| *** ijw has joined #openstack-nova | 00:01 | |
| *** Jeffrey4l has joined #openstack-nova | 00:03 | |
| *** browne has quit IRC | 00:04 | |
| openstackgerrit | Jay Pipes proposed openstack/nova: placement: allocations for generic PCI devices https://review.openstack.org/375168 | 00:08 |
|---|---|---|
| *** jaypipes has quit IRC | 00:09 | |
| *** yamahata has quit IRC | 00:09 | |
| *** armax has joined #openstack-nova | 00:14 | |
| *** ducttape_ has joined #openstack-nova | 00:14 | |
| *** ducttape_ has quit IRC | 00:15 | |
| *** ducttape_ has joined #openstack-nova | 00:15 | |
| *** jwcroppe_ has quit IRC | 00:17 | |
| *** browne has joined #openstack-nova | 00:20 | |
| *** kaisers has joined #openstack-nova | 00:22 | |
| *** kaisers2 has quit IRC | 00:22 | |
| *** kaisers1 has quit IRC | 00:23 | |
| *** tuanla has joined #openstack-nova | 00:23 | |
| *** kaisers1 has joined #openstack-nova | 00:23 | |
| *** marst has joined #openstack-nova | 00:29 | |
| *** kuzko has quit IRC | 00:32 | |
| *** kuzko has joined #openstack-nova | 00:34 | |
| *** stewie925 has quit IRC | 00:35 | |
| *** Apoorva_ has joined #openstack-nova | 00:36 | |
| *** Apoorva_ has quit IRC | 00:36 | |
| *** Apoorva has quit IRC | 00:40 | |
| *** ducttape_ has quit IRC | 00:40 | |
| *** phuongnh has joined #openstack-nova | 00:47 | |
| *** kuzko has quit IRC | 00:48 | |
| *** kuzko has joined #openstack-nova | 00:51 | |
| *** marst has quit IRC | 00:52 | |
| *** mtanin___ has joined #openstack-nova | 00:52 | |
| *** moshele has joined #openstack-nova | 00:52 | |
| *** mtanino has quit IRC | 00:53 | |
| *** markvoelker has joined #openstack-nova | 00:55 | |
| *** ijw has quit IRC | 00:57 | |
| *** hoangcx has joined #openstack-nova | 01:00 | |
| *** gyee has quit IRC | 01:01 | |
| *** newtime has joined #openstack-nova | 01:01 | |
| *** newtime is now known as Xingbin | 01:01 | |
| *** jamielennox is now known as jamielennox|away | 01:03 | |
| *** namnh has joined #openstack-nova | 01:07 | |
| *** ducttape_ has joined #openstack-nova | 01:09 | |
| *** ducttape_ has quit IRC | 01:13 | |
| *** ijw has joined #openstack-nova | 01:14 | |
| *** ducttape_ has joined #openstack-nova | 01:16 | |
| *** dimtruck is now known as zz_dimtruck | 01:21 | |
| *** jamesden_ has joined #openstack-nova | 01:22 | |
| *** jamielennox|away is now known as jamielennox | 01:22 | |
| *** ijw has quit IRC | 01:28 | |
| *** browne has quit IRC | 01:37 | |
| *** Sukhdev has joined #openstack-nova | 01:37 | |
| *** raunak has quit IRC | 01:39 | |
| *** hemanthm|afk is now known as hemanthm | 01:39 | |
| *** armax has quit IRC | 01:39 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: VMware: deprecate wsdl_location conf option https://review.openstack.org/342470 | 01:39 |
| *** salv-orl_ has joined #openstack-nova | 01:42 | |
| *** nic has quit IRC | 01:43 | |
| *** salv-orlando has quit IRC | 01:45 | |
| *** EinstCrazy has joined #openstack-nova | 01:50 | |
| *** unicell1 has joined #openstack-nova | 01:51 | |
| *** unicell has quit IRC | 01:51 | |
| *** mriedem has quit IRC | 01:53 | |
| *** ducttape_ has quit IRC | 02:01 | |
| *** gjayavelu has joined #openstack-nova | 02:04 | |
| *** thorst has joined #openstack-nova | 02:11 | |
| *** mtanino has joined #openstack-nova | 02:16 | |
| *** thorst has quit IRC | 02:16 | |
| *** Xingbin has quit IRC | 02:16 | |
| *** mtanin___ has quit IRC | 02:17 | |
| *** Sukhdev has quit IRC | 02:19 | |
| *** ducttape_ has joined #openstack-nova | 02:19 | |
| *** claudiub has quit IRC | 02:21 | |
| *** mtanino has quit IRC | 02:22 | |
| *** browne has joined #openstack-nova | 02:22 | |
| *** browne has quit IRC | 02:23 | |
| *** tbachman has quit IRC | 02:23 | |
| *** zz_dimtruck is now known as dimtruck | 02:24 | |
| *** rfolco has quit IRC | 02:24 | |
| *** Sukhdev has joined #openstack-nova | 02:28 | |
| *** jamesden_ has quit IRC | 02:30 | |
| *** rfolco has joined #openstack-nova | 02:32 | |
| *** TravT has joined #openstack-nova | 02:33 | |
| *** mugsie has quit IRC | 02:33 | |
| *** mugsie has joined #openstack-nova | 02:35 | |
| *** jamielennox is now known as jamielennox|away | 02:37 | |
| *** Sukhdev has quit IRC | 02:38 | |
| *** amotoki has quit IRC | 02:41 | |
| *** kaisers_ has joined #openstack-nova | 02:42 | |
| *** rfolco has quit IRC | 02:42 | |
| *** kaisers_ has quit IRC | 02:46 | |
| *** rama_y has quit IRC | 02:48 | |
| *** jamielennox|away is now known as jamielennox | 02:51 | |
| *** mlavalle has quit IRC | 02:51 | |
| *** amotoki has joined #openstack-nova | 02:53 | |
| *** kevinz has quit IRC | 02:55 | |
| *** ducttape_ has quit IRC | 03:01 | |
| *** andrewbogott has quit IRC | 03:01 | |
| *** andrewbogott has joined #openstack-nova | 03:01 | |
| *** thorst has joined #openstack-nova | 03:04 | |
| *** david-lyle has quit IRC | 03:04 | |
| *** amotoki has quit IRC | 03:09 | |
| *** moshele has quit IRC | 03:11 | |
| *** haplo37 has quit IRC | 03:13 | |
| *** unicell has joined #openstack-nova | 03:17 | |
| *** unicell1 has quit IRC | 03:17 | |
| *** amotoki has joined #openstack-nova | 03:19 | |
| *** prateek has joined #openstack-nova | 03:19 | |
| *** raunak has joined #openstack-nova | 03:22 | |
| *** rmcall_ has joined #openstack-nova | 03:24 | |
| *** rmcall has quit IRC | 03:26 | |
| *** rmcall_ is now known as rmcall | 03:26 | |
| *** ayoung has joined #openstack-nova | 03:26 | |
| *** mtanino has joined #openstack-nova | 03:26 | |
| *** armax has joined #openstack-nova | 03:27 | |
| *** rfolco has joined #openstack-nova | 03:27 | |
| *** rfolco has quit IRC | 03:27 | |
| *** dimtruck is now known as zz_dimtruck | 03:31 | |
| *** namnh has quit IRC | 03:31 | |
| *** rama_y has joined #openstack-nova | 03:32 | |
| openstackgerrit | John Hua proposed openstack/nova: XenAPI: device tagging https://review.openstack.org/333781 | 03:37 |
| *** amotoki has quit IRC | 03:39 | |
| *** mtanino has quit IRC | 03:41 | |
| *** adisky has joined #openstack-nova | 03:41 | |
| *** gouthamr has joined #openstack-nova | 03:44 | |
| *** Sukhdev has joined #openstack-nova | 03:46 | |
| *** haplo37 has joined #openstack-nova | 03:48 | |
| *** gouthamr has quit IRC | 03:49 | |
| *** amotoki has joined #openstack-nova | 03:49 | |
| *** dikonoor has joined #openstack-nova | 03:49 | |
| openstackgerrit | John Hua proposed openstack/nova: XenAPI: Remove ovs_integration_bridge default value https://review.openstack.org/299092 | 03:53 |
| *** baoli_ has joined #openstack-nova | 03:54 | |
| *** ijw has joined #openstack-nova | 03:56 | |
| *** raunak has quit IRC | 03:57 | |
| *** baoli has quit IRC | 03:58 | |
| openstackgerrit | LIU Yulong proposed openstack/nova-specs: Reset the instance keypair while rebuilding (spec) https://review.openstack.org/375221 | 04:01 |
| *** ijw has quit IRC | 04:01 | |
| *** sdake has quit IRC | 04:02 | |
| *** ManishD has joined #openstack-nova | 04:03 | |
| *** marst has joined #openstack-nova | 04:04 | |
| *** rkmrhj has quit IRC | 04:04 | |
| *** raunak has joined #openstack-nova | 04:07 | |
| openstackgerrit | LIU Yulong proposed openstack/nova-specs: Reset the instance keypair while rebuilding (spec) https://review.openstack.org/375221 | 04:08 |
| *** coolsvap has joined #openstack-nova | 04:08 | |
| *** thorst has quit IRC | 04:09 | |
| *** ManishD has quit IRC | 04:10 | |
| *** GB21 has joined #openstack-nova | 04:10 | |
| *** thorst has joined #openstack-nova | 04:12 | |
| *** baoli has joined #openstack-nova | 04:14 | |
| *** gjayavelu has quit IRC | 04:15 | |
| *** gjayavelu has joined #openstack-nova | 04:15 | |
| *** baoli_ has quit IRC | 04:17 | |
| *** ijw has joined #openstack-nova | 04:17 | |
| *** ijw has quit IRC | 04:19 | |
| *** gjayavelu has quit IRC | 04:20 | |
| *** thorst has quit IRC | 04:23 | |
| *** thorst has joined #openstack-nova | 04:24 | |
| *** haplo37_ has quit IRC | 04:27 | |
| *** thorst has quit IRC | 04:29 | |
| *** haplo37_ has joined #openstack-nova | 04:29 | |
| *** baoli has quit IRC | 04:29 | |
| *** haplo37 has quit IRC | 04:30 | |
| *** haplo37 has joined #openstack-nova | 04:31 | |
| *** tbachman has joined #openstack-nova | 04:32 | |
| *** swamireddy has quit IRC | 04:33 | |
| openstackgerrit | LIU Yulong proposed openstack/nova-specs: Reset the instance keypair while rebuilding (spec) https://review.openstack.org/375221 | 04:34 |
| openstackgerrit | Tony Xu proposed openstack/nova: Fix pep8 E501 line too long https://review.openstack.org/375231 | 04:37 |
| *** unicell1 has joined #openstack-nova | 04:44 | |
| *** unicell has quit IRC | 04:45 | |
| *** bks has joined #openstack-nova | 04:46 | |
| *** sudswas__ has joined #openstack-nova | 04:46 | |
| *** Guest56480 has joined #openstack-nova | 04:46 | |
| *** prateek has quit IRC | 04:47 | |
| *** thorst has joined #openstack-nova | 04:56 | |
| *** thorst has quit IRC | 05:01 | |
| *** bhagyashri has quit IRC | 05:01 | |
| *** haplo37_ has quit IRC | 05:05 | |
| *** lpetrut has joined #openstack-nova | 05:06 | |
| *** haplo37_ has joined #openstack-nova | 05:07 | |
| *** haplo37_ has quit IRC | 05:07 | |
| *** haplo37_ has joined #openstack-nova | 05:10 | |
| *** jaosorior has joined #openstack-nova | 05:10 | |
| *** haplo37_ has quit IRC | 05:10 | |
| *** haplo37_ has joined #openstack-nova | 05:13 | |
| *** ecelik has joined #openstack-nova | 05:13 | |
| openstackgerrit | huan proposed openstack/nova: xenapi: support the hotplug of a neutron port https://review.openstack.org/365505 | 05:14 |
| *** sabari has quit IRC | 05:17 | |
| *** haplo37 has quit IRC | 05:19 | |
| *** rmcall has quit IRC | 05:19 | |
| *** tbachman_ has joined #openstack-nova | 05:19 | |
| *** tbachman has quit IRC | 05:20 | |
| *** tbachman_ is now known as tbachman | 05:20 | |
| *** haplo37 has joined #openstack-nova | 05:21 | |
| openstackgerrit | huan proposed openstack/nova: XenAPI: Fix VM live-migrate with iSCSI SR volume https://review.openstack.org/359548 | 05:21 |
| *** rmcall has joined #openstack-nova | 05:22 | |
| *** paul-carlton2 has joined #openstack-nova | 05:25 | |
| *** tbachman has quit IRC | 05:26 | |
| *** thorst has joined #openstack-nova | 05:27 | |
| *** GordonTX has joined #openstack-nova | 05:28 | |
| *** armax has quit IRC | 05:29 | |
| *** thorst has quit IRC | 05:31 | |
| *** armax has joined #openstack-nova | 05:32 | |
| *** armax has quit IRC | 05:34 | |
| *** bhagyashri has joined #openstack-nova | 05:35 | |
| *** armax has joined #openstack-nova | 05:36 | |
| *** sabari has joined #openstack-nova | 05:36 | |
| *** raunak has quit IRC | 05:36 | |
| *** armax has quit IRC | 05:38 | |
| *** kaisers_ has joined #openstack-nova | 05:40 | |
| *** gszasz has joined #openstack-nova | 05:42 | |
| *** thorst has joined #openstack-nova | 05:43 | |
| *** lpetrut has quit IRC | 05:46 | |
| *** thorst has quit IRC | 05:48 | |
| *** haplo37_ has quit IRC | 05:48 | |
| *** haplo37_ has joined #openstack-nova | 05:50 | |
| *** armax has joined #openstack-nova | 05:50 | |
| *** mjura has joined #openstack-nova | 05:52 | |
| *** prateek has joined #openstack-nova | 05:57 | |
| *** salv-orl_ has quit IRC | 05:57 | |
| *** thorst has joined #openstack-nova | 05:59 | |
| *** mjura has quit IRC | 05:59 | |
| *** mjura has joined #openstack-nova | 06:00 | |
| *** prateek has quit IRC | 06:01 | |
| *** prateek has joined #openstack-nova | 06:01 | |
| *** ianw is now known as ianw_pto | 06:02 | |
| *** thorst has quit IRC | 06:03 | |
| *** yamahata has joined #openstack-nova | 06:09 | |
| *** thorst has joined #openstack-nova | 06:15 | |
| *** rcernin has joined #openstack-nova | 06:15 | |
| *** unicell has joined #openstack-nova | 06:16 | |
| *** unicell1 has quit IRC | 06:16 | |
| *** thorst has quit IRC | 06:19 | |
| *** claudiub has joined #openstack-nova | 06:20 | |
| *** mamandle has joined #openstack-nova | 06:21 | |
| *** aswadr_ has joined #openstack-nova | 06:25 | |
| *** andreas_s has joined #openstack-nova | 06:25 | |
| *** phuongnh has quit IRC | 06:29 | |
| openstackgerrit | javeme proposed openstack/nova: Failed to evacuate instances booted from volume https://review.openstack.org/372045 | 06:29 |
| *** phuongnh has joined #openstack-nova | 06:29 | |
| *** thorst has joined #openstack-nova | 06:31 | |
| openstackgerrit | John Hua proposed openstack/nova: XenAPI: Remove ovs_integration_bridge default value https://review.openstack.org/299092 | 06:34 |
| *** thorst has quit IRC | 06:36 | |
| *** lpetrut has joined #openstack-nova | 06:36 | |
| *** woodster_ has quit IRC | 06:40 | |
| *** haplo37 has quit IRC | 06:43 | |
| *** nherciu_ has joined #openstack-nova | 06:44 | |
| *** haplo37 has joined #openstack-nova | 06:44 | |
| *** GB21 has quit IRC | 06:45 | |
| *** rrecio_ has quit IRC | 06:45 | |
| *** lpetrut has quit IRC | 06:45 | |
| *** thorst has joined #openstack-nova | 06:47 | |
| *** thorst has quit IRC | 06:51 | |
| *** GordonTX has quit IRC | 06:52 | |
| *** sridharg has joined #openstack-nova | 06:53 | |
| *** matrohon has joined #openstack-nova | 06:54 | |
| *** hoangcx_ has joined #openstack-nova | 06:55 | |
| *** GB21 has joined #openstack-nova | 06:57 | |
| *** hoangcx has quit IRC | 06:57 | |
| *** jlanoux has joined #openstack-nova | 06:58 | |
| *** phuongnh has quit IRC | 06:58 | |
| *** phuongnh has joined #openstack-nova | 06:58 | |
| *** thorst has joined #openstack-nova | 07:02 | |
| *** salv-orlando has joined #openstack-nova | 07:03 | |
| *** salv-orlando has quit IRC | 07:03 | |
| *** andymaier has joined #openstack-nova | 07:04 | |
| *** salv-orlando has joined #openstack-nova | 07:05 | |
| *** thorst has quit IRC | 07:07 | |
| *** salv-orlando has quit IRC | 07:08 | |
| *** sahid has joined #openstack-nova | 07:10 | |
| *** Sukhdev has quit IRC | 07:11 | |
| *** thorst has joined #openstack-nova | 07:18 | |
| *** bauzas is now known as bauwser | 07:18 | |
| *** jlanoux_ has joined #openstack-nova | 07:20 | |
| *** ihrachys has joined #openstack-nova | 07:21 | |
| *** ihrachys has quit IRC | 07:22 | |
| *** ihrachys has joined #openstack-nova | 07:22 | |
| *** jlanoux has quit IRC | 07:23 | |
| *** thorst has quit IRC | 07:23 | |
| *** slaweq_ has quit IRC | 07:24 | |
| *** zenoway has joined #openstack-nova | 07:30 | |
| *** rmart04 has joined #openstack-nova | 07:32 | |
| openstackgerrit | Hieu LE proposed openstack/nova: Config options: improve libvirt help text (2) https://review.openstack.org/316550 | 07:37 |
| *** hoangcx_ has quit IRC | 07:38 | |
| *** hoangcx has joined #openstack-nova | 07:39 | |
| *** slaweq_ has joined #openstack-nova | 07:41 | |
| *** lpetrut has joined #openstack-nova | 07:42 | |
| *** dave-johnston has joined #openstack-nova | 07:46 | |
| openstackgerrit | javeme proposed openstack/nova: Mock.side_effects does not exist, use Mock.side_effect instead https://review.openstack.org/375298 | 07:48 |
| *** thorst has joined #openstack-nova | 07:50 | |
| *** thorst has quit IRC | 07:54 | |
| *** mamandle has quit IRC | 07:55 | |
| *** mariusv has joined #openstack-nova | 07:56 | |
