| *** dims has quit IRC | 00:00 | |
| *** ijw has quit IRC | 00:01 | |
| *** markvoelker has joined #openstack-nova | 00:02 | |
| *** achanda has quit IRC | 00:03 | |
| *** eliqiao has joined #openstack-nova | 00:03 | |
| *** jcru has joined #openstack-nova | 00:04 | |
| *** jcru has quit IRC | 00:04 | |
| *** ZZelle_ has quit IRC | 00:05 | |
| *** shaohe_feng has quit IRC | 00:06 | |
| *** baoli has quit IRC | 00:06 | |
| *** Marga_ has quit IRC | 00:08 | |
| *** Marga_ has joined #openstack-nova | 00:08 | |
| *** Marga__ has joined #openstack-nova | 00:10 | |
| *** Marga_ has quit IRC | 00:10 | |
| *** igordcard_ has quit IRC | 00:13 | |
| *** annashen has quit IRC | 00:13 | |
| *** annegentle has quit IRC | 00:14 | |
| *** emagana has quit IRC | 00:14 | |
| *** baoli has joined #openstack-nova | 00:15 | |
| *** rmoe has quit IRC | 00:16 | |
| *** melwitt has quit IRC | 00:19 | |
| *** sdake has quit IRC | 00:22 | |
| *** hblixt has quit IRC | 00:26 | |
| *** Marga__ has quit IRC | 00:28 | |
| *** tsekiyama has joined #openstack-nova | 00:32 | |
| *** Nic has quit IRC | 00:33 | |
| *** rmoe has joined #openstack-nova | 00:33 | |
| *** baoli has quit IRC | 00:35 | |
| *** jcoufal has joined #openstack-nova | 00:38 | |
| *** jhesketh has quit IRC | 00:43 | |
| *** jhesketh has joined #openstack-nova | 00:43 | |
| *** lpetrut has joined #openstack-nova | 00:46 | |
| *** asettle has quit IRC | 00:47 | |
| *** lpetrut has quit IRC | 00:50 | |
| *** coolsvap|afk is now known as coolsvap | 00:50 | |
| *** Longgeek has joined #openstack-nova | 00:51 | |
| openstackgerrit | Luo Gangyi proposed openstack/nova: Ensure bridge exists before resuming a VM after an offline snapshot https://review.openstack.org/149942 | 00:53 |
|---|---|---|
| *** tsekiyam_ has joined #openstack-nova | 00:56 | |
| *** tsekiyama has quit IRC | 00:57 | |
| *** shaohe_feng has joined #openstack-nova | 00:57 | |
| *** salv-orlando has quit IRC | 01:04 | |
| *** jwang_ has joined #openstack-nova | 01:06 | |
| *** jwang__ has quit IRC | 01:09 | |
| *** mtanino has joined #openstack-nova | 01:09 | |
| *** emagana has joined #openstack-nova | 01:11 | |
| *** baoli has joined #openstack-nova | 01:11 | |
| *** annashen has joined #openstack-nova | 01:13 | |
| *** annegentle has joined #openstack-nova | 01:14 | |
| *** liusheng has quit IRC | 01:16 | |
| *** emagana has quit IRC | 01:17 | |
| *** emagana has joined #openstack-nova | 01:18 | |
| *** annashen has quit IRC | 01:19 | |
| *** annegentle has quit IRC | 01:20 | |
| *** emagana has quit IRC | 01:23 | |
| *** tsekiyam_ has quit IRC | 01:30 | |
| *** mtanino has quit IRC | 01:32 | |
| *** unicell1 has quit IRC | 01:35 | |
| *** otter768 has joined #openstack-nova | 01:39 | |
| *** dsanders has quit IRC | 01:39 | |
| *** otter768 has quit IRC | 01:43 | |
| *** dims has joined #openstack-nova | 01:44 | |
| *** dims_ has quit IRC | 01:46 | |
| *** browne1 has quit IRC | 01:48 | |
| *** iamjarvo has joined #openstack-nova | 01:49 | |
| *** iamjarvo has quit IRC | 01:49 | |
| *** jcru has joined #openstack-nova | 01:50 | |
| *** iamjarvo has joined #openstack-nova | 01:50 | |
| *** jcru has quit IRC | 01:51 | |
| *** mriedem1 has quit IRC | 01:51 | |
| *** kaufer has joined #openstack-nova | 01:54 | |
| shz | who can help me to review this patch:https://review.openstack.org/#/c/164128/ | 01:54 |
| *** jcru has joined #openstack-nova | 01:56 | |
| *** takedakn has joined #openstack-nova | 01:59 | |
| *** r-daneel has quit IRC | 02:02 | |
| *** hdd has quit IRC | 02:04 | |
| *** hdd has joined #openstack-nova | 02:04 | |
| *** salv-orlando has joined #openstack-nova | 02:05 | |
| *** VW_ has joined #openstack-nova | 02:06 | |
| *** takedakn has quit IRC | 02:11 | |
| *** erkules_ has joined #openstack-nova | 02:14 | |
| *** baoli has quit IRC | 02:14 | |
| *** baoli has joined #openstack-nova | 02:14 | |
| *** neelashah has joined #openstack-nova | 02:15 | |
| *** popw has joined #openstack-nova | 02:16 | |
| *** erkules has quit IRC | 02:16 | |
| *** VW_ has quit IRC | 02:16 | |
| *** annegentle has joined #openstack-nova | 02:16 | |
| *** ijw_ has quit IRC | 02:19 | |
| *** neelashah has quit IRC | 02:21 | |
| *** browne has joined #openstack-nova | 02:21 | |
| *** annegentle has quit IRC | 02:21 | |
| *** apuimedo has quit IRC | 02:22 | |
| *** unicell has joined #openstack-nova | 02:22 | |
| *** VW_ has joined #openstack-nova | 02:23 | |
| *** kaisers1 has joined #openstack-nova | 02:26 | |
| *** kaisers has quit IRC | 02:28 | |
| *** haomaiwang has joined #openstack-nova | 02:28 | |
| *** jcoufal has quit IRC | 02:29 | |
| *** IanGovett has quit IRC | 02:29 | |
| *** pixelb has joined #openstack-nova | 02:30 | |
| *** Maike has quit IRC | 02:33 | |
| *** Maike has joined #openstack-nova | 02:33 | |
| *** achanda has joined #openstack-nova | 02:37 | |
| *** salv-orlando has quit IRC | 02:38 | |
| *** salv-orlando has joined #openstack-nova | 02:39 | |
| shz | hi all, who can help me to review this patch:https://review.openstack.org/#/c/164128/ | 02:39 |
| *** hemna has joined #openstack-nova | 02:43 | |
| *** zul has quit IRC | 02:47 | |
| openstackgerrit | Lingxian Kong proposed openstack/nova: Raise exception when using invalid az for vm creation https://review.openstack.org/163842 | 02:49 |
| *** hdd has quit IRC | 02:50 | |
| *** dims has quit IRC | 02:51 | |
| lifeless | shz: are you the author | 02:53 |
| lifeless | shz: if so, you don't need help to review, you need help from a reviewer :) | 02:53 |
| *** achanda has quit IRC | 02:53 | |
| shz | yes, I need a help | 02:53 |
| *** achanda has joined #openstack-nova | 02:56 | |
| *** patrickeast has quit IRC | 02:56 | |
| *** popw1 has joined #openstack-nova | 02:57 | |
| *** zul has joined #openstack-nova | 02:59 | |
| *** popw has quit IRC | 03:00 | |
| shz | lifeless: can you help me? | 03:01 |
| *** marun has quit IRC | 03:01 | |
| *** dsanders has joined #openstack-nova | 03:02 | |
| lifeless | shz: I'm not a core on nova, but I've got the review open and may look at it later | 03:03 |
| lifeless | shz: at a minimum you need to rebase it though | 03:03 |
| lifeless | shz: oh, you have cool | 03:03 |
| shz | I have rebase it | 03:04 |
| *** salv-orlando has quit IRC | 03:05 | |
| *** asettle has joined #openstack-nova | 03:07 | |
| *** rfolco has quit IRC | 03:07 | |
| *** alex_xu has quit IRC | 03:12 | |
| *** jcru has quit IRC | 03:13 | |
| *** otter768 has joined #openstack-nova | 03:13 | |
| *** Longgeek_ has joined #openstack-nova | 03:22 | |
| *** hdd has joined #openstack-nova | 03:24 | |
| *** achanda has quit IRC | 03:24 | |
| *** jcru has joined #openstack-nova | 03:24 | |
| *** Longgeek has quit IRC | 03:25 | |
| openstackgerrit | Fei Long Wang proposed openstack/nova: Fix nova backup for volume-backed instance https://review.openstack.org/164494 | 03:25 |
| *** bkopilov has quit IRC | 03:29 | |
| *** achanda has joined #openstack-nova | 03:32 | |
| *** sdake has joined #openstack-nova | 03:36 | |
| *** salv-orlando has joined #openstack-nova | 03:37 | |
| *** hemna has quit IRC | 03:37 | |
| *** baoli has quit IRC | 03:37 | |
| *** achanda has quit IRC | 03:41 | |
| *** sdake has quit IRC | 03:42 | |
| *** jcru has quit IRC | 03:42 | |
| *** achanda has joined #openstack-nova | 03:43 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge V2 and V2.1 hypervisor functional tests https://review.openstack.org/167082 | 03:43 |
| *** armax has quit IRC | 03:43 | |
| *** jcru has joined #openstack-nova | 03:43 | |
| *** sdake has joined #openstack-nova | 03:44 | |
| *** armax has joined #openstack-nova | 03:44 | |
| *** jcru has quit IRC | 03:48 | |
