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: neutron: remove redundant request.network_id assignment https://review.openstack.org/279295 | 22:42 |
| mriedem | there, my first contribution ^ | 22:42 |
| *** salv-orlando has joined #openstack-nova | 22:42 | |
| *** cdelatte has joined #openstack-nova | 22:42 | |
| *** thorst_ has joined #openstack-nova | 22:43 | |
| *** salv-orl_ has quit IRC | 22:44 | |
| *** mylu has joined #openstack-nova | 22:45 | |
| *** thorst_ has quit IRC | 22:48 | |
| *** penick has quit IRC | 22:51 | |
| *** penick has joined #openstack-nova | 22:53 | |
| *** mriedem has quit IRC | 22:53 | |
| *** aginwala_ has quit IRC | 22:56 | |
| *** aginwala has joined #openstack-nova | 22:56 | |
| *** xyang1 has quit IRC | 22:58 | |
| *** haomaiwang has quit IRC | 23:01 | |
| *** haomaiwang has joined #openstack-nova | 23:01 | |
| *** edleafe has quit IRC | 23:02 | |
| *** ijw has quit IRC | 23:02 | |
| *** thorst_ has joined #openstack-nova | 23:03 | |
| *** mdrabe has quit IRC | 23:07 | |
| anteaya | yay | 23:08 |
| *** vishwanathj has quit IRC | 23:08 | |
| *** edleafe has joined #openstack-nova | 23:09 | |
| *** rlrossit has quit IRC | 23:09 | |
| *** mragupat has quit IRC | 23:11 | |
| *** ihrachys has quit IRC | 23:11 | |
| *** dedery has quit IRC | 23:13 | |
| *** lpetrut has quit IRC | 23:13 | |
| *** edtubill has quit IRC | 23:15 | |
| *** shz has quit IRC | 23:16 | |
| *** shihanzhang has joined #openstack-nova | 23:17 | |
| *** thorst_ has quit IRC | 23:20 | |
| *** thorst_ has joined #openstack-nova | 23:21 | |
| *** burt has quit IRC | 23:23 | |
| openstackgerrit | Dan Smith proposed openstack/nova: WIP resource-provider versioned objects https://review.openstack.org/277466 | 23:27 |
| openstackgerrit | Dan Smith proposed openstack/nova: Allocate UUID for compute node https://review.openstack.org/277554 | 23:27 |
| openstackgerrit | Dan Smith proposed openstack/nova: Make Service.get_minimum_version() work for multiple binaries https://review.openstack.org/279312 | 23:27 |
| openstackgerrit | Dan Smith proposed openstack/nova: WIP: Migrate compute node stats to Inventory objects https://review.openstack.org/279313 | 23:27 |
| *** pushkaru has quit IRC | 23:27 | |
| *** thorst_ has quit IRC | 23:29 | |
| *** baoli has quit IRC | 23:33 | |
| *** baoli has joined #openstack-nova | 23:36 | |
| *** ccarmack has quit IRC | 23:41 | |
| *** ctrath has quit IRC | 23:42 | |
| *** ljxiash has quit IRC | 23:42 | |
| *** zenoway has quit IRC | 23:44 | |
| *** marcusvrn_ has quit IRC | 23:47 | |
| *** nekrodesk has quit IRC | 23:47 | |
| *** Bjoern has quit IRC | 23:47 | |
| *** ivase has quit IRC | 23:49 | |
| *** haplo37 has quit IRC | 23:51 | |
| *** mylu has quit IRC | 23:52 | |
| *** dimtruck is now known as zz_dimtruck | 23:52 | |
| *** mylu has joined #openstack-nova | 23:53 | |
| *** ducttape_ has quit IRC | 23:55 | |
| *** diana_clarke has quit IRC | 23:56 | |
| *** amotoki has joined #openstack-nova | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!