| *** zzzeek has quit IRC | 08:00 | |
| *** zzzeek has joined #openstack-nova | 08:00 | |
| *** amotoki_ has joined #openstack-nova | 08:00 | |
| *** amotoki has quit IRC | 08:03 | |
| *** pilgrimstack has joined #openstack-nova | 08:05 | |
| *** yassine has joined #openstack-nova | 08:07 | |
| *** yassine is now known as Guest49837 | 08:08 | |
| openstackgerrit | sahid proposed openstack/nova-specs: Libvirt driver emulator threads placement policy https://review.openstack.org/284094 | 08:08 |
| *** Guest49837 is now known as yassou | 08:08 | |
| *** avolkov has quit IRC | 08:17 | |
| *** GordonTX has joined #openstack-nova | 08:19 | |
| *** thorst has joined #openstack-nova | 08:21 | |
| *** danpb has joined #openstack-nova | 08:23 | |
| *** GordonTX has quit IRC | 08:24 | |
| *** cdent has joined #openstack-nova | 08:24 | |
| *** amotoki_ is now known as amotoki | 08:24 | |
| *** avolkov has joined #openstack-nova | 08:26 | |
| *** thorst has quit IRC | 08:27 | |
| paul-carlton2 | johnthetubaguy bauwser see updated https://bugs.launchpad.net/nova/+bug/1626230, creating a blueprint to make libvirt spawn method tidy up on failure like XenAPI does | 08:31 |
| openstack | Launchpad bug 1626230 in OpenStack Compute (nova) "instance artefacts are not removed by libvirt driver if it fails to spawn" [Undecided,In progress] - Assigned to Paul Carlton (paul-carlton2) | 08:31 |
| paul-carlton2 | bauwser, you happy with this approach? | 08:31 |
| *** jlanoux_ has quit IRC | 08:32 | |
| *** ihrachys has quit IRC | 08:33 | |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs: Support volume_type as a BDM parameter when boot instances https://review.openstack.org/362698 | 08:35 |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs: Add support for tag instances when boot https://review.openstack.org/366469 | 08:36 |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs: Add notifications for server groups https://review.openstack.org/375316 | 08:36 |
| *** CristinaPauna has quit IRC | 08:36 | |
| *** thorst has joined #openstack-nova | 08:38 | |
| *** swamireddy has joined #openstack-nova | 08:38 | |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs: Add notifications for server groups https://review.openstack.org/375316 | 08:40 |
| *** haplo37 has quit IRC | 08:41 | |
| *** haplo37 has joined #openstack-nova | 08:42 | |
| *** GB21 has quit IRC | 08:42 | |
| *** thorst has quit IRC | 08:42 | |
| *** vladikr has quit IRC | 08:42 | |
| *** derekh has joined #openstack-nova | 08:44 | |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/nova: Return instance of Guest from method write_instance_config https://review.openstack.org/265146 | 08:44 |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/nova: Don't use 'updated_at' to check service's status https://review.openstack.org/297085 | 08:44 |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/nova: Log warning when user set improper config option value https://review.openstack.org/354532 | 08:45 |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/nova: Use min parameter to restrict live-migration config options https://review.openstack.org/323313 | 08:45 |
| *** GB21 has joined #openstack-nova | 08:46 | |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs: Add notifications for server groups https://review.openstack.org/375316 | 08:47 |
| *** Dinesh_Bhor has joined #openstack-nova | 08:48 | |
| *** CristinaPauna has joined #openstack-nova | 08:48 | |
| *** yassou has quit IRC | 08:48 | |
| *** code-R has joined #openstack-nova | 08:51 | |
| *** code-R_ has joined #openstack-nova | 08:56 | |
| *** mamandle has joined #openstack-nova | 08:57 | |
| *** code-R has quit IRC | 08:59 | |
| *** kaisers_ has quit IRC | 09:02 | |
| openstackgerrit | Chris Dent proposed openstack/nova: [placement] ensure that allow headers are native strings https://review.openstack.org/374800 | 09:02 |
| *** ihrachys has joined #openstack-nova | 09:03 | |
| *** EinstCrazy has quit IRC | 09:03 | |
| *** openstackgerrit has quit IRC | 09:03 | |
| *** yamahata has quit IRC | 09:03 | |
| *** openstackgerrit has joined #openstack-nova | 09:04 | |
| *** EinstCrazy has joined #openstack-nova | 09:04 | |
| openstackgerrit | Jens Rosenboom proposed openstack/nova: Default vnc.keymap to None https://review.openstack.org/375333 | 09:06 |
| *** GB21 has quit IRC | 09:06 | |
| *** EinstCrazy has quit IRC | 09:07 | |
| openstackgerrit | veena proposed openstack/nova: Adding hugepage and NUMA support check for aarch64 https://review.openstack.org/372304 | 09:10 |
| frickler | danpb: as the original recommendation came from you 7 years ago, it would be great to have your feedback on https://review.openstack.org/375333 | 09:12 |
| *** kaisers_ has joined #openstack-nova | 09:12 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: libvirt: fixes python 3 related unit tests https://review.openstack.org/278834 | 09:13 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes python 3 unit tests https://review.openstack.org/278810 | 09:13 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: hacking: Always use 'assertIs(Not)None' https://review.openstack.org/363623 | 09:14 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: numa: Fixes NUMA topology related unit tests https://review.openstack.org/278835 | 09:17 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: hacking: Reword in the imperative https://review.openstack.org/363624 | 09:18 |
| *** GordonTX has joined #openstack-nova | 09:20 | |
| openstackgerrit | Stephen Finucane proposed openstack/nova: Use more specific asserts in tests https://review.openstack.org/363043 | 09:21 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: hacking: Use assertIs(Not), assert(True|False) https://review.openstack.org/363685 | 09:21 |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs: Add notifications for Flavors https://review.openstack.org/375346 | 09:22 |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs: Add notifications for server groups https://review.openstack.org/375316 | 09:24 |
| *** GordonTX has quit IRC | 09:25 | |
| openstackgerrit | Zhenyu Zheng proposed openstack/nova-specs: Add notifications for Flavors https://review.openstack.org/375346 | 09:25 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: conf: Deprecate all topic options https://review.openstack.org/351608 | 09:27 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: conf: Trivial cleanup of 'cells' https://review.openstack.org/374672 | 09:27 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: conf: Trivial cleanup of compute.py https://review.openstack.org/374673 | 09:27 |
| *** kaisers_ has quit IRC | 09:28 | |
| *** amotoki has quit IRC | 09:32 | |
| *** jlanoux has joined #openstack-nova | 09:33 | |
| *** cdent has quit IRC | 09:35 | |
| openstackgerrit | Hussain Chachuliya proposed openstack/nova: Remove unused code https://review.openstack.org/375356 | 09:36 |
| *** EinstCrazy has joined #openstack-nova | 09:36 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Hyper-V: Adds vNUMA implementation https://review.openstack.org/282407 | 09:37 |
| *** kaisers_ has joined #openstack-nova | 09:42 | |
| openstackgerrit | Hussain Chachuliya proposed openstack/nova: Remove unused code https://review.openstack.org/375356 | 09:42 |
| openstackgerrit | Hussain Chachuliya proposed openstack/nova: Remove redundant str typecasting https://review.openstack.org/375359 | 09:44 |
| *** kaisers_ has quit IRC | 09:46 | |
| johnthetubaguy | paul-carlton2: ah, OK, I think thats a good thing to do regardless | 09:46 |
| johnthetubaguy | paul-carlton2: ah, I see in the bug now, yeah, thats a bit nasty with the delete of stuff | 09:47 |
| *** GB21 has joined #openstack-nova | 09:49 | |
| *** sridharg has quit IRC | 09:50 | |
| *** zenoway has quit IRC | 09:51 | |
| openstackgerrit | Hussain Chachuliya proposed openstack/nova: Replace uuid4() with generate_uuid() from oslo_utils https://review.openstack.org/373794 | 09:53 |
| *** gouthamr has joined #openstack-nova | 09:53 | |
| danpb | for live upgrades, do we require that neutron is upgraded *before* nova is upgraded ? | 09:53 |
| * danpb thinks we do, but can't see it explicitly mentioned anywhere | 09:54 | |
| *** amotoki has joined #openstack-nova | 09:54 | |
| bauwser | paul-carlton2: sorry, busy Friday for me :( | 09:54 |
| *** amotoki has quit IRC | 09:55 | |
| bauwser | paul-carlton2: johnthetubaguy: okay, I think your updated description better defines what's a problem | 09:56 |
| johnthetubaguy | danpb: I didn't think we specified that order, i.e. we usually have supported older Cinder, Neutron, etc | 09:56 |
| johnthetubaguy | danpb: given the changes I am proposing this cycle, we probably need to get crisper about our levels of support there, but I would certainly expect us to work with N-1 | 09:57 |
| *** alexchadin has joined #openstack-nova | 09:57 | |
| sfinucan | danpb: I thought the promise was support for a maximum of one generation difference between projects? i.e. N of nova <-> M of neutron and vice versa | 09:58 |
| *** kaisers_ has joined #openstack-nova | 09:59 | |
| *** kaisers_ has quit IRC | 09:59 | |
| *** kaisers_ has joined #openstack-nova | 10:00 | |
| *** kaisers__ has joined #openstack-nova | 10:03 | |
| *** sridharg has joined #openstack-nova | 10:03 | |
| paul-carlton2 | johnthetubaguy, bauwser looking at this further I think there is a way to fix the immediate issue with evacuate by catching the exception raised by spawn in the compute manager and calling the driver again to destroy the instance and delete the files (if they are not on shared storage) | 10:04 |
| *** TravT has quit IRC | 10:04 | |
| bauwser | paul-carlton2: that's different from what you mention in the bug report | 10:04 |
| *** sridharg has quit IRC | 10:05 | |
| bauwser | but yeah, I think we should not modify where we say instance.host = xxx and rather make sure that the driver, when excepting, can be able to remove the temporary files | 10:05 |
| *** sridharg has joined #openstack-nova | 10:05 | |
| *** cdent has joined #openstack-nova | 10:05 | |
| *** kaisers_ has quit IRC | 10:06 | |
| johnthetubaguy | paul-carlton2: so if destroy is written defensively enough, that is probably good enough for all instances we put into an ERROR state during spawn | 10:06 |
| bauwser | the point is, we *evacuate* | 10:07 |
| bauwser | which means the instance is already gone | 10:07 |
| paul-carlton2 | Yes my concern is that changing the the libvirt driver to delete the files when it fails to spawn that would be a bad thing if shared storage is uswed | 10:07 |
| paul-carlton2 | ! | 10:07 |
| bauwser | okay, I was confusing | 10:08 |
| bauwser | I mean we should rather accept to consider the instance as being destroyable | 10:08 |
| paul-carlton2 | bauwser, migrate/resize deals with this differently, it explicitly cleans up the files it copied from source if the instance fails to start on target | 10:09 |
| bauwser | that's a very different usecase | 10:10 |
| paul-carlton2 | but evacuate doesn't copy files, it just calls rebuild which calls spawn on the target | 10:10 |
| sfinucan | johnthetubaguy: What's the chances of moving release notes into per-release folders? | 10:10 |
| sfinucan | a la nova-specs. It's already getting mighty messy in there... | 10:10 |
| *** sdague has joined #openstack-nova | 10:11 | |
| bauwser | paul-carlton2: resize is an user API based on the assumption that your instance is still alive | 10:11 |
| bauwser | paul-carlton2: evacuate is some of eletrical shock for an admin only saying their want to resurrect some | 10:11 |
| openstackgerrit | Maciej Szankin proposed openstack/nova: Improve consistency in serial_console https://review.openstack.org/348902 | 10:11 |
| paul-carlton2 | trouble is driver spawn doesn't know if it is part of a rebuild for evacuate or normal rebuild, boot or unshelve and worse still it doesn't know if the instance files are on shared storage | 10:12 |
| paul-carlton2 | so it could end up deleting the evacuating instance's files on shared storage | 10:12 |
| bauwser | sfinucan: the big concern to me is that we'd loose the benefit of backportable patches including reno files | 10:12 |
| *** gcb has quit IRC | 10:12 | |
| bauwser | sfinucan: reno doesn't work thru paths, rather looking at the git repo | 10:13 |
| bauwser | I understand that being messy, but I think reno is not aimed to be read directly by the YAML files | 10:13 |
| paul-carlton2 | bauwser, johnthetubaguy which is a nightmare scenario, we have users who use shared storage and use evacuate to recreate instance on a new node during compute node reboots | 10:14 |
| bauwser | sfinucan: instead, you have to look at the generated view | 10:14 |
| sfinucan | bauwser: It'd just be a case of 'cherry-pick --no-commit' though, right? | 10:14 |
| sfinucan | then move it | 10:14 |
| bauwser | sfinucan: you mean that our stable backports should be modified ? gosh :p | 10:14 |
| bauwser | I know it's feasible | 10:14 |
| sfinucan | bauwser: Hmm, idk | 10:15 |
| sfinucan | I just think code (even YAML) is for people to read | 10:15 |
| bauwser | I just say that the whole reno design is made on the assumption that we ship our relnote with the patch, so that backporting a change that include a relnote also includes the relnote | 10:15 |
| *** tuanla has quit IRC | 10:15 | |
| bauwser | sfinucan: that's *really* not the intent :p | 10:15 |
| bauwser | sfinucan: what we *could* is to point those references in trees | 10:16 |
| sfinucan | bauwser: yup, relnotes in code is a good move for sure. It would just be great if we could structure it a little better :) | 10:17 |
| sfinucan | trees? | 10:17 |
| sfinucan | kittens belong in trees - not relnotes | 10:17 |
| paul-carlton2 | johnthetubaguy, so in theory an instance being evacuated in an environment with XenAPI (HyperV too I think) and instances on shared storage would have its instance files deleted if the spawn failed on the target! | 10:17 |
| bauwser | sfinucan: like I said, it's pretty well structured if you read the generated output :) | 10:18 |
| sfinucan | bauwser: Yeah, it is in fairness http://docs.openstack.org/releasenotes/nova/unreleased.html | 10:19 |
| johnthetubaguy | paul-carlton2: we don't support shared storage in XenAPI, AFAIK | 10:20 |
| sfinucan | Hmm, let me think on it more. If there was a change, it would need upstream (reno) support anyway. I'll check out that codebase | 10:20 |
| *** eandersson has quit IRC | 10:20 | |
| paul-carlton2 | johnthetubaguy, ok, wondered about that, good | 10:20 |
| sfinucan | On a totally unrelated note, can someone please explain to me why I'd use 'stub_out' vs. 'mock' in tests? | 10:21 |
| *** kaisers__ has quit IRC | 10:23 | |
| openstackgerrit | Lucian Petrut proposed openstack/nova: Add Hyper-V storage QoS support https://review.openstack.org/175479 | 10:24 |
| *** EinstCrazy has quit IRC | 10:31 | |
| *** phuongnh has quit IRC | 10:33 | |