| *** sbfox has joined #openstack-nova | 03:49 | |
| *** armax has quit IRC | 03:50 | |
| *** achanda has quit IRC | 03:53 | |
| *** pixelbeat has joined #openstack-nova | 03:53 | |
| *** pixelb has quit IRC | 03:54 | |
| *** sdake has quit IRC | 03:55 | |
| openstackgerrit | rajiv proposed openstack/python-novaclient: nova client now displays absolute limits when available https://review.openstack.org/166703 | 03:55 |
| *** marun has joined #openstack-nova | 03:58 | |
| *** achanda has joined #openstack-nova | 03:59 | |
| *** marun has quit IRC | 04:03 | |
| *** VW_ has quit IRC | 04:05 | |
| *** Marga_ has joined #openstack-nova | 04:06 | |
| *** Marga_ has quit IRC | 04:06 | |
| *** VW_ has joined #openstack-nova | 04:07 | |
| *** Marga_ has joined #openstack-nova | 04:07 | |
| *** deepthi has joined #openstack-nova | 04:08 | |
| *** salv-orlando has quit IRC | 04:09 | |
| *** dims has joined #openstack-nova | 04:12 | |
| *** hdd has quit IRC | 04:16 | |
| *** annegentle has joined #openstack-nova | 04:18 | |
| *** abhijeetm has joined #openstack-nova | 04:18 | |
| *** dims has quit IRC | 04:19 | |
| *** shz has quit IRC | 04:19 | |
| *** shz has joined #openstack-nova | 04:19 | |
| *** abhijeetm has quit IRC | 04:20 | |
| *** dave-mccowan has quit IRC | 04:22 | |
| *** otter768 has quit IRC | 04:22 | |
| *** annegentle has quit IRC | 04:23 | |
| *** VW_ has quit IRC | 04:26 | |
| *** asettle has quit IRC | 04:27 | |
| *** otter768 has joined #openstack-nova | 04:27 | |
| *** VW_ has joined #openstack-nova | 04:31 | |
| *** pixelbeat has quit IRC | 04:32 | |
| *** VW_ has quit IRC | 04:34 | |
| *** kaufer has quit IRC | 04:35 | |
| *** xyang1 has quit IRC | 04:35 | |
| *** amotoki has joined #openstack-nova | 04:39 | |
| *** xyang1 has joined #openstack-nova | 04:41 | |
| *** xyang1 has quit IRC | 04:43 | |
| *** coolsvap is now known as coolsvap|afk | 04:44 | |
| *** ishant has joined #openstack-nova | 04:48 | |
| *** iamjarvo has quit IRC | 04:50 | |
| *** amotoki has quit IRC | 04:51 | |
| *** amotoki has joined #openstack-nova | 04:52 | |
| *** amotoki has quit IRC | 04:52 | |
| *** amotoki has joined #openstack-nova | 04:52 | |
| *** bkopilov has joined #openstack-nova | 04:57 | |
| *** marun has joined #openstack-nova | 04:59 | |
| *** marun has quit IRC | 05:04 | |
| *** salv-orlando has joined #openstack-nova | 05:08 | |
| *** Marga_ has quit IRC | 05:10 | |
| *** marun has joined #openstack-nova | 05:11 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge V2 and V2.1 Agent functional tests https://review.openstack.org/167096 | 05:13 |
| gmann | park: sdague: cyeoh : created Etherpad for merge functional tests - https://etherpad.openstack.org/p/merge_sample_tests | 05:18 |
| *** annegentle has joined #openstack-nova | 05:19 | |
| *** harlowja is now known as harlowja_away | 05:22 | |
| *** annegentle has quit IRC | 05:23 | |
| cyeoh | gmann: excellent, thx! | 05:24 |
| *** otter768 has quit IRC | 05:25 | |
| *** boris-42 has quit IRC | 05:28 | |
| *** shaohe_feng has quit IRC | 05:34 | |
| *** ajayaa has joined #openstack-nova | 05:36 | |
| *** veena has joined #openstack-nova | 05:37 | |
| *** Nikolay_St has quit IRC | 05:40 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Fix v2.1 hypervisor servers to return empty list https://review.openstack.org/167100 | 05:43 |
| *** shaohe_feng has joined #openstack-nova | 05:43 | |
| *** park_heijlong has joined #openstack-nova | 05:44 | |
| gmann | cyeoh: ^^ hypervisor patch for v2.1 | 05:44 |
| gmann | cyeoh: please let me know if more needed on commit message | 05:45 |
| *** emagana has joined #openstack-nova | 05:45 | |
| cyeoh | gmann: ok will have a look in a sec | 05:45 |
| gmann | cyeoh: thanks | 05:45 |
| *** jcru has joined #openstack-nova | 05:45 | |
| *** salv-orlando has quit IRC | 05:47 | |
| *** jcru has quit IRC | 05:50 | |
| *** _gryf|afk is now known as _gryf | 06:00 | |
| *** Maike has quit IRC | 06:03 | |
| *** dims has joined #openstack-nova | 06:04 | |
| *** yamahata has quit IRC | 06:06 | |
| *** dims has quit IRC | 06:10 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Imported Translations from Transifex https://review.openstack.org/162030 | 06:14 |
| *** achanda has quit IRC | 06:15 | |
| *** annashen has joined #openstack-nova | 06:18 | |
| *** abhijeetm has joined #openstack-nova | 06:20 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge V2 and V2.1 aggregate functional tests https://review.openstack.org/167113 | 06:22 |
| *** annashen has quit IRC | 06:24 | |
| *** ankit_ag has joined #openstack-nova | 06:27 | |
| openstackgerrit | zhangtralon proposed openstack/nova: resize-revert can break anti-affinity policy https://review.openstack.org/160110 | 06:30 |
| *** Mike_D_laptop has joined #openstack-nova | 06:31 | |
| *** oro_ has joined #openstack-nova | 06:34 | |
| *** oro has joined #openstack-nova | 06:34 | |
| *** nkrinner has joined #openstack-nova | 06:35 | |
| *** pcaruana has quit IRC | 06:41 | |
| *** sahid has joined #openstack-nova | 06:43 | |
| *** lennyb has joined #openstack-nova | 06:43 | |
| *** moshele has joined #openstack-nova | 06:50 | |
| *** Marga_ has joined #openstack-nova | 06:52 | |
| *** oro_ has quit IRC | 06:53 | |
| *** irenab has quit IRC | 06:53 | |
| *** oro has quit IRC | 06:53 | |
| *** fesp has quit IRC | 06:54 | |
| *** dulek_ has joined #openstack-nova | 06:55 | |
| *** dulek_ is now known as dulek | 06:55 | |
| Mike_D_laptop | Hi folks! Please can you reviewed patch if you have a time? :) -> https://review.openstack.org/#/c/152964/ | 06:56 |
| *** Bsony has joined #openstack-nova | 06:57 | |
| *** dulek_ has joined #openstack-nova | 07:01 | |
| *** dulek has quit IRC | 07:04 | |
| openstackgerrit | zhangtralon proposed openstack/nova: check the configuration item api_max_retries https://review.openstack.org/123319 | 07:04 |
| *** browne has quit IRC | 07:07 | |
| *** moshele has quit IRC | 07:07 | |
| *** garyk has joined #openstack-nova | 07:09 | |
| *** sbfox has quit IRC | 07:10 | |
| *** salv-orlando has joined #openstack-nova | 07:16 | |
| *** irenab has joined #openstack-nova | 07:19 | |
| *** oro_ has joined #openstack-nova | 07:19 | |
| *** oro has joined #openstack-nova | 07:20 | |
| *** annegentle has joined #openstack-nova | 07:20 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge cell functional tests for V2 and V2.1 https://review.openstack.org/167119 | 07:21 |
| *** emagana has quit IRC | 07:22 | |
| *** emagana has joined #openstack-nova | 07:22 | |
| shz | hi all, who can help to review this patch:https://review.openstack.org/#/c/164128/ | 07:23 |
| *** fesp has joined #openstack-nova | 07:24 | |
| *** annegentle has quit IRC | 07:25 | |
| *** otter768 has joined #openstack-nova | 07:25 | |
| *** emagana has quit IRC | 07:27 | |
| *** atuvenie has joined #openstack-nova | 07:29 | |
| *** otter768 has quit IRC | 07:30 | |
| *** Marga_ has quit IRC | 07:33 | |
| *** Marga_ has joined #openstack-nova | 07:33 | |
| *** fesp has quit IRC | 07:35 | |
| openstackgerrit | Abhishek Kekane proposed openstack/nova-specs: Improve performance of UnShelve API https://review.openstack.org/135387 | 07:39 |
| *** mpavone has joined #openstack-nova | 07:40 | |
| openstackgerrit | garyk proposed openstack/nova: Move suds into test-requirements.txt https://review.openstack.org/166770 | 07:40 |
| garyk | jogo: ^ - addressed you comment with the driver specifc | 07:40 |
| *** deepthi has quit IRC | 07:41 | |
| *** deepthi has joined #openstack-nova | 07:41 | |
| *** dulek_ has quit IRC | 07:43 | |
| *** veena has quit IRC | 07:43 | |
| *** dulek has joined #openstack-nova | 07:43 | |
| *** browne has joined #openstack-nova | 07:45 | |
| *** yamahata has joined #openstack-nova | 07:45 | |
| *** yamahata has quit IRC | 07:46 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: add in folder support on VC https://review.openstack.org/165060 | 07:47 |
| *** Nikolay_St has joined #openstack-nova | 07:47 | |
| *** Marga_ has quit IRC | 07:49 | |
| *** dsanders has quit IRC | 07:51 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: improve instance names on VC https://review.openstack.org/166608 | 07:51 |
| *** dims has joined #openstack-nova | 07:55 | |
| *** markus_z has joined #openstack-nova | 07:56 | |
| openstackgerrit | zhangtralon proposed openstack/nova: filter_class_names is never passed https://review.openstack.org/143285 | 07:56 |
| openstackgerrit | zhangtralon proposed openstack/nova: filter_class_names is never passed https://review.openstack.org/143285 | 07:57 |
| *** TobiasE has joined #openstack-nova | 07:57 | |
| *** yamahata has joined #openstack-nova | 07:58 | |
| *** Longgeek_ has quit IRC | 07:58 | |
| *** ildikov has quit IRC | 08:00 | |
| *** flwang1 has joined #openstack-nova | 08:01 | |
| *** browne has quit IRC | 08:01 | |
| *** dims has quit IRC | 08:02 | |
| *** salv-orlando has quit IRC | 08:03 | |
| *** yingjun has joined #openstack-nova | 08:07 | |
| *** jyuso1 has joined #openstack-nova | 08:08 | |
| *** liusheng has joined #openstack-nova | 08:10 | |
| *** veena has joined #openstack-nova | 08:12 | |
| *** romainh has joined #openstack-nova | 08:13 | |
| shz | hi all, who can help review this patch:https://review.openstack.org/#/c/164128/ | 08:15 |
| *** pask81 has joined #openstack-nova | 08:17 | |
| *** yingjun has quit IRC | 08:19 | |
| *** annegentle has joined #openstack-nova | 08:21 | |
| *** erkules_ is now known as erkules | 08:21 | |
| *** erkules has quit IRC | 08:21 | |
| *** erkules has joined #openstack-nova | 08:21 | |
| *** yamahata has quit IRC | 08:22 | |
| *** yingjun has joined #openstack-nova | 08:22 | |
| *** oro_ has quit IRC | 08:23 | |
| *** pask81 has quit IRC | 08:23 | |
| *** oro has quit IRC | 08:24 | |
| *** annegentle has quit IRC | 08:26 | |
| *** aloga has quit IRC | 08:26 | |
| *** gszasz has joined #openstack-nova | 08:26 | |
| *** ndipanov has joined #openstack-nova | 08:28 | |
| *** tellesnobrega has quit IRC | 08:28 | |
| *** apuimedo has joined #openstack-nova | 08:31 | |
| *** apuimedo has quit IRC | 08:32 | |
| *** apuimedo has joined #openstack-nova | 08:33 | |
| *** aloga has joined #openstack-nova | 08:39 | |
| *** tellesnobrega has joined #openstack-nova | 08:40 | |
| *** Bsony has quit IRC | 08:41 | |
| openstackgerrit | zhangtralon proposed openstack/nova: filter_class_names is never passed https://review.openstack.org/167137 | 08:42 |
| *** aloga has quit IRC | 08:44 | |
| openstackgerrit | Qin Zhao proposed openstack/nova: Raise invalid input if use invalid ip for network to attach interface https://review.openstack.org/162894 | 08:46 |
| *** aloga has joined #openstack-nova | 08:46 | |
| openstackgerrit | zhangtralon proposed openstack/nova: filter_class_names is never passed https://review.openstack.org/143285 | 08:47 |
| *** liusheng has quit IRC | 08:52 | |
| *** salv-orlando has joined #openstack-nova | 08:52 | |
| *** andreykurilin_ has joined #openstack-nova | 08:53 | |
| *** Marga_ has joined #openstack-nova | 08:53 | |
| *** jlanoux has joined #openstack-nova | 08:57 | |
| *** andreykurilin_ has quit IRC | 08:59 | |
| *** pixelbeat has joined #openstack-nova | 09:00 | |
| *** andreykurilin_ has joined #openstack-nova | 09:00 | |
| *** park_heijlong has quit IRC | 09:02 | |
| *** eglynn has joined #openstack-nova | 09:04 | |
| *** jistr has joined #openstack-nova | 09:04 | |
| *** krykowski has joined #openstack-nova | 09:09 | |
| openstackgerrit | rajiv proposed openstack/python-novaclient: nova client now displays absolute limits when available https://review.openstack.org/166703 | 09:09 |
| *** oro_ has joined #openstack-nova | 09:10 | |
| *** oro has joined #openstack-nova | 09:10 | |
| *** lucasagomes has joined #openstack-nova | 09:11 | |
| *** zz_johnthetubagu is now known as johnthetubaguy | 09:12 | |
| *** shaohe_feng has quit IRC | 09:15 | |
| *** dims has joined #openstack-nova | 09:15 | |
| *** derekh has joined #openstack-nova | 09:15 | |
| *** gszasz has quit IRC | 09:16 | |
| *** zhangtralon has joined #openstack-nova | 09:17 | |
| shz | hi all, who can help review this patch:https://review.openstack.org/#/c/164128/ | 09:17 |
| *** moshele has joined #openstack-nova | 09:19 | |
| *** dims has quit IRC | 09:20 | |
| *** Marga_ has quit IRC | 09:21 | |
| openstackgerrit | zhangtralon proposed openstack/nova: limits don't update without using filters https://review.openstack.org/156029 | 09:22 |
| *** krykowski has left #openstack-nova | 09:22 | |
| *** andrearo1 has joined #openstack-nova | 09:23 | |
| *** jlanoux_ has joined #openstack-nova | 09:23 | |
| *** andrearosa has quit IRC | 09:25 | |
| *** jlanoux has quit IRC | 09:25 | |
| *** otter768 has joined #openstack-nova | 09:26 | |
| *** otter768 has quit IRC | 09:31 | |
| *** gszasz has joined #openstack-nova | 09:31 | |
| *** yingjun has quit IRC | 09:33 | |
| *** salv-orlando has quit IRC | 09:35 | |
| *** jcru has joined #openstack-nova | 09:37 | |
| *** Bsony has joined #openstack-nova | 09:38 | |
| *** jcru has quit IRC | 09:38 | |
| *** garyk has quit IRC | 09:40 | |
| *** lpetrut has joined #openstack-nova | 09:40 | |
| *** garyk has joined #openstack-nova | 09:40 | |
| openstackgerrit | Abhijeet Malawade proposed openstack/nova: Remove update_db decorator from attach method https://review.openstack.org/167145 | 09:41 |
| *** johnthetubaguy is now known as zz_johnthetubagu | 09:43 | |
| *** ZZelle has quit IRC | 09:44 | |
| *** ZZelle has joined #openstack-nova | 09:44 | |
| *** zz_johnthetubagu is now known as johnthetubaguy | 09:46 | |
| openstackgerrit | Sergey Nikitin proposed openstack/nova: Added server tags support in nova-api https://review.openstack.org/128940 | 09:50 |
| *** yassine has joined #openstack-nova | 09:51 | |
| *** achanda has joined #openstack-nova | 09:51 | |
| *** achanda has quit IRC | 09:56 | |
| *** zhangtralon has quit IRC | 09:58 | |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: Replace get_dynamic_property with get_object_property https://review.openstack.org/164220 | 09:58 |
| openstackgerrit | Radoslav Gerganov proposed openstack/nova: VMware: Replace get_dynamic_properties with get_object_properties_dict https://review.openstack.org/164604 | 09:58 |
| *** andreykurilin_ has quit IRC | 09:59 | |
| *** TobiasE has quit IRC | 09:59 | |
| *** andreykurilin_ has joined #openstack-nova | 09:59 | |
| *** gszasz has quit IRC | 09:59 | |
| *** popw1 has quit IRC | 10:02 | |
| *** alex_xu has joined #openstack-nova | 10:07 | |
| *** eliqiao has quit IRC | 10:08 | |
| *** gszasz has joined #openstack-nova | 10:15 | |
| *** andreykurilin__ has joined #openstack-nova | 10:17 | |
| *** andreykurilin_ has quit IRC | 10:17 | |
| *** dims has joined #openstack-nova | 10:19 | |
| *** haomaiwang has quit IRC | 10:21 | |
| *** markvoelker has quit IRC | 10:26 | |
| *** halfss has joined #openstack-nova | 10:27 | |
| bauzas | alaski: when you have time, I would like to discuss with you about something I saw for the Service object returning a ServiceProxy object | 10:29 |
| bauzas | alaski: long story short, as the Service.get_by_cn is a remotable classmethod, it can be called from the API to the conductor, and then the conductor is not knowing that it's for a Cells RPC API | 10:31 |