| *** swamireddy has quit IRC | 10:35 | |
| bauwser | sfinucan: stub_out is an helper for getting out of a mox thing that was not good for py3 | 10:47 |
| johnthetubaguy | sfinucan: stub_out was just to help us transition from the old stub method from pre-mock, AFAIK | 10:47 |
| bauwser | sfinucan: it's better to use mock for a new UT | 10:48 |
| sdague | sfinucan: right, it was based on the analysis that there were a ton of unit tests that didn't use any mox except Stubs | 10:48 |
| *** kaisers_ has joined #openstack-nova | 10:48 | |
| sdague | and there was a close enough way to do the same thing with mock | 10:48 |
| openstackgerrit | Alex Szarka proposed openstack/nova-specs: Resume Guests State upgrade https://review.openstack.org/375388 | 10:48 |
| sfinucan | So use 'mock' ideally, but 'stub_out' if I have to? | 10:50 |
| openstackgerrit | John Garbutt proposed openstack/nova: Fixed deprecation status for xvp options https://review.openstack.org/353658 | 10:50 |
| sdague | sfinucan: right, use mock ideally. | 10:50 |
| bauwser | sfinucan: stub_out if you want to help a current UT | 10:50 |
| bauwser | mock if anything else | 10:51 |
| sdague | using stub_out is just about converting existing tests without a ton of churn (code or mental) | 10:51 |
| openstackgerrit | Alex Szarka proposed openstack/nova-specs: Resume Guests State upgrade https://review.openstack.org/375388 | 10:51 |
| sdague | johnthetubaguy / bauwser - https://review.openstack.org/#/c/375037/ easy test output cleanup | 10:51 |
| openstackgerrit | Alex Szarka proposed openstack/nova-specs: Resume Guests State upgrade https://review.openstack.org/375388 | 10:52 |
| * bauwser looking | 10:53 | |
| bauwser | oooh I se | 10:53 |
| bauwser | see | 10:53 |
| *** GB21 has quit IRC | 10:53 | |
| sfinucan | sdague: Ideal. Cheers | 10:54 |
| *** Guest56480 has quit IRC | 11:00 | |
| *** sudswas__ has quit IRC | 11:00 | |
| openstackgerrit | Sean Dague proposed openstack/nova: Capture stdout in for test_wsgi:test_debug https://review.openstack.org/375397 | 11:04 |
| *** lucas-afk is now known as lucasagomes | 11:05 | |
| *** eandersson has joined #openstack-nova | 11:10 | |
| openstackgerrit | Stephen Finucane proposed openstack/nova: relnotes: Fix config options deprecation notes https://review.openstack.org/375400 | 11:10 |
| *** smatzek has joined #openstack-nova | 11:12 | |
| *** alexchadin has quit IRC | 11:13 | |
| *** sahid has quit IRC | 11:14 | |
| mdbooth | artom: Can you have another look at https://review.openstack.org/#/c/369176/ for me. It looks so wrong to me I'm worried I've missed something obvious. | 11:15 |
| *** smatzek has quit IRC | 11:17 | |
| *** kaisers_ has quit IRC | 11:18 | |
| *** bvanhav has joined #openstack-nova | 11:19 | |
| hussain_c | alaski: Hi Andrew, can you please take a look.. I re-uploaded the patch after fixing merge conflict. https://review.openstack.org/#/c/373794/ | 11:21 |
| *** aswadr_ has quit IRC | 11:22 | |
| *** rtheis has joined #openstack-nova | 11:22 | |
| *** hoangcx has quit IRC | 11:23 | |
| hussain_c | alaski: I have removed some unused method in this patch. Please review. https://review.openstack.org/#/c/375356/ | 11:24 |
| *** kaisers_ has joined #openstack-nova | 11:27 | |
| *** kaisers__ has joined #openstack-nova | 11:28 | |
| *** kaisers_ has quit IRC | 11:32 | |
| *** amotoki has joined #openstack-nova | 11:38 | |
| mdbooth | hussain_c: More, please :) | 11:38 |
| mdbooth | hussain_c: FYI, I don't think TrivialFix ever took off, unfortunately. | 11:39 |
| mdbooth | There's no harm in it, but afaik there are no cores filtering on it. | 11:39 |
| *** jaypipes has joined #openstack-nova | 11:42 | |
| openstackgerrit | Chris Dent proposed openstack/nova: [placement] ensure that allow headers are native strings https://review.openstack.org/374800 | 11:43 |
| *** gouthamr has quit IRC | 11:44 | |
| cdent | sdague: if you get a chance to go back to this one https://review.openstack.org/#/c/371370/ it is now using a new gabbi so the gabbi is gone from the test name: http://logs.openstack.org/70/371370/2/check/gate-nova-tox-db-functional-ubuntu-xenial/b3f05c6/testr_results.html.gz | 11:44 |
| sdague | cdent: cool, thanks | 11:46 |
| cdent | sdague: the global req change has merged | 11:47 |
| artom | mdbooth, no, at first glance I you're right. It didn't dawn on me that ephemeral disks get their size from the flavor, so 0 size disks were "possible." | 11:48 |
| *** andreas_s has quit IRC | 11:49 | |
| *** jaypipes is now known as leakypipes | 11:50 | |
| openstackgerrit | Sean Dague proposed openstack/nova: Clean up stdout/stderr leakage in cmd testing https://review.openstack.org/375414 | 11:53 |
| *** sabari has quit IRC | 11:56 | |
| mdbooth | artom: Can you comment on the reviews? I also commented on your temptest test, btw. | 11:57 |
| *** jfding has quit IRC | 11:58 | |
| mdbooth | artom: For the tempest test, btw, wouldn't you have to skip it if flavor 1 doesn't have 2G of ephemeral disk in order to be fully tempesty? | 11:59 |
| *** dave-mccowan has quit IRC | 12:00 | |
| *** kaisers__ has quit IRC | 12:01 | |
| *** thorst has joined #openstack-nova | 12:03 | |
| artom | mdbooth, yeah, let me just get the kids daycare sorted out :) | 12:05 |
| artom | (And some coffee too, ideally) | 12:05 |
| mdbooth | artom: Which of those is the priority? | 12:05 |
| mdbooth | That's a tough one. | 12:05 |
| artom | mdbooth, err, yes. | 12:06 |
| artom | Those is the priority. | 12:06 |
| *** zul has quit IRC | 12:07 | |
| *** GB21 has joined #openstack-nova | 12:07 | |
| mdbooth | artom: I'm worried that without coffee you might arrive at daycare having forgotten the kids. | 12:07 |
| *** thorst has quit IRC | 12:08 | |
| *** catintheroof has quit IRC | 12:08 | |
| prateek | sdague, did you get a chance to write a spec for https://review.openstack.org/#/c/327706/ ? | 12:09 |
| prateek | johnthetubaguy, can you please take a look at https://review.openstack.org/#/c/294513/ ? | 12:09 |
| artom | mdbooth, then I'll stay in their stead :D | 12:10 |
| *** gouthamr has joined #openstack-nova | 12:11 | |
| *** maticue has joined #openstack-nova | 12:16 | |
| *** artom has quit IRC | 12:16 | |
| *** artom has joined #openstack-nova | 12:17 | |
| *** artom has quit IRC | 12:18 | |
| *** sc68cal_ is now known as sc68cal | 12:18 | |
| *** artom has joined #openstack-nova | 12:18 | |
| openstackgerrit | Béla Vancsics proposed openstack/nova: Reduced the complexity of the _init_instance method https://review.openstack.org/374661 | 12:19 |
| *** thorst has joined #openstack-nova | 12:19 | |
| bauwser | sdague: is https://review.openstack.org/#/c/374307/4 somehow impacted by https://review.openstack.org/#/c/362438/ ? | 12:20 |
| bauwser | sdague: I can see grenade failing on it | 12:20 |
| *** mriedem has joined #openstack-nova | 12:22 | |
| *** sabari has joined #openstack-nova | 12:24 | |
| *** thorst has quit IRC | 12:24 | |
| *** Rod2a has joined #openstack-nova | 12:24 | |
| mriedem | superdan: you want to update the link in the channel topic to this? https://wiki.openstack.org/wiki/Nova/Ocata_Release_Schedule | 12:25 |
| Rod2a | hello everybody | 12:25 |
| *** prateek has quit IRC | 12:26 | |
| Rod2a | I got a problem with nova-compute wich is loosing connection with my controller node | 12:26 |
| Rod2a | I use RDO kilo release | 12:26 |
| Rod2a | the only error message in compute is nova.servicegroup.drivers.db "model server went away" | 12:27 |
| sdague | bauwser: it's because we haven't branched grenade / devstack yet | 12:27 |
| bauwser | okay, that's what I thought | 12:27 |
| Rod2a | I can see that rabbitmq unack msg is growing up until I reboot nova-compute | 12:28 |
| bauwser | sdague: and we need https://review.openstack.org/#/c/362438/ before branching stable grenade, right? | 12:28 |
| Rod2a | does anybody have a clue ? | 12:28 |
| Rod2a | there is no load on any server, no error in rabbit logs... | 12:29 |
| bauwser | Rod2a: that looks like an operating problem related to rabbit message queue, so probably #openstack could be better for discussing | 12:29 |
| Rod2a | tcp connection are keepalived | 12:29 |
| sdague | bauwser: actually, I think we probably want to branch first | 12:29 |
| sdague | I was just thinking through that all | 12:29 |
| bauwser | okay | 12:29 |
| *** thorst has joined #openstack-nova | 12:29 | |
| Rod2a | bauwser: cinder and glance are still ok so I think rabbitmq is OK | 12:30 |
| Rod2a | bauwser: only server creation is broken | 12:30 |
| leakypipes | sdague: thx much for your log cleanup patches. really helpful in keeping my sanity when running tests :) | 12:31 |
| Rod2a | I use vmware driver to access vsphere 5.1 cluster, my vcenter is OK, no error and the API of vcenter seems ok (I request the sdk wsdl file to ensure) | 12:31 |
| sdague | leakypipes: no problem. Sometimes you just need to do the spring cleaning that's been pissing you off for years :) | 12:31 |
| *** zul has joined #openstack-nova | 12:32 | |
| sdague | leakypipes: after that last patch of mine there is one left that I'm not sure I can get rid of, then the oslo_db / pkg_resources / uuid warnings that are all valid and not something we can supress | 12:33 |
| *** artom has quit IRC | 12:33 | |
| sdague | but, basically everything in that stream now becomes something people should care about | 12:33 |
| *** yassine has joined #openstack-nova | 12:33 | |
| bauwser | Rod2a: server creation is broken because the scheduler checks the server health by looking at what we call the servicegroup API | 12:33 |
| *** yassine is now known as Guest37187 | 12:34 | |
| bauwser | Rod2a: in general, those healthchecks are using a simple DB call, verifying if some field is updated | 12:34 |
| mriedem | sdague: a pleasant surprise to see merged this morning https://review.openstack.org/#/c/374433/ | 12:34 |
| Rod2a | we use this plateform in production since nearly one year with about 500 actives VM and the problem occurs more often | 12:34 |
| sdague | leakypipes: 2 last ones on that series https://review.openstack.org/#/q/topic:fix_tests+status:open | 12:34 |
| sdague | mriedem: yep | 12:35 |
| bauwser | Rod2a: the compute nodes are periodically updating that field, but they don't update the DB directly, it goes thru the conductor | 12:35 |
| bauwser | Rod2a: for sending the DB update to the conductor, it pass thru a RPC cast | 12:35 |
| *** artom has joined #openstack-nova | 12:35 | |
| mriedem | lpetrut: can you get the unit tests passing on this? https://review.openstack.org/#/c/175479/ | 12:36 |
| bauwser | Rod2a: which means that if your queue is stuck or full, then it won't update correctly that field for that compute and then the scheduler will see that compute as "down" | 12:36 |
| bauwser | Rod2a: usual checks go thru checking the service state for that specific compute and also check the field (if using the DB SG driver) | 12:36 |
| bauwser | Rod2a: but all of that is very operational, hence me telling you that you should get better help for troubleshooting by moving to #openstack | 12:37 |
| bauwser | (see channel topic too) | 12:37 |
| lpetrut | @mriedem: sure, sending a new patch set right away | 12:37 |
| bauwser | lpetrut: oh gosh, we're not using Slack fortunately :p | 12:38 |
| *** rcernin has quit IRC | 12:38 | |
| mriedem | lpetrut: also i replied in https://review.openstack.org/#/c/175479/23/nova/virt/hyperv/vmops.py@884 | 12:39 |
| mriedem | i still think you're going to be setting disk qos specs even if nothing is specified b/c of the logic in there | 12:39 |
| *** daemontool has joined #openstack-nova | 12:40 | |
| *** Rod2a has quit IRC | 12:41 | |
| *** takedakn has quit IRC | 12:42 | |
| *** edleafe is now known as figleaf | 12:43 | |
| *** ducttape_ has joined #openstack-nova | 12:44 | |
| *** pcaruana|afk| has joined #openstack-nova | 12:44 | |
| mriedem | lpetrut: nevermind that last one, i've got it now | 12:44 |
| lpetrut | bytes_per_sec_to_iops(0) will be 0, I guess that was the misleading thing, right? | 12:45 |
| *** ducttape_ has quit IRC | 12:45 | |
| *** ducttape_ has joined #openstack-nova | 12:46 | |
| *** artom has quit IRC | 12:46 | |
| *** artom has joined #openstack-nova | 12:46 | |
| *** thorst has quit IRC | 12:46 | |
| *** alexchadin has joined #openstack-nova | 12:46 | |
| *** artom has quit IRC | 12:47 | |
| *** artom has joined #openstack-nova | 12:47 | |
| *** sdake has joined #openstack-nova | 12:48 | |
| mriedem | yeah | 12:48 |
| mriedem | 0 + 8192 - 1 // 8192 | 12:48 |
| leakypipes | sdague: done. nice work on both :) | 12:48 |
| *** armax has quit IRC | 12:49 | |
| openstackgerrit | Lucian Petrut proposed openstack/nova: Add Hyper-V storage QoS support https://review.openstack.org/175479 | 12:51 |
| *** GB21 has quit IRC | 12:54 | |
| *** mamandle has quit IRC | 12:54 | |
| mriedem | lpetrut: sorry just missed my comments in ps24 | 12:54 |
| mriedem | a couple of small issues | 12:54 |
| *** mamandle has joined #openstack-nova | 12:54 | |
| *** mriedem has quit IRC | 12:55 | |
| *** david-lyle has joined #openstack-nova | 12:56 | |
| *** prateek has joined #openstack-nova | 12:58 | |
| hussain_c | alaski: Thanks for the review.. :) Can you please take a look at this trivial fix? https://review.openstack.org/#/c/375360 | 12:59 |
| *** tblakes has joined #openstack-nova | 12:59 | |
| *** mamandle_ has joined #openstack-nova | 12:59 | |
| *** READ10 has joined #openstack-nova | 12:59 | |
| *** mamandle has quit IRC | 12:59 | |
| *** pilgrimstack has quit IRC | 12:59 | |
| *** mamandle_ is now known as mamandle | 12:59 | |
| *** thorst has joined #openstack-nova | 12:59 | |
| lpetrut | mriedem: no worries, thanks for reviewing this patch! I'm fixing the nits right now | 13:00 |
| *** pcaruana|afk| has quit IRC | 13:00 | |
| *** kaisers_ has joined #openstack-nova | 13:01 | |
| *** pcaruana has quit IRC | 13:01 | |
| mdbooth | danpb: Aren't we trying to move away from libvirt device names? | 13:02 |
| *** pcaruana has joined #openstack-nova | 13:02 | |
| openstackgerrit | Stephen Finucane proposed openstack/nova: objects: Move 'arch' to 'fields.Architecture' https://review.openstack.org/356067 | 13:03 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: objects: Move 'vm_mode' to 'fields.VMMode' https://review.openstack.org/356553 | 13:03 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: objects: Move 'hv_type' to 'fields.HVType' https://review.openstack.org/356424 | 13:03 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: objects: Move 'cpumodel' to 'fields.CPU*' https://review.openstack.org/356398 | 13:03 |
| *** mjura has quit IRC | 13:03 | |
| *** thorst has quit IRC | 13:04 | |
| *** rtheis has quit IRC | 13:05 | |
| *** rcernin has joined #openstack-nova | 13:06 | |
| danpb | mdbooth: away from users /choosing/ them - we still need a way to unique identify disks to users hough in some manner | 13:06 |
| *** prateek_ has joined #openstack-nova | 13:06 | |