| *** ajayaa has quit IRC | 10:31 | |
| *** veena has quit IRC | 10:31 | |
| *** salv-orlando has joined #openstack-nova | 10:31 | |
| bauzas | alaski: so it returns directly a Service object, instead of a Service proxy object | 10:32 |
| *** Bsony_ has joined #openstack-nova | 10:34 | |
| bauzas | alaski: see http://paste.openstack.org/show/195890/ | 10:35 |
| *** Bsony has quit IRC | 10:35 | |
| bauzas | alaski: instead, we need to call the Service classmethod by adding a new arg (like cell=true) | 10:36 |
| *** andreykurilin_ has joined #openstack-nova | 10:37 | |
| *** andreykurilin__ has quit IRC | 10:37 | |
| bauzas | alaski: ping me when you have time, so we can discuss about how to fix it properly | 10:37 |
| openstackgerrit | Abhijeet Malawade proposed openstack/nova: Remove update_db decorator from attach method https://review.openstack.org/167145 | 10:39 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: WIP: Fix how service objects are looked up for Cells https://review.openstack.org/160506 | 10:40 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Fix how the Cells API is returning ComputeNode objects https://review.openstack.org/157185 | 10:40 |
| bauzas | alaski: see https://review.openstack.org/160506, I uploaded it | 10:41 |
| *** andreykurilin_ has quit IRC | 10:45 | |
| *** Bsony has joined #openstack-nova | 10:46 | |
| *** Bsony has quit IRC | 10:47 | |
| *** Bsony_ has quit IRC | 10:48 | |
| shz | hi all, who can help to review this patch:https://review.openstack.org/#/c/164128/ | 10:49 |
| *** Bsony has joined #openstack-nova | 10:52 | |
| *** vigneshvar has joined #openstack-nova | 10:54 | |
| *** Bsony_ has joined #openstack-nova | 10:57 | |
| *** TobiasE has joined #openstack-nova | 10:58 | |
| *** Bsony has quit IRC | 10:59 | |
| *** Bsony_ has quit IRC | 11:02 | |
| *** Bsony has joined #openstack-nova | 11:05 | |
| *** ildikov has joined #openstack-nova | 11:07 | |
| openstackgerrit | Sean Dague proposed openstack/nova: Use migration object in resource_tracker https://review.openstack.org/79324 | 11:08 |
| *** pask81 has joined #openstack-nova | 11:10 | |
| *** Bsony_ has joined #openstack-nova | 11:13 | |
| *** shaohe_feng has joined #openstack-nova | 11:14 | |
| *** iamjarvo has joined #openstack-nova | 11:14 | |
| *** deepthi has quit IRC | 11:15 | |
| *** Bsony has quit IRC | 11:16 | |
| *** ildikov has quit IRC | 11:17 | |
| *** ajayaa has joined #openstack-nova | 11:17 | |
| *** Bsony_ has quit IRC | 11:18 | |
| *** exploreshaifali has joined #openstack-nova | 11:20 | |
| *** exploreshaifali has quit IRC | 11:23 | |
| *** annashen has joined #openstack-nova | 11:23 | |
| *** IanGovett has joined #openstack-nova | 11:24 | |
| *** iamjarvo has quit IRC | 11:24 | |
| mdbooth | Trivia: https://review.openstack.org/#/c/165773/ https://review.openstack.org/#/c/165784/ | 11:25 |
| mdbooth | 7 lines and 4 lines | 11:25 |
| mdbooth | Does anybody know how to pull a specific unmerged change out of gerrit using plain git? | 11:26 |
| mdbooth | I'm expecting the answer to be something like 'git fetch gerrit <refspec>' | 11:26 |
| *** markvoelker has joined #openstack-nova | 11:26 | |
| mdbooth | where refspec is something non-default | 11:26 |
| lxsli | mdbooth: what's wrong with the cherry-pick line Gerrit gives you? | 11:26 |
| mdbooth | lxsli: You can't cherry pick a commit you haven't fetched | 11:27 |
| lxsli | EG git fetch https://review.openstack.org/openstack/nova refs/changes/28/164128/3 && git cherry-pick FETCH_HEAD | 11:27 |
| *** otter768 has joined #openstack-nova | 11:27 | |
| mdbooth | lxsli: That looks like it :) | 11:27 |
| mdbooth | lxsli: Assuming 165128 is the change number, 3 is the patch revision | 11:28 |
| *** annashen has quit IRC | 11:28 | |
| mdbooth | What's 28? | 11:28 |
| lxsli | In the "Patch Set 3" box, besides Download, click "cherry-pick" tab | 11:28 |
| lxsli | 28 appears to be the last two digits of the change number | 11:28 |
| mdbooth | Ooh | 11:28 |
| mdbooth | looking | 11:28 |
| lxsli | I've experimented and that seems to work, don't ask me why! | 11:28 |
| *** jlanoux has joined #openstack-nova | 11:28 | |
| mdbooth | :q | 11:30 |
| mdbooth | Eurgh. | 11:30 |
| * mdbooth needs eyeball focus | 11:30 | |
| *** jlanoux_ has quit IRC | 11:31 | |
| *** Bsony has joined #openstack-nova | 11:31 | |
| *** markvoelker has quit IRC | 11:31 | |
| *** andrearo1 is now known as andrearosa | 11:32 | |
| *** otter768 has quit IRC | 11:32 | |
| mdbooth | lxsli: That's awesome. Exactly what I was looking for. Thanks! | 11:32 |
| *** Bsony has quit IRC | 11:36 | |
| jaypipes | bauzas: crazy plane crash right near you :( | 11:38 |
| *** zul has quit IRC | 11:40 | |
| *** zul has joined #openstack-nova | 11:41 | |
| *** shaohe_feng has quit IRC | 11:41 | |
| openstackgerrit | Yasuaki Nagata proposed openstack/nova: Do 'detach' if a timeout happened in the process of attach_volume https://review.openstack.org/167183 | 11:44 |
| lxsli | mdbooth: sweet :) | 11:44 |
| *** jamielennox is now known as jamielennox|away | 11:45 | |
| openstackgerrit | Hans Lindgren proposed openstack/nova: Set conductor use_local flag in compute manager tests https://review.openstack.org/167184 | 11:46 |
| *** salv-orlando has quit IRC | 11:46 | |
| *** iamjarvo has joined #openstack-nova | 11:47 | |
| *** jlanoux_ has joined #openstack-nova | 11:47 | |
| *** ociuhandu has joined #openstack-nova | 11:47 | |
| *** halfss has quit IRC | 11:48 | |
| *** wuhao_ has joined #openstack-nova | 11:48 | |
| *** flwang1 has quit IRC | 11:49 | |
| *** jlanoux has quit IRC | 11:50 | |
| *** dulek_ has joined #openstack-nova | 11:55 | |
| *** xuhanp has joined #openstack-nova | 11:58 | |
| *** dulek has quit IRC | 11:58 | |
| *** xuhanp_ has joined #openstack-nova | 11:58 | |
| *** markvoelker has joined #openstack-nova | 11:58 | |
| *** gszasz has quit IRC | 12:01 | |
| *** dulek_ has quit IRC | 12:04 | |
| *** _gryf has quit IRC | 12:06 | |
| *** atuvenie has quit IRC | 12:06 | |
| *** jaypipes has quit IRC | 12:06 | |
| *** _gryf has joined #openstack-nova | 12:07 | |
| *** jaypipes has joined #openstack-nova | 12:07 | |
| *** boris-42 has joined #openstack-nova | 12:08 | |
| *** Bsony has joined #openstack-nova | 12:09 | |
| *** dkliban_afk is now known as dkliban | 12:14 | |
| *** atuvenie has joined #openstack-nova | 12:14 | |
| *** johnthetubaguy is now known as zz_johnthetubagu | 12:15 | |
| *** ZZelle has quit IRC | 12:15 | |
| *** lucasagomes is now known as lucas-hungry | 12:15 | |
| *** gszasz has joined #openstack-nova | 12:15 | |
| *** dims has quit IRC | 12:15 | |
| *** ZZelle has joined #openstack-nova | 12:15 | |
| *** _gryf has quit IRC | 12:15 | |
| *** dims has joined #openstack-nova | 12:16 | |
| *** ildikov has joined #openstack-nova | 12:18 | |
| *** Bsony has quit IRC | 12:20 | |
| *** jcoufal has joined #openstack-nova | 12:22 | |
| bauzas | jaypipes: yeah :-( that's weird to see the plane there because it's not the right route | 12:23 |
| *** salv-orlando has joined #openstack-nova | 12:24 | |
| *** zz_johnthetubagu is now known as johnthetubaguy | 12:25 | |
| jaypipes | bauzas: usually goes over your house? ;) | 12:28 |
| *** pask81 has quit IRC | 12:29 | |
| *** dulek has joined #openstack-nova | 12:30 | |
| *** jistr is now known as jistr|class | 12:30 | |
| alaski | bauzas: just looked over what you were talking about on https://review.openstack.org/#/c/160506/8 | 12:32 |
| bauzas | alaski: yup, and ?:) | 12:33 |
| alaski | bauzas: one option is to require that conductors at the cell level be configured as api cells | 12:33 |
| bauzas | jaypipes: yeah, kind of, when traveling over the Alps, it's usually taking a direct route | 12:33 |
| *** aysyd has joined #openstack-nova | 12:33 | |
| alaski | bauzas: the other would be to do what dheeraj suggests, and move the swith up to the API code | 12:34 |
| *** dave-mccowan has joined #openstack-nova | 12:34 | |
| bauzas | alaski: mmm, thinking about it | 12:34 |
| alaski | I really don't want to update the rpc to pass a cells param | 12:35 |
| bauzas | alaski: agreed with you, it's a bad thing, hence why I asked you | 12:35 |
| bauzas | alaski: well, let's try what dheeraj suggests | 12:35 |
| *** irenab has quit IRC | 12:36 | |
| alaski | bauzas: I do think it's reasonable to expect conductors in a cells setup to have cells turned on and designated as being at the api level | 12:36 |
| alaski | bauzas: sounds good | 12:36 |
| *** ishant has quit IRC | 12:36 | |
| bauzas | alaski: yeah about conductors but I think it's a huge change for something which would be deprecated soon | 12:37 |
| bauzas | by soon, I would rather say "sooner or later' | 12:37 |
| *** rushiagr_away is now known as rushiagr | 12:37 | |
| *** rfolco has joined #openstack-nova | 12:37 | |
| alaski | bauzas: it's not a code change though, just deployment expectation. We're not running conductors yet but when we do we'll have them configured that way | 12:38 |
| bauzas | alaski: I don't get your point, conductors are not knowing cells, right ? | 12:38 |
| alaski | bauzas: I mean that "if cells_opts.get_cell_type() == 'api':" should work | 12:39 |
| bauzas | alaski: yeah, but that means that we need to change the base conductor proxy | 12:39 |
| bauzas | alaski: because objects are not calling conductor methods, but rather using them for calling the DB remotely | 12:40 |
| bauzas | thru a proxy | 12:40 |
| bauzas | alaski: https://github.com/openstack/nova/blob/master/nova/conductor/manager.py#L412-L460 | 12:41 |
| alaski | ahh, I see | 12:42 |
| bauzas | alaski: and https://github.com/openstack/nova/blob/master/nova/objects/base.py#L154-L171 | 12:42 |
| alaski | I thought the issue was that the wrong thing would be called on the conductor, it's the translation on the way back | 12:42 |
| *** Marga_ has joined #openstack-nova | 12:42 | |
| bauzas | alaski: to be clear, the Service.get_by_cn method is then called on the conductor side and is returning the service object because it doesn't know about the cells at all | 12:44 |
| *** ildikov has quit IRC | 12:44 | |
| *** dprince has joined #openstack-nova | 12:44 | |
| *** iamjarvo has quit IRC | 12:48 | |
| alex_xu | good morning, a question: do we support upgrade the conductor service separate with other service now? or now we still need upgrade all the controller service at once? | 12:48 |
| alaski | bauzas: right, I see that now. I think bypassing the object will be the best solution for now. | 12:48 |
| bauzas | alaski: I'm just worried by what dansmith could say :D | 12:49 |
| bauzas | alaski: because we (dan and I) both agreed that just proxying the object methods in such a compute api thing was a bad thing | 12:49 |
| alaski | bauzas: another possibility would be to add a routing_path field to the object which isn't tied to the db model and have accesses for 'id' append the routing path if it's set | 12:49 |
| alaski | or access the routing path if cells are enabled | 12:50 |
| bauzas | alaski: mmm | 12:51 |
| *** exploreshaifali has joined #openstack-nova | 12:51 | |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova: libvirt: introduce libosinfo library to set hardware policy https://review.openstack.org/149625 | 12:51 |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova: libvirt: use osinfo when configuring network model https://review.openstack.org/149627 | 12:51 |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova: libvirt: adding libosinfo configuration https://review.openstack.org/149626 | 12:51 |
| openstackgerrit | Vladik Romanovsky proposed openstack/nova: libvirt: use osinfo when configuring the disk bus https://review.openstack.org/149628 | 12:51 |
| alaski | bauzas: yeah, I agree with you and dan. just trying to think of minimally invasive solutions | 12:52 |
| bauzas | alaski: yeah... | 12:52 |
| bauzas | alaski: trying to summarize the problem | 12:53 |
| bauzas | alaski: so, remotable object methods are directly calling the conductor without proxying thru the cells RPCAPI | 12:53 |
| *** jang1 has joined #openstack-nova | 12:53 | |
| bauzas | alaski: probably a change in the remotable decorator on the obbjects base module could prevent that behaviour | 12:54 |
| alaski | bauzas: well, the call will go through the cells rpcapi on the conductor side | 12:54 |
| alaski | right? | 12:55 |
| *** mpaolino has joined #openstack-nova | 12:55 | |
| bauzas | alaski: yeah but that's not the right path, right ? | 12:55 |
| bauzas | alaski: I'm not sure that https://review.openstack.org/#/c/160506/8/nova/objects/service.py,cm L151 is true on the conductor side... | 12:56 |
| bauzas | alaski: see http://logs.openstack.org/06/160506/8/check/gate-nova-python27/b3bf259/console.html | 12:56 |
| alaski | bauzas: I think it is, though that depends on the setup | 12:56 |
| openstackgerrit | garyk proposed openstack/nova: Virt: update shared storage base methods https://review.openstack.org/162457 | 12:57 |
| bauzas | alaski: http://logs.openstack.org/06/160506/8/check/gate-nova-python27/b3bf259/console.html#_2015-03-24_10_58_06_440 | 12:57 |
| alaski | bauzas: to me it seems that the issue is https://github.com/openstack/nova/blob/master/nova/conductor/manager.py#L436 and what it hydrates to on the other side | 12:57 |
| bauzas | alaski: that just means we're not entering the conditional | 12:57 |
| openstackgerrit | Monty Taylor proposed openstack/nova-specs: Fetch SSH Hostkey https://review.openstack.org/167202 | 12:57 |
| *** ferest has joined #openstack-nova | 12:57 | |
| *** ildikov has joined #openstack-nova | 12:58 | |
| *** matrohon has joined #openstack-nova | 12:58 | |
| *** subscope has quit IRC | 12:58 | |
| bauzas | alaski: well, unfortunately I can't trace this using pdb... | 12:58 |
| bauzas | alaski: I can only trace the client side, and I don't know exactly how to trace it for the fake RPC driver | 12:58 |
| *** mpaolino has quit IRC | 12:59 | |
| alaski | bauzas: I have resorted to writing logs to a file on the remote side before... | 13:00 |
| *** mpaolino has joined #openstack-nova | 13:00 | |
| *** rushiagr is now known as rushiagr_away | 13:00 | |
| alaski | I'm pulling the change down to test | 13:01 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Fix API links and labels https://review.openstack.org/166985 | 13:01 |
| bauzas | alaski: re: https://github.com/openstack/nova/blob/master/nova/conductor/manager.py#L436 that's just because we're serializing the object when routing it, but then we rehydrate it on the client side | 13:02 |
| bauzas | IIUC of course | 13:02 |
| *** thomasem has quit IRC | 13:02 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Fix API links and labels https://review.openstack.org/166985 | 13:03 |
| bauzas | alaski: that's for compability checking | 13:03 |
| alaski | bauzas: yeah, but that will hydrate to a Service object, not the ServiceProxy | 13:03 |
| bauzas | alaski: if it was a ServiceProxy object, it would be returned directly | 13:03 |
| bauzas | alaski: here, I'm expecting a ServiceProxy object and I got a Service NovaObject instead | 13:04 |
| *** salv-orlando has quit IRC | 13:04 | |
| alaski | ok, ignore me. I should read the end of the else statement | 13:04 |
| bauzas | alaski: IMHO, the only way to get it is due to the fact I'm not entering the conditional | 13:04 |