| mdbooth | danpb: Just commented again, btw, after remembering that device_name can change. | 13:06 |
| *** rtheis has joined #openstack-nova | 13:06 | |
| openstackgerrit | Tony Xu proposed openstack/nova: Fix pep8 E501 line too long https://review.openstack.org/375231 | 13:07 |
| *** prateek has quit IRC | 13:07 | |
| alaski | hussain_c: that's not a nova patch. And you can use the "Add..." button to add me to a review if you want me to take a look rather than pinging me here each time | 13:07 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: hacking: Use assertIs(Not), assert(True|False) https://review.openstack.org/363685 | 13:08 |
| *** pilgrimstack has joined #openstack-nova | 13:08 | |
| mdbooth | danpb: How would you feel about explicitly exposing hypervisor and machine type to users, btw? | 13:09 |
| *** alaski is now known as lascii | 13:09 | |
| mdbooth | Given that these things can be inferred anyway. | 13:09 |
| *** rlrossit has joined #openstack-nova | 13:09 | |
| leakypipes | mdbooth: -10 | 13:11 |
| danpb | mdbooth: exposing where ? | 13:11 |
| mdbooth | leakypipes: Explain. I know we explicitly don't do it, but why? | 13:11 |
| mdbooth | danpb: In the diagnostics, for eg. Perhaps even instance metadata? | 13:12 |
| leakypipes | mdbooth: because it's a cloud. users should not need or want to know hardware-specific or hypervisor-specific details about where their instances are located. | 13:12 |
| danpb | leakypipes: we already allow users to /set/ those in image properties | 13:12 |
| mdbooth | leakypipes: My thinking is that these things are part of the interface that we expose to the user. | 13:12 |
| danpb | so refusing to report the value of something we already let the user set makes little sense | 13:13 |
| *** jaosorior has quit IRC | 13:13 | |
| mdbooth | Specifically thinking of non-admin users, btw | 13:13 |
| leakypipes | danpb: on the image properties, yes. but I'm thinking more regarding the stuff returned in a GET /servers/{uuid} call. | 13:13 |
| *** mamandle has quit IRC | 13:13 | |
| mdbooth | leakypipes: So yeah, it's still cloudy. The end-user doesn't get to ask for something specific. However they might reasonably want to know what they got. | 13:14 |
| *** jaosorior has joined #openstack-nova | 13:14 | |
| mdbooth | leakypipes: Also, they can already find that out if they want to. | 13:14 |
| *** jamesdenton has joined #openstack-nova | 13:14 | |
| mdbooth | Why wouldn't we make it slightly easier? | 13:14 |
| leakypipes | mdbooth: in what way? | 13:14 |
| mdbooth | Which bit? | 13:14 |
| leakypipes | mdbooth: can they "already find that out if they want to"? | 13:14 |
| mdbooth | e.g. the virt-what tool | 13:14 |
| danpb | leakypipes: the guest os can probe it to some degree | 13:15 |
| mdbooth | Will tell you which hypervisor you're running on | 13:15 |
| danpb | its not pretty, but its ultimately doable | 13:15 |
| leakypipes | heh, ok :) | 13:15 |
| mdbooth | Probing devices will tell you what they are, and you can infer machine type. | 13:15 |
| leakypipes | well, like I said I'd prefer to keep as much of the abstraction from hw and hv as possible. | 13:15 |
| dhellmann | sfinucan, bauwser : wow, nova has a lot of notes. If someone wants to add support for subdirectories under the notes dir that would be OK with me. I'm not sure how hard it would be off the top of my head. | 13:15 |
| danpb | it is useful troubleshooting info at the very least | 13:16 |
| bauwser | dhellmann: well, that's the problem I mentioned related to backports | 13:16 |
| sfinucan | dhellmann: Maybe we're *that* good at documentation... | 13:16 |
| sfinucan | o.O | 13:16 |
| sfinucan | :) | 13:16 |
| mdbooth | leakypipes: If a guest knows it's running on hypervisor X rather than hypervisor Y, it might configure itself to run optimally. | 13:16 |
| danpb | eg user has 2 vms running, and wondering why they have slightly different behaviour / characteristics | 13:16 |
| bauwser | dhellmann: I wouldn't like to have backports being different just because of a reno file | 13:16 |
| sfinucan | bauwser: Yeah, I've thought about it a little more and we're better keeping it in one directory | 13:16 |
| danpb | being able to see the virt/machine type is useful info for diagnosis | 13:16 |
| openstackgerrit | Chris Dent proposed openstack/nova: [placement] 404 responses do not cause exception logs https://review.openstack.org/374955 | 13:17 |
| openstackgerrit | Chris Dent proposed openstack/nova: [placement] 404 responses do not cause exception logs https://review.openstack.org/374955 | 13:17 |
| leakypipes | mdbooth, danpb: not necessarily disagreeing that it might be useful diagnostic information, agreed. I guess a more specific example of what bits/pieces of information would be useful for a user to know to "configure their instance" would be useful. | 13:17 |
| sfinucan | bauwser: dhellmann: Having written changelogs that follow the keepachangelog.com standard, I've seen how tough it is to version these things otherwise | 13:17 |
| dhellmann | bauwser : yeah, you might not want the dirs to be named for series. maybe the first 2 chars of the random bit of the filename? | 13:17 |
| sfinucan | dhellmann: that could work... | 13:18 |
| dhellmann | might be overkill, I dunno | 13:18 |
| leakypipes | mdbooth, danpb: I mean, are we talking about things like x86 CPU instruction set extensions here? | 13:18 |
| bauwser | dhellmann: or thinking about leaving all the notes by the same subdir, but having some way in tree to have indirections for each cycle ? | 13:18 |
| leakypipes | mdbooth, danpb: or are we talking about things as specific as, say, the version of libvirt/QEMU running on the host? | 13:18 |
| bauwser | I dunno | 13:18 |
| mdbooth | leakypipes: So, after you agree that this is already visible in the user interface, you then get to wondering how much to expose and how to expose it :) | 13:19 |
| bauwser | I honestly thought about that, and nothing but saying "please rather look at the generated output" does seem better | 13:19 |
| mdbooth | I don't think libvirt version is relevant, beause it's not part of the user interface. | 13:19 |
| leakypipes | mdbooth, danpb: because the latter I see as a bit too much information and potentially a security issue if the guest can query the host for a potentially-insecure version of the hypervisor? | 13:19 |
| *** mtanino has joined #openstack-nova | 13:19 | |
| mdbooth | I don't think qemu version is something we would want to expose for the reason you say | 13:19 |
| bauwser | dhellmann: or, maybe a debug mode for reno that would mention the filename for each output | 13:19 |
| mdbooth | However qemu machine type probably is | 13:19 |
| bauwser | filename(s) even | 13:19 |
| leakypipes | mdbooth: gotcha. | 13:20 |
| danpb | leakypipes: security through obscurity is no security at all | 13:20 |
| dhellmann | bauwser : why don't we talk about it in #openstack-release | 13:20 |
| bauwser | dhellmann: because the only usecase for checking the YAML files is when you see some mistake in the generated output and you want to fix it, so Ctrl-F is the only possibility | 13:20 |
| bauwser | +1 | 13:20 |
| leakypipes | mdbooth, danpb: of course, I'm thinking in the back of my head here about the potential os-traits library and how/what information could be communicated to an end-user in a safe fashion. | 13:20 |
| dhellmann | bauwser : actually, I have to step offline for a bit. I'll be back in ~45 minuters | 13:21 |
| danpb | that said, i don't think we need to report software versions | 13:21 |
| *** xyang1 has joined #openstack-nova | 13:21 | |
| leakypipes | danpb: I agree with you on that. However, I'm just playing devil's advocate a bit here :) | 13:21 |
| *** jamesdenton has quit IRC | 13:21 | |
| bauwser | dhellmann: np | 13:22 |
| mdbooth | leakypipes: Anyway, it was just a thought I had whilst reading that spec. I know we don't do it, but I wonder if we should. | 13:22 |
| *** parora has joined #openstack-nova | 13:22 | |
| bauwser | dhellmann: I'll have to step out at 1400UTC, but we can async discuss that | 13:22 |
| danpb | things like the hypervisor type, and machine type are fundamentally not secret though as they're guest ABI | 13:22 |
| *** eharney has joined #openstack-nova | 13:22 | |
| *** jamesdenton has joined #openstack-nova | 13:23 | |
| mdbooth | Does anybody still use PV xen? | 13:24 |
| *** prateek_ has quit IRC | 13:24 | |
| *** pilgrimstack has quit IRC | 13:24 | |
| johnthetubaguy | in general, yes | 13:24 |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Adds Hyper-V OVS ViF driver https://review.openstack.org/140045 | 13:24 |
| *** pilgrimstack has joined #openstack-nova | 13:24 | |
| bauwser | cdent: is https://review.openstack.org/#/c/374870/2 something defined by the API WG ? | 13:24 |
| johnthetubaguy | particularly now PV driver bits are build into the kernel | 13:24 |
| openstackgerrit | Lucian Petrut proposed openstack/nova: Add Hyper-V storage QoS support https://review.openstack.org/175479 | 13:25 |
| *** mdrabe has joined #openstack-nova | 13:25 | |
| cdent | bauwser: looking | 13:25 |
| mdbooth | That's a classic example of where the guest totally needs to know what its hypervisor is. | 13:25 |
| bauwser | cdent: I mean, having the root name == the ending / | 13:25 |
| mdbooth | Because otherwise it won't boot. | 13:25 |
| *** prateek_ has joined #openstack-nova | 13:25 | |
| johnthetubaguy | mdbooth: well its all auto detected in the kernel | 13:25 |
| johnthetubaguy | mdbooth: well 3.x and onwards | 13:25 |
| cdent | bauwser: no, the api-wg hasn't touched on that, yet, in part because there hasn't been much discussion about mixing services in the same web server | 13:25 |
| mdbooth | johnthetubaguy: You may infer from that, then, that I haven't used PV xen myself in some time :) | 13:26 |
| *** rcernin has quit IRC | 13:26 | |
| mdbooth | In my day... | 13:26 |
| cdent | the issue only shows up when there is a prefix (as in devstack for placement) | 13:26 |
| bauwser | cdent: I mean, I understand the motivation, and I appreciate it, but I think it would be cool if we could clarify that "officially" as a guideline | 13:26 |
| johnthetubaguy | mdbooth: yeah, no more recompile crap | 13:26 |
| cdent | bauwser: we can create a bug at https://bugs.launchpad.net/openstack-api-wg to remind ourself to do something about it? | 13:26 |
| bauwser | cdent: I just don't want to +A something I could regret in the future because of the API contract | 13:26 |
| johnthetubaguy | mdbooth: it now uses the same system that make it go faster on KVM, AFAIK | 13:26 |
| *** parora has quit IRC | 13:27 | |
| cdent | bauwser: under what circumstances would you _want_ that to 404? | 13:27 |
| mdbooth | johnthetubaguy: I'm thinking of the Xen that existed before hardware virt was a commodity thing. | 13:27 |
| bauwser | cdent: I don't know, that's just that accepting that means accepting that *for the eternity* | 13:27 |
| bauwser | cdent: the *for the eternity* thing is chilling me | 13:28 |
| johnthetubaguy | mdbooth: yeah, when the Xen team where helping build that with Intel and AMD | 13:28 |
| *** prithiv has joined #openstack-nova | 13:28 | |
| bauwser | sdague: thoughts on that https://review.openstack.org/#/c/374870/2 ? | 13:28 |
| *** parora has joined #openstack-nova | 13:28 | |
| cdent | bauwser: let's see if sdague and alex_xu have opinions? | 13:28 |
| mdbooth | johnthetubaguy: And RH used Xen :) | 13:28 |
| *** eandersson has quit IRC | 13:29 | |
| *** sahid has joined #openstack-nova | 13:29 | |
| *** gszasz has quit IRC | 13:30 | |
| bauwser | cdent: yeah, I'd appreciate | 13:30 |
| openstackgerrit | Sergey Nikitin proposed openstack/nova-specs: Restore V3 Diagnostics info https://review.openstack.org/357884 | 13:31 |
| *** prateek_ has quit IRC | 13:32 | |
| *** tongli has joined #openstack-nova | 13:36 | |
| *** ducttape_ has quit IRC | 13:36 | |
| *** ekarlso_ has joined #openstack-nova | 13:37 | |
| *** annegentle has joined #openstack-nova | 13:39 | |
| bauwser | cdent: FYI, https://review.openstack.org/#/c/371370/2 Depends-On was cool but wasn't preventing to have Nova being trampled in case we were capping the max version | 13:42 |
| bauwser | cdent: because Nova syncs asynchronously the requirements, so we could have taken an earlier version, but fortunatly, we're not capping the max, so all of our placement jobs are running with the latest gabbi | 13:43 |
| openstackgerrit | Ludovic Beliveau proposed openstack/nova: PCI: Avoid looping over PCI devices twice https://review.openstack.org/375489 | 13:43 |
| jianghuaw | johnthetubaguy: could you help to review this patch? | 13:43 |
| jianghuaw | https://review.openstack.org/#/c/366825/ | 13:43 |
| openstackgerrit | Ludovic Beliveau proposed openstack/nova: PCI: Avoid looping over PCI devices twice https://review.openstack.org/375489 | 13:45 |
| *** baoli has joined #openstack-nova | 13:45 | |
| *** baoli_ has joined #openstack-nova | 13:46 | |
| *** kaisers_ has quit IRC | 13:46 | |
| *** akshai has joined #openstack-nova | 13:47 | |
| bauwser | Vek: around ? I was reviewing https://bugs.launchpad.net/python-novaclient/+bug/1616415 and I actually wonder if that's a bug | 13:47 |
| openstack | Launchpad bug 1616415 in python-novaclient "The incorrect output of "nova show" when VM has long user data" [Undecided,In progress] - Assigned to viswesuwara nathan (viswesn) | 13:47 |
| bauwser | Vek: if the user is having trouble with paginating the output, there are couple of ways to do that out of Nova | 13:47 |
| *** kragniz has quit IRC | 13:48 | |
| *** kragniz has joined #openstack-nova | 13:48 | |
| *** zenoway has joined #openstack-nova | 13:49 | |
| *** dave-mccowan has joined #openstack-nova | 13:49 | |
| *** baoli has quit IRC | 13:49 | |
| *** tblakes has quit IRC | 13:50 | |
| openstackgerrit | Sean Dague proposed openstack/nova: properly capture logging during db functional tests https://review.openstack.org/375498 | 13:51 |
| sdague | leakypipes: starting to clean up functional side now ^^^ | 13:52 |
| leakypipes | rock on sdague | 13:52 |
| leakypipes | will review after yummy time | 13:52 |
| *** rcernin has joined #openstack-nova | 13:53 | |
| *** hshiina has joined #openstack-nova | 13:53 | |
| *** brian_price has quit IRC | 13:53 | |
| *** rlrossit has quit IRC | 13:54 | |
| cdent | bauwser: isn't upper-constraints used? | 13:54 |
| *** woodster_ has joined #openstack-nova | 13:54 | |
| *** dansmith is now known as superdan | 13:55 | |
| bauwser | cdent: yup, you're actually right for UTs and functional tests | 13:55 |
| bauwser | my bad | 13:56 |
| openstackgerrit | Simona Iuliana Toader proposed openstack/nova-specs: Adds Hyper-V Shielded VMs spec https://review.openstack.org/274709 | 13:56 |
| *** rlrossit has joined #openstack-nova | 13:56 | |
| *** hussain_c has quit IRC | 13:56 | |
| openstackgerrit | Andrew Laski proposed openstack/nova: Annotate online db migrations with cycle added https://review.openstack.org/374323 | 13:59 |
| *** GordonTX has joined #openstack-nova | 13:59 | |
| *** ecelik has quit IRC | 14:00 | |
| *** GordonTX has quit IRC | 14:01 | |
| *** jheroux has joined #openstack-nova | 14:02 | |
| *** prithiv has quit IRC | 14:02 | |
| *** ducttape_ has joined #openstack-nova | 14:02 | |
| *** Guest56480 has joined #openstack-nova | 14:02 | |
| *** sudswas__ has joined #openstack-nova | 14:02 | |
| *** GordonTX has joined #openstack-nova | 14:02 | |
| *** mriedem has joined #openstack-nova | 14:02 | |
| openstackgerrit | Sean Dague proposed openstack/nova: properly capture logging during db functional tests https://review.openstack.org/375498 | 14:03 |
| openstackgerrit | Sean Dague proposed openstack/nova: remove extra error_log from gabbi test fixtures https://review.openstack.org/375506 | 14:03 |
| *** thorst has joined #openstack-nova | 14:04 | |
| openstackgerrit | Lucian Petrut proposed openstack/nova: Hyper-V: fix image handling when shared storage is being used https://review.openstack.org/347437 | 14:05 |
| *** zenoway has quit IRC | 14:06 | |
| *** ducttape_ has quit IRC | 14:06 | |
| *** parora has quit IRC | 14:06 | |
| *** zenoway has joined #openstack-nova | 14:07 | |
| *** sandanar has joined #openstack-nova | 14:08 | |
| *** yamahata has joined #openstack-nova | 14:09 | |
| *** tblakes has joined #openstack-nova | 14:10 | |
| *** tblakes has quit IRC | 14:11 | |
| *** zenoway has quit IRC | 14:11 | |
| *** tblakes has joined #openstack-nova | 14:11 | |
| cdent | sdague: there's gotta be a better way on https://review.openstack.org/#/c/375506/1 | 14:12 |
| *** daemontool has quit IRC | 14:12 | |
| *** daemontool_ has joined #openstack-nova | 14:12 | |
| *** daemontool__ has joined #openstack-nova | 14:13 | |
| *** cmartin has joined #openstack-nova | 14:13 | |
| sdague | cdent: no, you explicitly bypassed the way we attempt to control output | 14:14 |
| sdague | we need to take that back out | 14:14 |
| cmartin | Good morning (my time) everyone, happy friday | 14:14 |
| cdent | yes! because it _needs_ to be bypassed | 14:14 |
| *** tbachman has joined #openstack-nova | 14:15 | |
| sdague | cdent: then come up with a generic solution for the whole test suite, not just in a corner | 14:15 |
| cdent | sdague: I'm trying to, I ask right there in my comment for us to work together to figure something out, but in the meantime, don't _you_ break things | 14:16 |
| *** rtheis has quit IRC | 14:16 | |
| sdague | cdent: you already broke things, I'm trying to correct it | 14:16 |
| cdent | if you commit that change, when a test fails for unexpected reasons you'll have _nothing_ | 14:16 |
| sdague | no, that's not true | 14:16 |
| sdague | it all gets dumped in the error logs when a test fails | 14:17 |
| *** daemontool_ has quit IRC | 14:17 | |
| *** artom has quit IRC | 14:17 | |
| *** artom has joined #openstack-nova | 14:18 | |
| cdent | did you test that sdague ? | 14:18 |
| sdague | yes | 14:18 |
| cdent | because when I tried to get things working beore I did that workaround, I could not get it work appropriately | 14:19 |
| sdague | here is a patch, you can see | 14:19 |
| openstackgerrit | Sean Dague proposed openstack/nova: WIP: demonstrate what failed tests look like https://review.openstack.org/375516 | 14:19 |
| sdague | that's how those standard fixtures work | 14:19 |
| sdague | pull that, run it, you'll see | 14:19 |
| cdent | I got a traceback with the test failure, but not with the internal exception | 14:19 |
| cdent | I've pulled your change and am testing it with a forced breakage now | 14:19 |
| cdent | sdague: the situation where it doesn't work is if if the code within the wsgi application is failing for some reason | 14:20 |
| cdent | I'll produce an output diff | 14:21 |
| cdent | and see if my fears are true | 14:21 |
| BobBall | mriedem: Could you consider removing procedural -2 from https://review.openstack.org/#/c/333781/ ? It now has a spec-less blueprint open as discussed yesterday | 14:21 |
| sdague | cdent: ok, so that needs to be captured somewhere else. Because that also explains what I've seen where if there is anything broken in wsgi you get nothing in logs | 14:21 |
| sdague | my guess is the wsgi stack needs a better backstop middleware to catch that | 14:21 |
| *** vladikr has joined #openstack-nova | 14:23 | |
| *** zenoway has joined #openstack-nova | 14:23 | |
| *** TravT has joined #openstack-nova | 14:25 | |
| *** thorst has quit IRC | 14:26 | |
| cdent | sdague: so what I've done is raise a typerror in one of the OVO, that causes a 500 in a test | 14:26 |
| *** thorst has joined #openstack-nova | 14:26 | |
| cdent | when your patch is in use we do no get this output anywhere: http://paste.openstack.org/show/582764/ | 14:27 |
| openstackgerrit | Andrew Laski proposed openstack/nova: DNM https://review.openstack.org/375106 | 14:27 |
| cdent | sdague: that's pretty useful output. I don't know what the problem is, and I don't understand the things that the nova fixtures are doing with logging and output capture to figure this out (quickly) solo, but I don't want to lose that info | 14:28 |
| cdent | I'm extremely happy to work with someone to try to figure it out | 14:28 |
| cdent | but right now my battery is dying so I need to get home | 14:28 |
| sdague | so that means that there is an exception thrown for which there is no handler in the stack to catch it and log the uncaught exception, right? | 14:29 |
| sdague | because the expectation is there would be something backstopping the uncaught exceptions and turning them into logs | 14:29 |
| cdent | sdague: isn't that what FaultWrap is supposed to do? | 14:30 |
| cdent | and it _is_ being caught before then too because that's what this does: | 14:31 |
| *** thorst has quit IRC | 14:31 | |
| cdent | https://github.com/openstack/nova/blob/master/nova/api/openstack/placement/handler.py#L179 | 14:31 |
| sdague | I would have thought, I don't know why this isn't getting to the subunit in gabbi tests when it is everywhere else | 14:32 |
| cdent | (and we see that 'uncaught exception' message in the paste I made) | 14:32 |
| *** marst has quit IRC | 14:32 | |
| sdague | if you push a DNM change that has the kind of issue that you want exposed, I can poke a little | 14:32 |
| mriedem | BobBall: i'll take a look, in meetings | 14:32 |
| sdague | right, but that's not logged in the standard logger | 14:32 |
| sdague | otherwise it would look like oslo_log format | 14:33 |
| BobBall | No rush :) | 14:33 |
| cdent | sdague: yes, I'm agreeing with you: something's off | 14:33 |
| sdague | cdent: ok, push a DNM before you leave | 14:33 |
| sdague | then I have something concrete to look at | 14:33 |
| cdent | ya, thanks, | 14:33 |
| mriedem | BobBall: on that patch, we're going to want to see the citrix xenserver ci run with artom's tempest patch to test 2.32 | 14:35 |
| mriedem | with device tags | 14:35 |
| cdent | ossum, my vm just died | 14:36 |
| mriedem | BobBall: this one https://review.openstack.org/#/c/305120/ | 14:36 |
| *** rtheis has joined #openstack-nova | 14:36 | |
| BobBall | Yeah - I see that. Just thinking of the best way to get Zuul to run a test with both of them. Ideally don't want the Nova change to depend on Tempest or other way round either :) | 14:37 |
| mriedem | BobBall: the hyper-v wizards were able to get a single run with both patches somehow | 14:37 |
| mriedem | BobBall: we could...have a separate DNM patch that depends on both | 14:37 |
| mriedem | like roll one on top of the xenserver nova patch | 14:38 |
| mriedem | that depends on the tempest patch | 14:38 |
| mriedem | that should do it | 14:38 |
| BobBall | Ah - yes - that'd be a nice easy way to do that | 14:38 |
| BobBall | Will make it happen. | 14:38 |
| mriedem | thanks | 14:38 |
| openstackgerrit | Stephen Finucane proposed openstack/nova: hacking: Always use 'assertIs(Not)None' https://review.openstack.org/363623 | 14:41 |
| openstackgerrit | Francesco Santoro proposed openstack/os-vif: os-vif: add new VIFVHostUserFP class https://review.openstack.org/375533 | 14:44 |
| openstackgerrit | Chris Dent proposed openstack/nova: DNM Force gabbi and other placement tests to fail https://review.openstack.org/375534 | 14:45 |
| cdent | sdague: that ^ should exercise it, i _think_ | 14:46 |
| cdent | biab | 14:46 |
| *** dikonoor has quit IRC | 14:46 | |
| *** cdent has quit IRC | 14:47 | |
| *** stewie925 has joined #openstack-nova | 14:47 | |
| *** alexchadin has quit IRC | 14:47 | |
| *** sridharg has quit IRC | 14:48 | |
| *** brian_price has joined #openstack-nova | 14:48 | |
| *** rrecio has joined #openstack-nova | 14:48 | |
| *** tonytan_brb has joined #openstack-nova | 14:50 | |
| *** zz_dimtruck is now known as dimtruck | 14:50 | |
| *** rrecio_ has joined #openstack-nova | 14:50 | |
| *** mlavalle has joined #openstack-nova | 14:52 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Adds Hyper-V OVS ViF driver https://review.openstack.org/140045 | 14:52 |
| *** rrecio has quit IRC | 14:53 | |
| *** tonytan4ever has quit IRC | 14:53 | |
| openstackgerrit | Simona Iuliana Toader proposed openstack/nova-specs: Adds Hyper-V Shielded VMs spec https://review.openstack.org/274709 | 14:54 |
| *** amotoki has quit IRC | 14:58 | |
| *** tjones has joined #openstack-nova | 14:58 | |
| *** annegentle has quit IRC | 14:59 | |
| *** burgerk has joined #openstack-nova | 14:59 | |
| *** annegentle has joined #openstack-nova | 15:00 | |
| *** f10d4 has joined #openstack-nova | 15:01 | |
| *** jistr is now known as jistr|call | 15:03 | |
| *** Oku_OS is now known as Oku_OS-away | 15:04 | |
| *** lpetrut has quit IRC | 15:04 | |
| *** vishwanathj has joined #openstack-nova | 15:05 | |
| *** ducttape_ has joined #openstack-nova | 15:07 | |
| *** mat128 has quit IRC | 15:08 | |
| *** mat128 has joined #openstack-nova | 15:08 | |
| mdbooth | artom: Does my -1 on your tempest test make sense, btw? | 15:09 |
| * mdbooth is prepared to be told he doesn't understand tempest :) | 15:10 | |
| artom | mdbooth, it totally does | 15:11 |
| mdbooth | \o/ I understood something :) | 15:11 |
| *** ijw has joined #openstack-nova | 15:11 | |
| *** p01nt3r75 has joined #openstack-nova | 15:11 | |
| *** ducttape_ has quit IRC | 15:12 | |
| artom | mdbooth, hah :) Well, your -1 is more about me not understanding flavors imposing ephemeral disk sizes, so it's not strictly speaking Tempest | 15:12 |
| artom | So I'd take that \o/ back if I were you ;) | 15:12 |
| *** burgerk has quit IRC | 15:13 | |
| mdbooth | never | 15:13 |
| *** cdent has joined #openstack-nova | 15:13 | |
| *** jheroux has quit IRC | 15:14 | |
| *** amotoki has joined #openstack-nova | 15:14 | |
| *** tanvirt has joined #openstack-nova | 15:14 | |
| *** raunak has joined #openstack-nova | 15:15 | |
| *** ijw has quit IRC | 15:16 | |
| *** ijw has joined #openstack-nova | 15:16 | |
| *** amotoki has quit IRC | 15:17 | |
| *** tjones has left #openstack-nova | 15:18 | |
| *** raunak has quit IRC | 15:18 | |
| *** jheroux has joined #openstack-nova | 15:19 | |
| *** rcernin has quit IRC | 15:19 | |
| *** lucasagomes is now known as lucas-hungry | 15:20 | |
| *** tonytan_brb is now known as tonytan4ever | 15:21 | |
| *** dikonoor has joined #openstack-nova | 15:24 | |
| *** Dinesh_Bhor has quit IRC | 15:24 | |
| *** dave-mccowan has quit IRC | 15:25 | |
| *** mdrabe has quit IRC | 15:25 | |
| *** jistr|call is now known as jistr | 15:26 | |
| *** jistr is now known as jistr|biab | 15:28 | |
| *** bhagyashri has quit IRC | 15:30 | |
| *** mdrabe has joined #openstack-nova | 15:31 | |
| openstackgerrit | Simona Iuliana Toader proposed openstack/nova-specs: Adds Hyper-V Shielded VMs spec https://review.openstack.org/274709 | 15:33 |
| *** jistr|biab is now known as jistr | 15:36 | |
| *** dimtruck is now known as zz_dimtruck | 15:37 | |
| *** lamt has quit IRC | 15:39 | |
| *** thorst has joined #openstack-nova | 15:39 | |
| *** bks has quit IRC | 15:39 | |
| *** thorst has quit IRC | 15:41 | |
| *** thorst has joined #openstack-nova | 15:41 | |
| *** parora has joined #openstack-nova | 15:42 | |
| *** vladikr has quit IRC | 15:43 | |
| openstackgerrit | John Garbutt proposed openstack/nova-specs: WIP: Use neutron's new port binding API https://review.openstack.org/375580 | 15:44 |
| *** sandanar has quit IRC | 15:44 | |
| *** thorst has quit IRC | 15:45 | |
| *** Guest37187 has quit IRC | 15:45 | |
| *** thorst has joined #openstack-nova | 15:46 | |
| *** rmart04 has quit IRC | 15:46 | |
| cdent | sdague: I'm home now, with a plug, but not making any headway on what the issue is. I can say, however, that with the change associated with this log: http://logs.openstack.org/55/374955/3/check/gate-nova-tox-db-functional-ubuntu-xenial/9cb901e/console.html the number of "leaks" is less and those that are remaing are bugs to be fixed, such that once those are fixed, the only leaks will be unexpected stuff | 15:46 |
| cdent | not ideal, I know, but in the short term, not the worst | 15:46 |
| *** edmondsw has quit IRC | 15:46 | |
| *** panbalag1 has joined #openstack-nova | 15:47 | |
| *** panbalag1 has left #openstack-nova | 15:47 | |
| mdbooth | johnthetubaguy: With that neutron api spec, could we reasonably be aggressive about ditching the old one asap? | 15:48 |
| johnthetubaguy | mdbooth: the old one? | 15:48 |
| johnthetubaguy | api or spec? | 15:49 |
| mdbooth | johnthetubaguy: So there's a proposed new neutron port plugging api, right? | 15:49 |
| *** yamahata has quit IRC | 15:49 | |
| mdbooth | And an old one. | 15:49 |
| *** parora has quit IRC | 15:49 | |
| mdbooth | If we adopted the new one, how quickly could we drop the old one? | 15:49 |
| johnthetubaguy | ah, API, yeah, we shouldn't have an overlap of more than two cycles | 15:49 |
| johnthetubaguy | at least, thats my take | 15:49 |
| *** edmondsw has joined #openstack-nova | 15:49 | |
| mdbooth | It's just that we've been historically bad at maintaining 2 code paths like that. | 15:49 |
| mdbooth | Cool. | 15:50 |
| *** panbalag has joined #openstack-nova | 15:50 | |
| *** rama_y has quit IRC | 15:52 | |
| *** thorst has quit IRC | 15:53 | |
| sdague | cdent: I think the crux of the issue is that gabbi is building base tests out of unit tests raw, as such the run is not reporting into the subunit stream in the same way that testtools + useFixture sets up | 15:54 |
| *** panbalag has quit IRC | 15:54 | |
| *** panbalag has joined #openstack-nova | 15:54 | |
| cdent | there's special reporting into the into the subunit stream? where's that located? | 15:54 |
| sdague | https://github.com/testing-cabal/testtools/blob/bfc88a45a3d94b4b030f9210471b2f700d294f15/testtools/testcase.py#L721-L761 - that gather_details phase is how we capture all those things | 15:54 |
| *** thorst has joined #openstack-nova | 15:55 | |
| cdent | ah | 15:55 |
| sdague | this is the reason we use testtools.TestCase as our base through all of openstack | 15:55 |