| bauzas | alaski: np | 13:05 |
| *** changbl has quit IRC | 13:06 | |
| alaski | bauzas: oh, the flags probably aren't being carried over the the conductor service that starts up | 13:07 |
| bauzas | alaski: that's my opinion | 13:07 |
| *** amotoki has quit IRC | 13:09 | |
| *** salv-orlando has joined #openstack-nova | 13:09 | |
| *** ferest has quit IRC | 13:10 | |
| *** amotoki has joined #openstack-nova | 13:12 | |
| openstackgerrit | wuhao proposed openstack/nova-specs: Add previous page links for Nova GET servers API https://review.openstack.org/165041 | 13:13 |
| *** diegows has joined #openstack-nova | 13:13 | |
| *** thomasem has joined #openstack-nova | 13:14 | |
| *** doude has quit IRC | 13:16 | |
| *** eharney has quit IRC | 13:16 | |
| *** iamjarvo has joined #openstack-nova | 13:17 | |
| *** iamjarvo has quit IRC | 13:17 | |
| *** iamjarvo has joined #openstack-nova | 13:18 | |
| *** iamjarvo has joined #openstack-nova | 13:19 | |
| *** kaufer has joined #openstack-nova | 13:19 | |
| *** neelashah has joined #openstack-nova | 13:19 | |
| *** neelashah1 has joined #openstack-nova | 13:20 | |
| *** Marga_ has quit IRC | 13:21 | |
| *** zzzeek has joined #openstack-nova | 13:21 | |
| wuhao_ | I want to add a spec to enable previous page query when using /servers API. I wonder if anyone could make a review. | 13:22 |
| *** neelashah has quit IRC | 13:24 | |
| *** salv-orl_ has joined #openstack-nova | 13:24 | |
| *** sakthi has joined #openstack-nova | 13:25 | |
| *** lpetrut has quit IRC | 13:26 | |
| *** salv-orlando has quit IRC | 13:27 | |
| *** lpetrut has joined #openstack-nova | 13:27 | |
| *** otter768 has joined #openstack-nova | 13:28 | |
| *** salv-orlando has joined #openstack-nova | 13:28 | |
| sakthi | resize of vcpu from libvirt -> KVM doesn't support add module level, how can establish in openstack | 13:29 |
| *** lucas-hungry is now known as lucasagomes | 13:29 | |
| *** iamjarvo has quit IRC | 13:30 | |
| *** baoli has joined #openstack-nova | 13:30 | |
| *** salv-orl_ has quit IRC | 13:32 | |
| *** otter768 has quit IRC | 13:33 | |
| *** rushiagr_away is now known as rushiagr | 13:33 | |
| *** claudiub has joined #openstack-nova | 13:35 | |
| *** Bsony has joined #openstack-nova | 13:35 | |
| openstackgerrit | Sean Dague proposed openstack/nova: don't 500 on invalid security group format https://review.openstack.org/167220 | 13:35 |
| *** ildikov has quit IRC | 13:36 | |
| *** vladikr has joined #openstack-nova | 13:37 | |
| *** dboik has joined #openstack-nova | 13:38 | |
| *** dboik has quit IRC | 13:38 | |
| *** dboik has joined #openstack-nova | 13:39 | |
| ankit_ag | Hi all, can someone please review https://review.openstack.org/145738 | 13:39 |
| alaski | bauzas: oslo_messaging is serializing it to a Service object | 13:40 |
| bauzas | alaski: erm ? | 13:40 |
| bauzas | alaski: say it again ? | 13:40 |
| alaski | bauzas: https://github.com/openstack/oslo.messaging/blob/master/oslo_messaging/rpc/dispatcher.py#L131 | 13:40 |
| alaski | that converts it from ServiceProxy to a serialized ServiceObject | 13:40 |
| alaski | bauzas: the conductor method is returning ServiceProxy properly though | 13:41 |
| *** burt has joined #openstack-nova | 13:42 | |
| alaski | bauzas: http://paste.openstack.org/show/195975/ | 13:42 |
| *** armax has joined #openstack-nova | 13:42 | |
| dansmith | alaski: it should only hydrate to a NovaObject if it's serialized like one | 13:43 |
| dansmith | alaski: ServiceProxy is just a dict, no? | 13:44 |
| *** mriedem has joined #openstack-nova | 13:44 | |
| alaski | just a class with the Service object as an attribute | 13:44 |
| *** r-daneel has joined #openstack-nova | 13:44 | |
| dansmith | the thing you're returning from conductor? | 13:46 |
| *** mwagner_lap has quit IRC | 13:46 | |
| sdague | so, who is actually theoretically working on pci support? Because it seems pretty broken based on the bugs I'm seeing. | 13:46 |
| *** mtanino has joined #openstack-nova | 13:46 | |
| dansmith | sdague: it works, but it's very flaky and temperamental | 13:47 |
| sdague | https://bugs.launchpad.net/nova/+bug/1435483 | 13:47 |
| openstack | Launchpad bug 1435483 in OpenStack Compute (nova) " [pci-passthrough] Failed to launch vm after restarting nova compute " [Undecided,New] | 13:47 |
| dansmith | sdague: yjiang5 does the most consistent work on it I think | 13:47 |
| alaski | dansmith: before serialization, yes | 13:47 |
| *** jcru has joined #openstack-nova | 13:47 | |
| dansmith | sdague: yes, that bug has been in since we crammed stuff into juno at the end | 13:48 |
| alaski | dansmith: it does have __getitem__ defined to proxy through to the object, so it's basically a dict | 13:48 |
| dansmith | sdague: there have been patches up for it since almost then, let me see if they merged | 13:48 |
| sdague | so that's a dupe of the critical bug? | 13:48 |
| dansmith | alaski: okay, I probably need to re-read more of the backlog | 13:48 |
| dansmith | sdague: probably.. we just changed the format of our config option :( | 13:49 |
| alaski | dansmith: mainly we're tracking where exactly the failure was to figure out the best workaround | 13:49 |
| dansmith | sdague: https://bugs.launchpad.net/nova/+bug/1383465 | 13:49 |
| *** eglynn has quit IRC | 13:49 | |
| openstack | Launchpad bug 1383465 in OpenStack Compute (nova) "[pci-passthrough] nova-compute fails to start" [Critical,Triaged] | 13:49 |
| sdague | right, so that's different | 13:49 |
| bauzas | alaski: checking | 13:49 |
| sdague | because nova-compute is up here, but no vms can be started | 13:49 |
| *** jcru has quit IRC | 13:49 | |
| *** jcru has joined #openstack-nova | 13:50 | |
| bauzas | alaski: interesting... | 13:50 |
| dansmith | ah, sure enough | 13:50 |
| bauzas | alaski: why the heck a python object is transformed into a Service object ? | 13:50 |
| *** ildikov has joined #openstack-nova | 13:51 | |
| bauzas | alaski: I'm testing it locally | 13:51 |
| dansmith | sdague: oye, the older critical bug is still stuck in discussion | 13:52 |
| dansmith | sdague: it's _so_ bad that we broke the config and still haven't fixed it :( | 13:52 |
| sdague | yep | 13:52 |
| sdague | ok, so... how do we get out of this hole? | 13:52 |
| dansmith | the one you just linked is something I can take a look at | 13:53 |
| *** xgsa has quit IRC | 13:53 | |
| alaski | bauzas: not sure how that works quite yet | 13:53 |
| sdague | it feels like we should mark all the pci code as experimental until issues are resolved here | 13:53 |
| *** jistr|class is now known as jistr | 13:53 | |
| bauzas | alaski: looking at oslo.messaging now | 13:53 |
| *** xgsa has joined #openstack-nova | 13:53 | |
| sdague | I really don't like people thinking this stuff is going to work, and it falls down a bunch if you look at it incorrectly | 13:53 |
| sdague | dansmith: before you go dive, want to take a quick look at - https://review.openstack.org/167220 bug fix | 13:53 |
| sdague | it should be quick | 13:54 |
| dansmith | sdague: I think maybe mentioning that pci continues to be flaky in the release notes "known issues" would be fine | 13:54 |
| *** doude has joined #openstack-nova | 13:55 | |
| dansmith | sdague: I certainly don't want to start tagging whole things that have been in-tree for a while as experimental every time we see a surge of bugs | 13:55 |
| sdague | dansmith: ok | 13:55 |
| *** achanda has joined #openstack-nova | 13:55 | |
| dansmith | sdague: yeah, I looked at that one | 13:55 |
| dansmith | sdague: string freeze sucks. | 13:55 |