| * cdent blinks | 15:56 | |
| clarkb | the subunit runner should actually be able to patch unittest too fwiw | 15:56 |
| clarkb | the real reason we pushed for testtools was consistent behavior in python2.6 2.7 and 3.x | 15:56 |
| *** thorst has quit IRC | 15:56 | |
| clarkb | neither unittest nor unittest2 provided that at the time. But now python2.6 isn't supported and unittest2 should work aiui | 15:56 |
| *** thorst has joined #openstack-nova | 15:57 | |
| cdent | we can probably make the gabbi fixture do _something_ but it is going to be relevant that gabbi fixtures surround and entire TestSuite not a TestCase | 15:57 |
| cdent | because each file is an ordered graph | 15:57 |
| cdent | if you choose one leaf, everything before it is run | 15:57 |
| *** bnemec is now known as beekneemech | 15:58 | |
| *** matrohon has quit IRC | 15:58 | |
| *** code-R_ has quit IRC | 15:59 | |
| *** panbalag has quit IRC | 15:59 | |
| sdague | cdent: so, I expect we need to figure out out to pass fixtures down to the tests to setup themselves | 16:00 |
| sdague | anyway, https://github.com/sdague/gabbi has a totally hack and smash patch which I think demonstrates the end behavior we want | 16:00 |
| *** mdrabe has quit IRC | 16:01 | |
| *** mdrabe has joined #openstack-nova | 16:01 | |
| sdague | clarkb: when running on unittest we don't seem to get the subunit attachments for things like python logging | 16:01 |
| cdent | An alternative might be to duplicate the stuff that the fixtures are doing, as gabbi fixturs, and just ignore the nova fixtures? | 16:01 |
| *** panbalag has joined #openstack-nova | 16:01 | |
| cdent | brb | 16:01 |
| sdague | cdent: possibly, but the thing that's actually important is that details reporting at the end | 16:02 |
| sdague | which is just part of the useFixtures interface on testtools.TestCase | 16:02 |
| clarkb | sdague: thats likely because you aren't inheriting the base class that sets that all up | 16:02 |
| clarkb | sdague: but that base class could inherit unittest and work (I think) as long as you don't use python2.6 | 16:02 |
| johnthetubaguy | mdbooth: a very, very valid call out | 16:02 |
| *** gongysh has joined #openstack-nova | 16:03 | |
| sdague | clarkb: no, just changing base class doesn't impact anything | 16:03 |
| *** thorst has quit IRC | 16:03 | |
| clarkb | huh | 16:03 |
| *** raunak has joined #openstack-nova | 16:03 | |
| *** raunak has quit IRC | 16:03 | |
| *** thorst has joined #openstack-nova | 16:04 | |
| *** thorst has quit IRC | 16:04 | |
| cdent | sdague: the reason I'm suggesting the alternative is because I'm not sure I want to change gabbi's model just for nova, but I'll have a poke and see which way goes better. I suppose your way will provide the most flexibility in the long run | 16:05 |
| *** raunak has joined #openstack-nova | 16:05 | |
| cdent | s/just for/just because it doesn't work for/ | 16:05 |
| *** raunak has quit IRC | 16:06 | |
| mriedem | artom: you might have an opinion on throwing device tags into the os-diagnostics api https://review.openstack.org/#/c/357884/ | 16:06 |
| mriedem | superdan: ^ | 16:06 |
| mriedem | we don't expose the device tags from any REST API today so this would be a place to do it | 16:06 |
| danpb | hmm, interesting idea | 16:06 |
| *** Apoorva has joined #openstack-nova | 16:07 | |
| *** thorst has joined #openstack-nova | 16:07 | |
| superdan | yeah | 16:07 |
| danpb | that would remove the only compelling reason to expose device names in the diagnostics | 16:07 |
| *** raunak has joined #openstack-nova | 16:07 | |
| danpb | which is nice to avoid if we can help it | 16:07 |
| openstackgerrit | Andrew Laski proposed openstack/nova: DNM https://review.openstack.org/375106 | 16:07 |
| danpb | so i'm +1 on exposing device tags, and -1 on device names | 16:07 |
| mriedem | danpb: well you don't always have tags | 16:07 |
| mriedem | you always have a device name | 16:07 |
| *** raunak has quit IRC | 16:08 | |
| superdan | mriedem: but device names are almost always wrong, so .. :) | 16:08 |
| danpb | mriedem: but device name might be auto-generated | 16:08 |
| sdague | cdent: sure, anyway, hopefully that above hacky patch gives at least more bread crumbs to understanding why gabbi's output isn't getting where we expect it | 16:08 |
| danpb | mriedem: so you don't neccessarily know which device name matches your original config | 16:08 |
| danpb | whereas you would always know which device tag matchs | 16:08 |
| mriedem | we don't know the device name after we've created the guest? | 16:08 |
| superdan | no | 16:08 |
| mriedem | ffs | 16:08 |
| danpb | mriedem: exposing device names in diagnostics might encourage people to provide device names in the BDM at instance boot, which we don't want :-) | 16:09 |
| superdan | tbh, we probably should also auto-generate tags when they're not proided | 16:09 |
| mriedem | danpb: we can disable that on instance boot with a microversion | 16:09 |
| superdan | like 'root', 'ephemeral', and 'volume-%uuid' | 16:09 |
| mriedem | but that's a side thing we probably want anyway | 16:09 |
| danpb | superdan: yeah that'd be quite nice | 16:09 |
| superdan | yeah | 16:09 |
| superdan | not sure why we didn't do that in the first place | 16:09 |
| mriedem | i blame artom | 16:09 |
| *** dave-mccowan has joined #openstack-nova | 16:09 | |
| mriedem | for not predicting our future needs | 16:10 |
| sdague | we need a time machine | 16:10 |
| superdan | mriedem: careful, he scares easily | 16:10 |
| mdbooth | superdan: That was my thought too, btw. | 16:10 |
| mdbooth | i.e. auto-generated tags | 16:10 |
| danpb | as long as we don't allow anything to be dependant on the device names we generate, that's fine | 16:10 |
| danpb | ie, we should be clear that apps using nova should not rely on particular device tag strings being used | 16:11 |
| *** vladikr has joined #openstack-nova | 16:11 | |
| *** jlanoux has quit IRC | 16:11 | |
| *** lamt has joined #openstack-nova | 16:12 | |
| openstackgerrit | sahid proposed openstack/nova-specs: Libvirt driver emulator threads placement policy https://review.openstack.org/284094 | 16:13 |
| *** thorst has quit IRC | 16:13 | |
| *** bkopilov has quit IRC | 16:14 | |
| cfriesen | sahid: sfinucan: can you sanity-check my last comment in https://review.openstack.org/#/c/286744/ ? | 16:14 |
| *** thorst has joined #openstack-nova | 16:14 | |
| leakypipes | diana_clarke: we all choose our own nicks in IRC :) Casual Friday or otherwise! | 16:14 |
| *** Swami has joined #openstack-nova | 16:15 | |
| sahid | cfriesen: looking | 16:16 |
| sfinucan | leakypipes: Ahhh, I was wondering... | 16:16 |
| *** ariesyous has joined #openstack-nova | 16:16 | |
| sfinucan | cfriesen: ditto | 16:16 |
| ariesyous | Hello and Happy Friday everybody | 16:17 |
| ariesyous | Was wondering if Nova supports importing an already running kvm or qemu machine into Nova | 16:17 |
| ariesyous | so I can run nova commands against it | 16:17 |
| cdent | sdague: I think your way can probably work pretty well. How do you like the name 'inner_fixtures' which is a list of classes? | 16:17 |
| superdan | ariesyous: no, and also please see topic | 16:17 |
| ariesyous | Ah apologies | 16:18 |
| sdague | cdent: that seems fine to me | 16:18 |
| *** armax has joined #openstack-nova | 16:18 | |
| sdague | cdent: it also is going to require some interface change to the loader to define the base test class, right? | 16:18 |
| *** jistr is now known as jistr|afk | 16:18 | |
| cdent | oh, no, that can just be testtools by default, it was that early on and I took it out because it was overhead | 16:19 |
| sdague | as I'm honestly not super clear how much of testtools you'd need to duplicate to get the equiv of useFixtures | 16:19 |
| sdague | oh, ok, that's simpler then | 16:19 |
| sdague | I thought there were reasons to not use it in all cases | 16:19 |
| diana_clarke | leakypipes: I know ;) | 16:19 |
| sahid | cfriesen: yes your comment make sense we should use the host topology of the destination host | 16:19 |
| sahid | as for the vcpu_pin_set | 16:20 |
| cdent | sdague: I only thought of it as a weight problem | 16:20 |
| sdague | yeh, then if you just have a way to give it a list of fixture classes to apply to tests I think we might have it all sorted | 16:20 |
| *** raunak has joined #openstack-nova | 16:20 | |
| *** salv-orlando has joined #openstack-nova | 16:20 | |
| *** raunak has quit IRC | 16:20 | |
| *** jaosorior has quit IRC | 16:21 | |
| leakypipes | figleaf: lol on your tweet :) | 16:21 |
| *** adisky has quit IRC | 16:22 | |
| *** sahid has quit IRC | 16:22 | |
| cfriesen | danpb: in LibvirtDriver._live_migration_monitor() where we call guest.abort_job(), what do we expect info.type to be on the next time through the loop? | 16:22 |
| *** nic has joined #openstack-nova | 16:24 | |
| *** nherciu_ has quit IRC | 16:24 | |
| mriedem | sdague: fyi, seeing more osic cloud ceph snapshot timeout failures http://logs.openstack.org/79/175479/26/check/gate-tempest-dsvm-full-devstack-plugin-ceph-ubuntu-xenial/0853944/console.html#_2016-09-23_15_23_57_092311 | 16:24 |
| danpb | cfriesen: it can be quite a few different things | 16:25 |
| *** hshiina has quit IRC | 16:25 | |
| *** code-R has joined #openstack-nova | 16:26 | |
| danpb | cfriesen: normally it would change to CANCELLED | 16:26 |
| sdague | mriedem: only osic? | 16:26 |
| danpb | cfriesen: but it could still be UNBOUNDED as the abort takes place asynchronously | 16:27 |
| mriedem | yeah | 16:27 |
| danpb | cfriesen: similarly it could be COMPLETED if migration finished before the abort got processed | 16:27 |
| mriedem | but only 5 hits in 7 days according to logstash | 16:27 |
| danpb | cfriesen: or FAILED if migration failed before abort got processed | 16:27 |
| danpb | IOW, 'abort_job()' is merely a friendly request, not a guarantee that it'll abort | 16:27 |
| cfriesen | danpb: we hit a case where it was stuck as UNBOUNDED and we were stuck in the loop forever | 16:27 |
| sdague | mriedem: well osic is 50% of our capacity I think, so fails will happen there more often | 16:28 |
| danpb | that makes sense from nova's pov, as the migration is still ongoing | 16:28 |
| danpb | it could be a sign of a QEMU bug though | 16:28 |
| clarkb | sdague: mriedem correct its the majority so statistically will see more fails | 16:29 |
| *** yamahata has joined #openstack-nova | 16:29 | |
| *** lucas-hungry is now known as lucasagomes | 16:29 | |
| cdent | sdague: this is working in the desired way http://paste.openstack.org/show/582788/ | 16:29 |
| *** rlrossit has quit IRC | 16:29 | |
| cdent | the limitation is that there is no way to pass arguments to the fixtures | 16:30 |
| sdague | cdent: awesome | 16:30 |
| cdent | so you'd have to partials or something like | 16:30 |
| sdague | well... we can probably live with it for now | 16:30 |
| cdent | but I think it is reasonable enough as is | 16:30 |
| sdague | yeh | 16:30 |
| sdague | it will definitely work for what we currently need, and most of the fixtures that I know we have (should we need others) | 16:31 |
| cdent | it will take me a few days to package it into gabbi, but I think it's a good solution and won't destroy the world | 16:31 |
| *** thorst has quit IRC | 16:31 | |
| sdague | it's pretty rare fixtures that need constructor elements | 16:31 |
| *** raunak has joined #openstack-nova | 16:31 | |
| sdague | cdent: cool | 16:31 |
| cdent | you still have to launch the same fixtures in the gabbi fixture as well, to catch things like the database logging | 16:31 |
| cdent | but that's fine | 16:31 |
| cfriesen | danpb: thanks, that's helpful | 16:31 |
| *** thorst has joined #openstack-nova | 16:31 | |
| *** pilgrimstack has quit IRC | 16:31 | |
| cdent | sdague: thanks for helping me push it through to something that keeps it useful | 16:32 |
| sdague | cdent: yeh, no prob | 16:32 |
| *** rama_y has joined #openstack-nova | 16:32 | |
| sdague | also, why do you think we need it at the gabbi fixture as well? | 16:32 |
| sdague | I would have thought this would capture everything there | 16:32 |
| *** gongysh has quit IRC | 16:32 | |
| cdent | because the database is setup before any tests are run | 16:32 |
| sdague | ah, ok | 16:32 |
| *** thorst has quit IRC | 16:33 | |
| cdent | since once database is used per suite | 16:33 |
| sdague | though, you could just pass the db fixture to the inner fixture | 16:33 |
| sdague | but sure | 16:33 |
| cdent | that woud break all the tests | 16:33 |
| sdague | oh, because these chain | 16:33 |
| * cdent nods | 16:33 | |
| sdague | right, I forgot they aren't isolated | 16:33 |
| sdague | yeh, so that's fine | 16:33 |
| cdent | they are isolate, just differently ;) | 16:34 |
| sdague | :) | 16:34 |
| cdent | I'm going to go somewhere else for a while, I have a disappointed wife | 16:34 |
| cdent | "are you _still_ working" | 16:34 |
| cdent | bb<not sure when> | 16:35 |
| *** cdent has quit IRC | 16:35 | |
| *** thorst has joined #openstack-nova | 16:36 | |
| *** p01nt3r75 has quit IRC | 16:38 | |
| *** vladikr has quit IRC | 16:40 | |
| *** wjhofer has joined #openstack-nova | 16:41 | |
| *** dave-johnston has quit IRC | 16:41 | |
| *** code-R has quit IRC | 16:42 | |
| wjhofer | Hello everyone. I am running Mitaka 8.1 and have it configured to use neutron networking, but I am still seeing nova create iptable rules on my hypervisor nodes. It seems that each nova rule has a neutron counter part and depending on the order the networking will break. Thoughts? | 16:43 |
| *** code-R has joined #openstack-nova | 16:43 | |
| superdan | wjhofer: please see topic | 16:43 |
| superdan | bauwser: on https://review.openstack.org/#/c/374307/, did you see that I was looking for your feedback on your request spec migration specifically? | 16:44 |
| superdan | melwitt: you too for instance groups ^ | 16:44 |
| *** unicell has quit IRC | 16:45 | |
| *** wjhofer has quit IRC | 16:45 | |
| *** thorst has quit IRC | 16:47 | |
| *** code-R_ has joined #openstack-nova | 16:47 | |
| *** thorst has joined #openstack-nova | 16:48 | |
| *** code-R_ has quit IRC | 16:48 | |
| *** code-R has quit IRC | 16:48 | |
| *** code-R has joined #openstack-nova | 16:48 | |
| *** zenoway has quit IRC | 16:48 | |
| *** zenoway has joined #openstack-nova | 16:50 | |
| *** zenoway has quit IRC | 16:51 | |
| *** thorst has quit IRC | 16:54 | |
| *** thorst has joined #openstack-nova | 16:55 | |
| *** derekh has quit IRC | 16:56 | |
| *** jistr|afk is now known as jistr | 16:59 | |
| melwitt | superdan: I can't think of a reason why forcing instance_groups migration would be a problem, assuming we don't have any policy around it. could you tell an example of a reason one release later wouldn't be okay? | 16:59 |
| superdan | melwitt: it just means we have to do it later, and we have to keep all the load-from-here-or-from-there code | 17:00 |
| superdan | if we put it into the blocker migration, then we can remove all that stuff and assume they're all moved | 17:00 |
| melwitt | superdan: yeah, I understand that part but I meant what would be an example of a gotcha or have we not had one yet | 17:01 |
| superdan | melwitt: a gotcha for removing it earlier than later you mean? | 17:01 |
| melwitt | yes | 17:02 |
| *** markvoelker has quit IRC | 17:02 | |
| openstackgerrit | Paul Carlton proposed openstack/nova: Clean up instance on target node if evacuate fails https://review.openstack.org/375623 | 17:02 |
| *** markvoelker has joined #openstack-nova | 17:02 | |
| melwitt | superdan: I'm concerned I'm not aware or thinking of the possible problems of removing it earlier and thought an example might help | 17:02 |
| superdan | just that we require people to have done that homework before they can move on.. on the pci one we left it in there for a couple cycles because we couldn't as easily block on completion, which provided more grace period | 17:02 |
| superdan | but in this case, they had to have completed most/all of the newton ones anyway, which would include instance groups | 17:03 |
| melwitt | oh, okay | 17:03 |
| superdan | so probably not much reason to wait | 17:03 |
| superdan | but just wanted to ask | 17:03 |
| melwitt | yeah. cool, sounds fine to me then | 17:03 |
| superdan | okay | 17:05 |
| openstackgerrit | Paul Carlton proposed openstack/nova: HNOV-815 Clean up instance on target node if evacuate fails https://review.openstack.org/375623 | 17:06 |
| openstackgerrit | Paul Carlton proposed openstack/nova: Clean up instance on target node if evacuate fails https://review.openstack.org/375623 | 17:07 |
| *** ducttape_ has joined #openstack-nova | 17:08 | |
| *** catintheroof has joined #openstack-nova | 17:09 | |
| *** thorst has quit IRC | 17:10 | |
| *** thorst has joined #openstack-nova | 17:10 | |
| *** kaisers_ has joined #openstack-nova | 17:11 | |
| *** kaisers_ has quit IRC | 17:11 | |
| *** kaisers_ has joined #openstack-nova | 17:11 | |
| *** ducttape_ has quit IRC | 17:12 | |
| *** ihrachys has quit IRC | 17:13 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Add blocker migration to ensure for newton online migrations https://review.openstack.org/374307 | 17:14 |
| superdan | melwitt: ^ | 17:14 |
| *** thorst has quit IRC | 17:15 | |
| *** thorst has joined #openstack-nova | 17:16 | |
| *** thorst has quit IRC | 17:16 | |
| *** vladikr has joined #openstack-nova | 17:20 | |
| *** unicell has joined #openstack-nova | 17:20 | |
| *** paul-carlton2 has quit IRC | 17:21 | |
| *** thorst has joined #openstack-nova | 17:21 | |
| *** daemontool__ has quit IRC | 17:21 | |
| *** Apoorva has quit IRC | 17:21 | |
| *** rhagarty_ has quit IRC | 17:22 | |
| *** rhagarty_ has joined #openstack-nova | 17:22 | |
| *** thorst has quit IRC | 17:23 | |
| *** thorst has joined #openstack-nova | 17:23 | |
| *** ihrachys has joined #openstack-nova | 17:25 | |
| *** code-R_ has joined #openstack-nova | 17:28 | |
| *** code-R has quit IRC | 17:31 | |
| *** thorst has quit IRC | 17:34 | |
| *** thorst has joined #openstack-nova | 17:34 | |
| *** akshai has quit IRC | 17:35 | |
| *** code-R has joined #openstack-nova | 17:35 | |
| *** gouthamr_ has joined #openstack-nova | 17:36 | |
| *** ihrachys has quit IRC | 17:36 | |
| *** gouthamr has quit IRC | 17:36 | |
| *** akshai has joined #openstack-nova | 17:37 | |
| *** thorst has quit IRC | 17:38 | |
| *** code-R_ has quit IRC | 17:38 | |
| *** ariesyous has left #openstack-nova | 17:38 | |
| openstackgerrit | Vu Tran proposed openstack/nova: libvirt: Cannot live-migrate VM using default cpu model https://review.openstack.org/375644 | 17:39 |
| *** thorst has joined #openstack-nova | 17:39 | |
| *** gouthamr has joined #openstack-nova | 17:40 | |
| *** beagles has joined #openstack-nova | 17:41 | |
| *** gouthamr_ has quit IRC | 17:41 | |
| beagles | melwitt: thanks for the review etc. on this https://review.openstack.org/#/c/358024/ I somehow missed the feedback re: "there is no launchpad bug" | 17:42 |
| *** rmart04 has joined #openstack-nova | 17:43 | |
| beagles | melwitt: of course now I'm asked if it can be backported... which case I know that isn't going to happen without a bug. Question is, even with an appropriate bug, does this meet the criteria for a backport to mitaka | 17:44 |
| *** rmart04 has quit IRC | 17:44 | |
| *** thorst has quit IRC | 17:45 | |
| *** thorst has joined #openstack-nova | 17:46 | |
| *** rlrossit has joined #openstack-nova | 17:46 | |
| *** melwitt has left #openstack-nova | 17:46 | |
| *** melwitt has joined #openstack-nova | 17:47 | |
| *** gouthamr has quit IRC | 17:47 | |
| *** sudswas__ has quit IRC | 17:47 | |
| *** Guest56480 has quit IRC | 17:47 | |
| *** gouthamr_ has joined #openstack-nova | 17:48 | |
| *** gouthamr_ has quit IRC | 17:49 | |
| *** gouthamr has joined #openstack-nova | 17:50 | |
| *** mbirru has quit IRC | 17:50 | |
| *** danpb has quit IRC | 17:51 | |
| melwitt | beagles: whether or not it can be backported to mitaka is a better question for mriedem. if it's appropriate, I think we'd just want to open a bug now and refer to the patches there | 17:52 |
| beagles | melwitt: thanks! | 17:52 |
| *** thorst has quit IRC | 17:53 | |
| *** thorst has joined #openstack-nova | 17:54 | |
| *** hemanthm is now known as hemanthm|lunch | 17:55 | |
| *** psachin has joined #openstack-nova | 17:56 | |
| * beagles notes that it is dead quiet in here.... I haven't been around for awhile but wow... | 17:58 | |
| openstackgerrit | Andrew Laski proposed openstack/nova: DNM https://review.openstack.org/375106 | 17:58 |
| melwitt | beagles: here's the doc on stable branch policy. ultimately it's up to the stable team to decide the appropriateness of each backport http://docs.openstack.org/project-team-guide/stable-branches.html#appropriate-fixes | 17:59 |
| beagles | mriedem: what do you think? backportable? https://bugs.launchpad.net/nova/+bug/1627131 | 17:59 |
| openstack | Launchpad bug 1627131 in OpenStack Compute (nova) "Bigswitch related command filters are not included compute filters file" [Undecided,New] | 17:59 |
| mriedem | backporting rootwrap filters is generally not cool | 17:59 |
| mriedem | beagles: not imo | 17:59 |
| beagles | mriedem: agreed.. in this case it is just "moving" it from one file to another, but I do get what you mean | 18:00 |
| *** thorst has quit IRC | 18:01 | |
| beagles | depending on your interpretation of what goes where in these files, it may have been misplaced to begin with | 18:01 |
| *** krtaylor has quit IRC | 18:01 | |
| *** thorst has joined #openstack-nova | 18:01 | |
| *** ariesyous has joined #openstack-nova | 18:02 | |
| openstackgerrit | Evgeny Antyshev proposed openstack/nova: Don't expect block job to exist after rebase https://review.openstack.org/375652 | 18:03 |
| *** mariusv has quit IRC | 18:04 | |
| *** thorst has quit IRC | 18:04 | |
| *** mariusv has joined #openstack-nova | 18:04 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Adds Hyper-V OVS ViF driver https://review.openstack.org/140045 | 18:06 |
| *** thorst has joined #openstack-nova | 18:07 | |
| *** salv-orlando has quit IRC | 18:10 | |
| *** psachin has quit IRC | 18:11 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes python 3 unit tests https://review.openstack.org/278810 | 18:14 |
| *** armax_ has joined #openstack-nova | 18:14 | |
| *** armax has quit IRC | 18:15 | |
| *** armax_ is now known as armax | 18:15 | |
| *** thorst has quit IRC | 18:16 | |
| *** armax has quit IRC | 18:16 | |
| *** ducttape_ has joined #openstack-nova | 18:16 | |
| *** thorst has joined #openstack-nova | 18:16 | |
| *** thorst has quit IRC | 18:22 | |
| *** prateek has joined #openstack-nova | 18:23 | |
| *** thorst has joined #openstack-nova | 18:28 | |
| *** zz_dimtruck is now known as dimtruck | 18:28 | |
| *** thorst has quit IRC | 18:31 | |
| *** ducttape_ has quit IRC | 18:31 | |
| *** thorst has joined #openstack-nova | 18:32 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Add Hyper-V storage QoS support https://review.openstack.org/175479 | 18:34 |
| *** andymaier has quit IRC | 18:38 | |
| *** artom has quit IRC | 18:39 | |
| *** thorst has quit IRC | 18:40 | |
| *** rcarrillocruz has quit IRC | 18:40 | |
| mriedem | mdbooth: nice catch on https://review.openstack.org/#/c/305120/ | 18:42 |
| mriedem | mdbooth: btw device tags are only tested in the gate-tempest-dsvm-neutron-full-ssh job | 18:42 |
| *** rcarrillocruz has joined #openstack-nova | 18:45 | |
| *** hemanthm|lunch is now known as hemanthm | 18:50 | |
| *** sdague has quit IRC | 18:51 | |
| *** prateek has quit IRC | 18:55 | |
| *** devananda|lunch is now known as devananda | 18:56 | |
| *** rlrossit has quit IRC | 18:59 | |
| leakypipes | mriedem: I'm currently looking at https://review.openstack.org/#/c/306010/ from vdrok and the thing that strikes me as exceedingly odd is the "recreate" parameter to rebuild_instance(). In what scenarios is rebuild_instance() *NOT* going to do a recreate of the instance? | 18:59 |
| *** rlrossit has joined #openstack-nova | 18:59 | |
| *** kfarr has quit IRC | 18:59 | |
| *** rlrossit_ has joined #openstack-nova | 19:01 | |
| mriedem | ironic? | 19:02 |
| mriedem | ironic is the only one that has a non-default impl for rebuild | 19:02 |
| mriedem | also, i think that recreate parameter is just the flag to tell us it's an evacuate operation | 19:03 |
| *** rlrossit has quit IRC | 19:04 | |
| *** ihrachys has joined #openstack-nova | 19:05 | |
| mriedem | god that's always a mess to see what's passing recreate=True vs False | 19:05 |
| mriedem | compute.api explicitly passes recreate=True for evacuate | 19:06 |
| mriedem | but doesn't pass recreate=False for rebuild, it leaves that to conductor to default recreate=False | 19:06 |
| mriedem | anyway | 19:06 |
| *** armax has joined #openstack-nova | 19:07 | |
| mriedem | so i guess the difference is rebuild you stay on the same host/node don't you? | 19:08 |
| mriedem | evacuate you move | 19:08 |
| *** sdague has joined #openstack-nova | 19:08 | |
| superdan | mriedem: yeah, evac I think | 19:10 |
| superdan | actually | 19:10 |
| superdan | it might always be recreate=False in the case of a volume | 19:10 |
| *** david-lyle_ has joined #openstack-nova | 19:10 | |
| mriedem | ah yeah, compute api rebuild does the rpc cast to the instance.host | 19:12 |
| mriedem | so the rebuild happens on that same host | 19:12 |
| *** READ10 has quit IRC | 19:13 | |
| *** david-lyle has quit IRC | 19:13 | |
| *** raunak has quit IRC | 19:13 | |
| leakypipes | mriedem: sorry, got emergency call :( back shortly, hopefully | 19:16 |
| *** tonytan4ever has quit IRC | 19:17 | |
| *** catintheroof has quit IRC | 19:17 | |
| *** ihrachys has quit IRC | 19:19 | |
| mriedem | vdrok: i commented on ^ | 19:20 |
| mriedem | why not just pass scheduled_node=instance.node from the rebuild_instance method in the compute API as jroll pointed out in the ironic patch? | 19:20 |
| mriedem | i think either will work, but you won't get this fix until all computes are upgraded | 19:21 |
| mriedem | which in the case of ironic is maybe not an issue? | 19:21 |
| openstackgerrit | Andrew Laski proposed openstack/nova: DNM https://review.openstack.org/375106 | 19:21 |
| *** lpetrut has joined #openstack-nova | 19:24 | |
| *** vladikr has quit IRC | 19:25 | |
| vdrok | leakypipes: mriedem thanks for reviewing. I can move this to compute API, was not sure that it is OK for all virt drivers | 19:26 |
| mriedem | vdrok: well, either way it's going to be a change for all virt drivers | 19:27 |
| *** annegentle has quit IRC | 19:27 | |
| *** burgerk has joined #openstack-nova | 19:27 | |
| mriedem | when was scheduled_node added to rebuild instance in the compute? | 19:27 |
| leakypipes | vdrok, mriedem: k, off phone, sorry bout that | 19:28 |
| mriedem | 4.5 https://github.com/openstack/nova/blob/master/nova/compute/rpcapi.py#L291 | 19:28 |
| mriedem | added scheduled_node to the rebuild_instance method in the compute | 19:28 |
| mriedem | which is liberty | 19:28 |
| *** armax has quit IRC | 19:29 | |
| *** tonytan4ever has joined #openstack-nova | 19:29 | |
| mriedem | "A node of the host chosen by the scheduler. If a host was specified by the user, this will be None" | 19:30 |
| mriedem | so, i'm not sure how much logic is baked into assumptions about how when and what scheduled_node should be set to | 19:31 |
| mriedem | i think that's only supposed to be set for evacuate? | 19:31 |
| mriedem | to force the instance to be scheduled to a certain host during the evacuate | 19:31 |
| mriedem | so passing it in for rebuild probably breaks any assumptions about that value | 19:31 |
| mriedem | superdan: ^ does any of that sound familiar? | 19:31 |
| mriedem | looking at https://review.openstack.org/#/c/200485/ | 19:32 |
| vdrok | mriedem: ooo, I think I tried what you're suggesting | 19:32 |
| *** ducttape_ has joined #openstack-nova | 19:32 | |
| *** bvanhav has quit IRC | 19:32 | |
| *** lpetrut has quit IRC | 19:32 | |
| vdrok | Iirc this rebuild claim caused troubles when passed | 19:33 |
| vdrok | It was starting a migration even tho it was same node | 19:33 |
| vdrok | And failing | 19:33 |
| superdan | mriedem: not really | 19:33 |
| mriedem | yeah if you pass scheduled_node in it does a rebuild claim | 19:33 |
| mriedem | b/c of evacuate i think | 19:33 |
| mriedem | claiming resources on this new host | 19:34 |
| mriedem | but in the case of rebuild, we've already claimed those resources | 19:34 |
| vdrok | I don't really remember the exact reason of the failure, and can't test now as already out of office :( | 19:34 |
| mriedem | vdrok: so i think you're right | 19:34 |
| mriedem | sweet baby jesus i wish we had separate methods in the compute manager for rebuild and evacuate | 19:34 |
| mriedem | this is too complicated | 19:34 |
| vdrok | ++ :) | 19:34 |
| *** dimtruck is now known as zz_dimtruck | 19:35 | |
| *** zz_dimtruck is now known as dimtruck | 19:35 | |
| *** ducttape_ has quit IRC | 19:37 | |
| mriedem | +2 | 19:40 |
| *** nherciu_ has joined #openstack-nova | 19:41 | |
| *** salv-orlando has joined #openstack-nova | 19:41 | |
| mriedem | leakypipes: all yours if it makes sense | 19:42 |
| vdrok | Thank you! | 19:42 |
| leakypipes | mriedem: thx, appreciated. | 19:43 |
| openstackgerrit | Jay Pipes proposed openstack/nova: placement: handle generic PCI devices in inventory https://review.openstack.org/374681 | 19:44 |
| openstackgerrit | Jay Pipes proposed openstack/nova: placement: allocations for generic PCI devices https://review.openstack.org/375168 | 19:44 |
| *** devananda is now known as devananda|bbl | 19:44 | |
| *** dimtruck is now known as zz_dimtruck | 19:44 | |
| *** salv-orlando has quit IRC | 19:46 | |
| *** andymaier has joined #openstack-nova | 19:46 | |
| sdague | Vek: which was your base -1 on https://review.openstack.org/#/c/375397? | 19:50 |