| dansmith | jaypipes: bauzas: what recently merged around resource tracker that could cause update_resource_stats() to choke on pci stuff? | 13:56 |
| bauzas | dansmith: which one are you thinking about ? | 13:57 |
| dansmith | bauzas: https://launchpadlibrarian.net/201061744/n-cpu.log | 13:57 |
| *** yamahata has joined #openstack-nova | 13:58 | |
| bauzas | dansmith: sec, checking | 13:58 |
| sdague | dansmith: well, it should be fine in string freeze because it uses an existing string | 13:59 |
| *** thangp has joined #openstack-nova | 13:59 | |
| sdague | so that's already in the corpus | 13:59 |
| dansmith | sdague: I know, the error string is confusing because it's reusing something else | 13:59 |
| *** baoli has quit IRC | 13:59 | |
| jaypipes | dansmith: I've been wondering that too. | 13:59 |
| dansmith | sdague: I know | 13:59 |
| sdague | dansmith: it's actually not too bad | 13:59 |
| dansmith | jaypipes: you've been seeing that? | 13:59 |
| *** obutenko has quit IRC | 13:59 | |
| jaypipes | dansmith: this right? https://bugs.launchpad.net/nova/+bug/1435483 | 13:59 |
| openstack | Launchpad bug 1435483 in OpenStack Compute (nova) " [pci-passthrough] Failed to launch vm after restarting nova compute " [Undecided,Incomplete] | 13:59 |
| dansmith | jaypipes: yeah | 14:00 |
| *** baoli has joined #openstack-nova | 14:00 | |
| *** achanda has quit IRC | 14:00 | |
| *** dboik_ has joined #openstack-nova | 14:00 | |
| *** neelashah1 has quit IRC | 14:00 | |
| jaypipes | dansmith: you want me to have someone on my team look into it? | 14:01 |
| *** wuhao_ has quit IRC | 14:01 | |
| dansmith | jaypipes: I'm looking at it now, I'm just wondering if it sparks any ideas of things that might've merged recently that could be related | 14:02 |
| jaypipes | dansmith: not off the top of my head, no... | 14:02 |
| bauzas | dansmith: well, the update_resource_stats code was not touched recently IIRC | 14:02 |
| *** obutenko has joined #openstack-nova | 14:02 | |
| *** mriedem has quit IRC | 14:03 | |
| *** mriedem has joined #openstack-nova | 14:03 | |
| *** eharney has joined #openstack-nova | 14:03 | |
| *** lpetrut has quit IRC | 14:03 | |
| *** dboik has quit IRC | 14:04 | |
| jogo | garyk: thanks | 14:04 |
| *** xgsa has quit IRC | 14:05 | |
| dansmith | bauzas: it's failing when sending update_resource_stats to the scheduler | 14:05 |
| *** obutenko has quit IRC | 14:05 | |
| bauzas | dansmith: yup, that's what I'm seeing | 14:05 |
| *** jecarey has joined #openstack-nova | 14:05 | |
| bauzas | dansmith: nothing changed on the scheduler cli side | 14:05 |
| *** xgsa has joined #openstack-nova | 14:06 | |
| bauzas | dansmith: so that's coming from a change in the RT | 14:06 |
| *** neelashah has joined #openstack-nova | 14:06 | |
| bauzas | dansmith: currently blaming it | 14:06 |
| dansmith | bauzas: so it's setting a non-object on compute_node.stats or something? | 14:06 |
| *** VW_ has joined #openstack-nova | 14:06 | |
| bauzas | dansmith: well, the pci_device_pools field is waiting a PCIDevicePoolList object while it's not | 14:08 |
| *** vigneshvar has quit IRC | 14:08 | |
| dansmith | right | 14:08 |
| bauzas | dansmith: now I'm looking at the trace to see where the RT is called for updating those stats | 14:09 |
| dansmith | yeah, it's something in self.stats in RT, right? | 14:09 |
| garyk | mriedem: can you please look at https://review.openstack.org/166770 (this moves suds from requirements to test-requirements - it is only used by vmwareapi) | 14:09 |
| sdague | garyk: can you triage this bug - https://bugs.launchpad.net/nova/+bug/1430941 ? | 14:09 |
| openstack | Launchpad bug 1430941 in OpenStack Compute (nova) "vmware: snapshot not passing image properties" [Undecided,New] | 14:09 |
| garyk | sdague: sure | 14:09 |
| *** sdake has joined #openstack-nova | 14:10 | |
| sdague | garyk: +A on the req patch above | 14:10 |
| bauzas | dansmith: that's just horrible to look at a plain traceback... :) | 14:10 |
| mriedem | garyk: looks like sdague already got it | 14:10 |
| dansmith | bauzas: although we've already done compute_node.update(values) there | 14:10 |
| bauzas | dansmith: but sounds like it's called because it wants to delete an instance | 14:10 |
| *** ildikov has quit IRC | 14:10 | |
| garyk | sdague: mriedem: thanks | 14:11 |
| bauzas | delete_instance is calling RT.update_usage() | 14:11 |
| bauzas | which itself calls RT._update() which calls the sched_cli.update_resource_stats() methof | 14:11 |
| dansmith | bauzas: right | 14:12 |
| bauzas | dansmith: now that we know that, RT is updating the compute node using what's in update_usage | 14:13 |
| sdague | mriedem: responded on - https://review.openstack.org/#/c/167220/ - in the past we've not wanted to leak internal stack traces all the way back to the caller | 14:13 |
| dansmith | bauzas: right, self.stats must have stats['pci_device_pools'] right? | 14:13 |
| bauzas | dansmith: that's horrible, because pci stats have their own module | 14:14 |
| bauzas | dansmith: ie. PCIStats | 14:14 |
| dansmith | bauzas: but I don't see how else we could get here, right? | 14:14 |
| bauzas | dansmith: and RT is hooking up that module | 14:14 |
| anteaya | swamireddy: since the meeting time you selected will conflict with Neutron's meeting time every other week, I highly suggest you take another look at the meeting schedule and select an alternate meeting time | 14:14 |
| anteaya | swamireddy: many meetings alternate on a bi-weekly basis so keep that in mind when you pick a time | 14:15 |
| *** ank2090 has joined #openstack-nova | 14:15 | |
| anteaya | swamireddy: we also have 4 meeting channels to chose from so one of them might have your prefered time open: https://wiki.openstack.org/wiki/Meetings/CreateaMeeting | 14:16 |
| ank2090 | hello all :) | 14:16 |
| *** iamjarvo has joined #openstack-nova | 14:16 | |
| dansmith | bauzas: check this: | 14:16 |
| dansmith | # NOTE(pmurray): the pci_device_pools field maps to the | 14:16 |
| dansmith | # pci_stats field in the database | 14:16 |
| dims | mriedem: still snowed in? | 14:17 |
| ank2090 | can anyone send me a nova-boot command to launch multiple instances? | 14:17 |
| swamireddy | anteaya: Yep...I will go with meeting-3 room | 14:17 |
| mriedem | dims: nope | 14:17 |
| ank2090 | "instance count" like parameter? | 14:17 |
| swamireddy | anteaya: and update the wiki appropriately. | 14:17 |
| dansmith | bauzas: hmm, well, I was wondering if maybe we were getting a raw db object in there that was setting pci_device_pools during the update(), but that would be backwards | 14:17 |
| mriedem | just meeting-'ed in | 14:17 |
| dansmith | bauzas: I wonder if I could reproduce this with a regular system and just create some fake whitelist | 14:18 |
| * dansmith tries | 14:18 | |
| *** atuvenie has quit IRC | 14:18 | |
| mriedem | sdague: 2 comments in the tests | 14:19 |
| *** gszasz has quit IRC | 14:19 | |
| *** gszasz has joined #openstack-nova | 14:20 | |
| anteaya | swamireddy: sounds like you will have better success | 14:21 |
| anteaya | swamireddy: thanks for understanding | 14:21 |
| garyk | ank2090: nova boot —num-instances X …. | 14:22 |
| bauzas | dansmith: okay, just to be clear, RT is managing a compute_node primitive | 14:23 |
| bauzas | so a dict | 14:23 |
| dansmith | bauzas: it's not | 14:23 |
| bauzas | dansmith: it is | 14:23 |
| bauzas | :) | 14:23 |
| bauzas | dansmith: until PaulMurray's patch land | 14:23 |
| garyk | ank2090: here is the full example - nova boot --image cirros-0.3.2-i386-disk --flavor 1 --num-instances 2 X | 14:23 |
| dansmith | bauzas: oh, I missed that it's returning obj_to_primitive(), I see | 14:23 |