| mriedem | claudiub: want to fix the release note here? https://review.openstack.org/#/c/209581/ then i'm +2 | 19:50 |
| *** lpetrut has joined #openstack-nova | 19:53 | |
| claudiub | sure, just a sec | 19:53 |
| *** burgerk has quit IRC | 19:53 | |
| *** cmartin has quit IRC | 19:54 | |
| *** code-R has quit IRC | 19:58 | |
| *** sdake has quit IRC | 20:00 | |
| *** krtaylor has joined #openstack-nova | 20:00 | |
| openstackgerrit | Alex Szarka proposed openstack/nova-specs: Resume Guests State upgrade https://review.openstack.org/375388 | 20:02 |
| *** ijw has quit IRC | 20:02 | |
| *** TonyXu has joined #openstack-nova | 20:03 | |
| *** david-lyle_ is now known as david-lyle | 20:06 | |
| *** haplo37 has quit IRC | 20:06 | |
| *** TonyXu has quit IRC | 20:06 | |
| *** TonyXu has joined #openstack-nova | 20:07 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Hyper-V: Adds Hyper-V UEFI Secure Boot https://review.openstack.org/209581 | 20:07 |
| *** TonyXu has quit IRC | 20:07 | |
| *** TonyXu has joined #openstack-nova | 20:07 | |
| *** leakypipes has quit IRC | 20:08 | |
| *** annegentle has joined #openstack-nova | 20:08 | |
| *** haplo37 has joined #openstack-nova | 20:08 | |
| *** lucasagomes is now known as lucas-afk | 20:09 | |
| *** ijw has joined #openstack-nova | 20:10 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Fixes python 3 unit tests https://review.openstack.org/278810 | 20:10 |
| *** kaisers_ has quit IRC | 20:12 | |
| *** tonytan4ever has quit IRC | 20:12 | |
| *** dikonoor has quit IRC | 20:13 | |
| openstackgerrit | Sujitha proposed openstack/nova: Improve help text for upgrade_levels options https://review.openstack.org/329154 | 20:13 |
| *** ijw has quit IRC | 20:14 | |
| *** zz_dimtruck is now known as dimtruck | 20:15 | |
| *** akshai has quit IRC | 20:15 | |
| *** ducttape_ has joined #openstack-nova | 20:15 | |
| *** code-R has joined #openstack-nova | 20:16 | |
| *** akshai has joined #openstack-nova | 20:20 | |
| *** armax has joined #openstack-nova | 20:21 | |
| openstackgerrit | Matt Fischer proposed openstack/nova: docs: Clarify sections & note on filter scheduler https://review.openstack.org/375702 | 20:22 |
| *** rmart04 has joined #openstack-nova | 20:23 | |
| *** slaweq__ has joined #openstack-nova | 20:23 | |
| *** rmart04 has quit IRC | 20:24 | |
| *** slaweq_ has quit IRC | 20:25 | |
| *** salv-orlando has joined #openstack-nova | 20:26 | |
| *** dave-mccowan has quit IRC | 20:26 | |
| *** clenimar has quit IRC | 20:26 | |
| *** tjones has joined #openstack-nova | 20:29 | |
| *** tjones has left #openstack-nova | 20:29 | |
| *** haplo37_ has quit IRC | 20:30 | |
| *** ducttape_ has quit IRC | 20:30 | |
| *** code-R has quit IRC | 20:31 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Hyper-V: Adds Hyper-V UEFI Secure Boot https://review.openstack.org/209581 | 20:31 |
| *** code-R has joined #openstack-nova | 20:31 | |
| *** slaweq__ has quit IRC | 20:31 | |
| *** dimtruck is now known as zz_dimtruck | 20:31 | |
| mriedem | claudiub: +2 on ^ | 20:31 |
| *** haplo37_ has joined #openstack-nova | 20:32 | |
| *** Jeffrey4l_ has joined #openstack-nova | 20:35 | |
| *** Jeffrey4l has quit IRC | 20:35 | |
| claudiub | yeay! :D | 20:35 |
| *** tongli has quit IRC | 20:37 | |
| *** rmart04 has joined #openstack-nova | 20:37 | |
| *** raunak has joined #openstack-nova | 20:38 | |
| *** rlrossit_ has quit IRC | 20:39 | |
| mriedem | claudiub: and https://review.openstack.org/#/c/175479/ | 20:41 |
| *** lascii is now known as alaski | 20:41 | |
| claudiub | another yeay! :D | 20:43 |
| *** rmart04 has quit IRC | 20:45 | |
| *** jheroux has quit IRC | 20:45 | |
| *** slaweq__ has joined #openstack-nova | 20:46 | |
| *** krtaylor has quit IRC | 20:47 | |
| *** mriedem has quit IRC | 20:50 | |
| *** sdague has quit IRC | 20:50 | |
| openstackgerrit | Anusha Unnam proposed openstack/nova: Improve help text for libvirt options https://review.openstack.org/372761 | 20:50 |
| diana_clarke | I have 1 last database poison warning fix up for review, if anyone has some spare cycles. https://review.openstack.org/#/c/347531/ | 20:53 |
| *** mriedem has joined #openstack-nova | 20:53 | |
| *** swamireddy has joined #openstack-nova | 20:53 | |
| diana_clarke | I've set a goal to review 100 mox patches this cycle, if that helps show commitment in the other direction. Thanks! | 20:54 |
| *** ducttape_ has joined #openstack-nova | 20:54 | |
| *** ariesyous has left #openstack-nova | 20:54 | |
| *** thorst has joined #openstack-nova | 20:56 | |
| *** thorst has quit IRC | 20:56 | |
| *** thorst has joined #openstack-nova | 20:57 | |
| *** thorst has quit IRC | 21:01 | |
| *** thorst has joined #openstack-nova | 21:02 | |
| *** TonyXu has quit IRC | 21:02 | |
| *** TonyXu has joined #openstack-nova | 21:03 | |
| *** thorst has quit IRC | 21:04 | |
| *** maticue has quit IRC | 21:05 | |
| *** ducttape_ has quit IRC | 21:07 | |
| *** thorst has joined #openstack-nova | 21:07 | |
| openstackgerrit | Simona Iuliana Toader proposed openstack/nova-specs: Adds Hyper-V Shielded VMs spec https://review.openstack.org/274709 | 21:08 |
| *** xinliang has quit IRC | 21:10 | |
| *** lpetrut has quit IRC | 21:11 | |
| mriedem | claudiub: can you tell why the hyper-v ci failed here? https://review.openstack.org/#/c/273504/ | 21:11 |
| bauwser | superdan: oops, sorry, missed your -W on https://review.openstack.org/#/c/374307/ | 21:11 |
| bauwser | superdan: so, about the req_spec migration, that's actually a good point | 21:11 |
| bauwser | superdan: in theory, old instances should have been migrated if they run the main db online_migrations script | 21:12 |
| superdan | bauwser: right, this is to enforce that they have done that | 21:12 |
| *** thorst has quit IRC | 21:12 | |
| *** TonyXu has quit IRC | 21:12 | |
| superdan | bauwser: we need (a) a way to ensure that it's been done and (b) to decide if we're cool with forcing that it was done in newton | 21:13 |
| *** kberger has quit IRC | 21:13 | |
| superdan | I think we probably need it for cells anyway, so probably good to force it now | 21:13 |
| superdan | but wanted you to opine | 21:13 |
| bauwser | superdan: the main problem I see for operators is that they could have a lot of instances | 21:13 |
| *** TonyXu has joined #openstack-nova | 21:13 | |
| superdan | right | 21:13 |
| bauwser | superdan: meaning that maybe one cycle could be short for them | 21:13 |
| *** kberger has joined #openstack-nova | 21:13 | |
| bauwser | we could keep the backwards-compat code for yet another cycle | 21:14 |
| *** pradk has quit IRC | 21:14 | |
| bauwser | but in case we want to force, then yeah, we could perhaps look at the marker | 21:14 |
| superdan | this review is purely about forcing, and being able to delete the compat code | 21:15 |
| bauwser | yeah | 21:15 |
| superdan | I don't want to have morethan one blocker migration | 21:15 |
| superdan | so we need to decide | 21:15 |
| superdan | but, we can't do this until we branch, so think on it over the weekend and we can decide next week | 21:16 |
| bauwser | honestly, I think it's more a question for the operators | 21:16 |
| bauwser | ie. do they feel enough confident that they'll have time to run over all their instances? | 21:16 |
| * superdan winces | 21:16 | |
| superdan | I figure they won't know that until they're doing it, not knowing how much work it is | 21:17 |
| superdan | and I don't think we can keep this in for two years | 21:17 |
| bauwser | superdan: if you feel that's reasonable, then we could just lookup the existing marker | 21:17 |
| *** jamesdenton has quit IRC | 21:17 | |
| bauwser | if there is no marker, that's either they never ran the migration (which is nearly impossible if some other stuff migrated too) or that the migration ended | 21:18 |
| superdan | we have to be more specific about that I think | 21:18 |
| bauwser | superdan: one other way would be to just compare the number of non-deleted instances and the number of Requestspec records... | 21:19 |
| superdan | yeah | 21:19 |
| superdan | mriedem: I should review the release notes for newton, but we'll probably want to add some language around the fact that there are a *lot* of online migrations in newton and to make sure to do them, | 21:20 |
| superdan | and then again in ocata warning them to have done those before upgrading | 21:20 |
| superdan | because I know we're going to get complaints | 21:20 |
| bauwser | superdan: mriedem: we could just write a reno file for that like we did for Mitaka | 21:21 |
| *** coolsvap has quit IRC | 21:22 | |
| bauwser | superdan: the only thing I actually worry is that we don't delete the RequestSpec object when the instance is deleted too AFAIR | 21:22 |
| superdan | bauwser: really? | 21:22 |
| * bauwser checking | 21:23 | |
| mriedem | superdan: write the reno against the stable/newton branch | 21:23 |
| superdan | mriedem: yeah | 21:23 |
| bauwser | superdan: that's alaski who wrote the persisted side, so I need to doublecheck | 21:23 |
| superdan | bauwser: can we talk about this next week? | 21:24 |
| bauwser | agreed | 21:24 |
| superdan | bauwser: I'm running out of fucks | 21:24 |
| superdan | okay thanks :) | 21:24 |
| bauwser | k :) | 21:24 |
| *** xinliang has joined #openstack-nova | 21:24 | |
| bauwser | alaski: do you think we could have the same race conditions to handle like for the BuildRequest if we would like to delete the RequestSpec object when an instance is deleted ? | 21:25 |
| * mriedem lends superdan some fucks | 21:25 | |
| superdan | mriedem: I do not want used fucks | 21:26 |
| superdan | gross | 21:26 |
| mriedem | sloppy seconds bro | 21:27 |
| mriedem | i'm sorry | 21:27 |
| mriedem | even i'm offended by that | 21:27 |
| * superdan walks away | 21:27 | |
| *** gouthamr has quit IRC | 21:28 | |
| *** thumpba has joined #openstack-nova | 21:31 | |
| *** tanvirt has quit IRC | 21:31 | |
| *** thorst has joined #openstack-nova | 21:32 | |
| clarkb | why does nova config have auth_uri and auth_url set to different valeus under keystone_authtoken? | 21:33 |
| clarkb | example can eb seen at http://logs.openstack.org/74/372374/16/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ssh-tinyipa-nv/5dac5e8/logs/etc/nova/nova.conf.txt.gz not sure how users are supposed to understand whatever subtlety is at play there | 21:33 |
| *** cybojanek has joined #openstack-nova | 21:34 | |
| mriedem | i've wondered the same | 21:35 |
| mriedem | stevemar: ^ | 21:35 |
| *** thorst has quit IRC | 21:36 | |
| *** annegent_ has joined #openstack-nova | 21:37 | |
| *** thorst has joined #openstack-nova | 21:37 | |
| *** thorst has quit IRC | 21:37 | |
| *** panbalag has quit IRC | 21:38 | |
| *** rtheis has quit IRC | 21:38 | |
| *** GordonTX has quit IRC | 21:39 | |
| mriedem | BobBall: one thing on that device tagging tempest test, it's conditional on quite a few things, like ssh being enabled | 21:39 |
| mriedem | just fyi | 21:39 |
| *** GordonTX has joined #openstack-nova | 21:39 | |
| *** annegentle has quit IRC | 21:40 | |
| *** haplo37 has quit IRC | 21:40 | |
| *** mriedem has quit IRC | 21:41 | |
| *** thorst has joined #openstack-nova | 21:41 | |
| openstackgerrit | Anusha Unnam proposed openstack/nova: Config options consistency of ephemeral_storage.py https://review.openstack.org/348487 | 21:42 |
| *** haplo37 has joined #openstack-nova | 21:43 | |
| *** tonytan4ever has joined #openstack-nova | 21:43 | |
| *** code-R has quit IRC | 21:43 | |
| *** gouthamr has joined #openstack-nova | 21:44 | |
| *** thumpba has quit IRC | 21:46 | |
| *** thorst has quit IRC | 21:46 | |
| *** tonytan4ever has quit IRC | 21:47 | |
| *** tblakes has quit IRC | 21:54 | |
| *** yamahata has quit IRC | 21:57 | |
| *** rhagarty_ has quit IRC | 21:57 | |
| *** rhagarty has joined #openstack-nova | 21:58 | |
| *** yamahata has joined #openstack-nova | 21:59 | |
| *** kaisers_ has joined #openstack-nova | 22:00 | |
| *** mdrabe has quit IRC | 22:01 | |
| *** lamt has quit IRC | 22:01 | |
| *** mdrabe has joined #openstack-nova | 22:01 | |
| *** kberger has quit IRC | 22:02 | |
| *** kaisers_ has quit IRC | 22:05 | |
| *** xyang1 has quit IRC | 22:06 | |
| *** ducttape_ has joined #openstack-nova | 22:08 | |
| *** mdrabe has quit IRC | 22:11 | |
| *** GordonTX has quit IRC | 22:11 | |
| *** ducttape_ has quit IRC | 22:13 | |
| *** rrecio_ has quit IRC | 22:20 | |
| *** salv-orlando has quit IRC | 22:21 | |
| openstackgerrit | Sujitha proposed openstack/nova: Improve help text for network options. https://review.openstack.org/341203 | 22:22 |
| *** akshai has quit IRC | 22:25 | |
| *** armax has quit IRC | 22:28 | |
| openstackgerrit | Sujitha proposed openstack/nova: Remove deprecated manager classes in service.py https://review.openstack.org/338553 | 22:29 |
| openstackgerrit | Anusha Unnam proposed openstack/nova: Improve help text for interval_opts https://review.openstack.org/331761 | 22:32 |
| openstackgerrit | Eric Brown proposed openstack/nova: Fix config option types https://review.openstack.org/168481 | 22:40 |
| *** cdelatte has joined #openstack-nova | 22:43 | |
| *** tonytan4ever has joined #openstack-nova | 22:44 | |
| *** markvoelker has quit IRC | 22:47 | |
| *** tonytan4ever has quit IRC | 22:48 | |
| *** annegent_ has quit IRC | 22:48 | |
| *** Swami has quit IRC | 22:51 | |
| *** nherciu_ has quit IRC | 22:57 | |
| *** erlon has quit IRC | 22:57 | |
| *** annegentle has joined #openstack-nova | 22:58 | |
| oomichi | mtanino: that is super complex even in nova only | 22:59 |
| oomichi | oops, typo sorry | 23:00 |
| mtanino | sure | 23:00 |
| oomichi | mtreinish: that is super complex even in nova only | 23:00 |
| *** akshai has joined #openstack-nova | 23:03 | |
| openstackgerrit | Eric Brown proposed openstack/nova: VMware: Enforce minimum vCenter version of 5.5 https://review.openstack.org/253666 | 23:04 |
| *** vishwanathj has quit IRC | 23:05 | |
| *** haplo37 has quit IRC | 23:16 | |
| *** sdake has joined #openstack-nova | 23:17 | |
| *** ducttape_ has joined #openstack-nova | 23:18 | |
| *** ducttape_ has quit IRC | 23:22 | |
| *** figleaf has quit IRC | 23:24 | |
| *** haplo37 has joined #openstack-nova | 23:24 | |
| *** rama_y has quit IRC | 23:38 | |
| *** ijw has joined #openstack-nova | 23:39 | |
| *** akshai has quit IRC | 23:41 | |
| *** ducttape_ has joined #openstack-nova | 23:41 | |
| *** armax has joined #openstack-nova | 23:43 | |
| *** ducttape_ has quit IRC | 23:45 | |
| *** markvoelker has joined #openstack-nova | 23:48 | |
| *** ijw has quit IRC | 23:49 | |
| *** kaisers_ has joined #openstack-nova | 23:49 | |
| *** salv-orlando has joined #openstack-nova | 23:52 | |
| alaski | bauwser: no, deleting the req_spec is easy, just delete it. build_req is complex because it's used to synchronize between boot/delete | 23:52 |
| *** markvoelker has quit IRC | 23:52 | |
| *** kaisers_ has quit IRC | 23:54 | |
| *** salv-orlando has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!