| dansmith | bauzas: but still, that won't convert the attribute name | 14:24 |
| *** lpetrut has joined #openstack-nova | 14:24 | |
| bauzas | dansmith: so, somewhere in the RT lifecycle, we're updating self.compute_node['pci_device_pools'] with a dict or a list of something else | 14:24 |
| dansmith | right, okay, that's a wider surface | 14:25 |
| bauzas | dansmith: so when the scheduler will pick each field and instanciate a new object, it won't coerce | 14:25 |
| dansmith | bauzas: I'm spinning up now to see if I can repro this | 14:25 |
| dansmith | if so, it will be much easier to track down | 14:25 |
| dims | bauzas: talking about this? https://bugs.launchpad.net/nova/+bug/1435483 | 14:25 |
| openstack | Launchpad bug 1435483 in OpenStack Compute (nova) " [pci-passthrough] Failed to launch vm after restarting nova compute " [Undecided,Incomplete] | 14:25 |
| bauzas | dansmith: so, the question is : "where in the code, that field is written to self.compute_node ?" | 14:25 |
| dansmith | yeah | 14:25 |
| dansmith | dims: yep | 14:25 |
| *** annashen has joined #openstack-nova | 14:26 | |
| * dims follows along | 14:26 | |
| bauzas | dansmith: oh | 14:26 |
| bauzas | the DB pci_stats field is turned into a pci_device_pool object in compute._from_db_object() | 14:27 |
| *** tonytan4ever has joined #openstack-nova | 14:27 | |
| dansmith | bauzas: ...sure... | 14:28 |
| *** nelsnelson has joined #openstack-nova | 14:28 | |
| swamireddy | anteaya: Thank you. Sorry for confusion...:( | 14:28 |
| dansmith | bauzas: ah | 14:28 |
| anteaya | swamireddy: it happens, you handled it well, thank you | 14:28 |
| bauzas | dansmith: so in order to do this properly, we should probably call _from_db_object() with a dict primitive in the scheduler cli, right ? | 14:28 |
| dansmith | bauzas: so since we're doing to_primitive() on an object, we end up with a dict in that spot | 14:28 |
| bauzas | dansmith: exactly | 14:29 |
| edleafe | bauzas: both compute_node.save() and .create() change back to pci_stats dict, right? | 14:29 |
| dansmith | edleafe: we're talking about the other way | 14:29 |
| bauzas | dansmith: but to_primitive() is not the opposite of _from_db_object(), hence the problem :) | 14:29 |
| dansmith | bauzas: yeah | 14:29 |
| edleafe | bauzas: yeah | 14:29 |
| *** ildikov has joined #openstack-nova | 14:29 | |
| edleafe | bauzas: could that be the disconnect? | 14:30 |
| bauzas | dansmith: because it will serialize it to the same key as the field | 14:30 |
| bauzas | edleafe: yeah | 14:30 |
| *** annashen has quit IRC | 14:30 | |
| dansmith | bauzas: the problem is, we need to handle the dict form of the pci device specially I think, right? i.e. create a PciDevicePool object and call .update() on it with the dict | 14:30 |
| *** browne has joined #openstack-nova | 14:30 | |
| bauzas | I mean, if an object is having a field called 'foo', the dict will be keyed by 'foo' | 14:30 |
| edleafe | dansmith: we call pci_device_pool.from_pci_stats() | 14:31 |
| *** eglynn has joined #openstack-nova | 14:31 | |
| bauzas | dansmith: the problem is how to reconstruct it properly | 14:31 |
| dansmith | edleafe: is that a direct opposite of the generic to-primitive operation? | 14:32 |
| dansmith | bauzas: right | 14:32 |
| *** achanda has joined #openstack-nova | 14:32 | |
| edleafe | dansmith: heh - it depends :) | 14:32 |
| bauzas | dansmith: tbh, the problem seems really old, since Juno | 14:33 |
| bauzas | dansmith: because I wrote that scheduler cli method that cycle | 14:33 |
| dansmith | bauzas: I think it must have come in when we converted the scheduler report.py to use an object | 14:33 |
| dansmith | which was only like a month ago | 14:33 |
| bauzas | dansmith: oh then I missed the patch | 14:33 |
| bauzas | dansmith: lemme blame the code | 14:33 |
| edleafe | bauzas: but the from_pci_stats method of pci_device_pool.py is from last November | 14:34 |
| bauzas | dansmith: because if I would have reviewed the code, I should have pointed out the problem of reversibility for special fields | 14:34 |
| *** ajayaa has quit IRC | 14:34 | |
| bauzas | dansmith: and if not, I totally suck | 14:34 |
| moshele | bauzas: I tested juno before I opened the bug I didn't see pci_device_pools key in self.compute_node | 14:35 |
| *** ajayaa has joined #openstack-nova | 14:35 | |
| bauzas | moshele: right, as said dansmith it's because the scheduler client now uses objects | 14:35 |
| edleafe | moshele: yes, that was added during Kilo | 14:35 |
| bauzas | moshele: previously, it was calling the conductor | 14:35 |
| dansmith | hmm, I can't repro just by having pci_passthrough_whitelist set to something | 14:37 |
| bauzas | dansmith: so, yeah, I think we just need to call the _from_db_object in the scheduler code and make sure that the primitive we pass to the client is consistent with what's in the DB | 14:37 |
| *** ank2090 has quit IRC | 14:37 | |
| dansmith | bauzas: in the client you mean? | 14:37 |
| bauzas | dansmith: my bad, yeah | 14:37 |
| dansmith | bauzas: and you mean on the compute_node or on the pci_device_pools bit? | 14:38 |
| dansmith | moshele: is there any other trick to reproducing this? | 14:38 |
| bauzas | dansmith: sec, trying to find an identic pattern | 14:38 |
| bauzas | identical* | 14:38 |
| *** lpetrut has quit IRC | 14:38 | |
| bauzas | dansmith: because I had to write that reversibility somewhere in the code for another object but I don't recall which one | 14:39 |
| edleafe | bauzas: do we need special handling for the setattr call? | 14:39 |
| bauzas | edleafe: yeah, we really need to avoid it | 14:39 |
| bauzas | edleafe: because keys are not consistent between the NovaObject and the corresponding SQLA object | 14:39 |
| dansmith | so, | 14:39 |
| moshele | dansmith: I don't know, but I have setup with eclipse maybe you can point me where to look | 14:39 |
| dansmith | I think that all we need to do is really just create an object there and do .update() with the dict primitive | 14:40 |
| bauzas | and so, I'm a jerk (seen the review) | 14:40 |
| edleafe | bauzas: I was thinking that instead of L55-56 in the report client, call a method on the compute node object | 14:40 |
| dansmith | because the dict looks like the object, not like the db | 14:40 |
| dansmith | moshele: well, it would be much easier if I could reproduce it myself | 14:41 |
| bauzas | dansmith: yeah, either you're considering the object keys, or the db keys but you can't play with both | 14:41 |
| *** ildikov has quit IRC | 14:41 | |
| *** sdake__ has joined #openstack-nova | 14:41 | |
| bauzas | dansmith: sec, checking again | 14:41 |
| dansmith | bauzas: just a sec, let me post a patch and see if you agree | 14:41 |
| *** mwagner_lap has joined #openstack-nova | 14:42 | |
| bauzas | dansmith: sure, in the meantime, I'm trying to find the best pattern | 14:42 |
| *** achanda has quit IRC | 14:42 | |
| *** mjturek1 has quit IRC | 14:43 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Convert pci_device_pools dict to object before passing to scheduler https://review.openstack.org/167256 | 14:43 |
| dansmith | moshele: could you apply that locally and try it? ^^ | 14:44 |
| *** tonytan4ever has quit IRC | 14:44 | |
| moshele | dansmith: sure | 14:44 |
| *** sdake has quit IRC | 14:44 | |
| *** lpetrut has joined #openstack-nova | 14:45 | |
| edleafe | dansmith: I don't like mixing the internals of compute node into a scheduler rpc client | 14:45 |
| edleafe | dansmith: I would prefer keeping that all in compute_node | 14:46 |
| mriedem | sdague: what's the difference between functional tests for like flavors and the old integration api_samples tests for flavors? | 14:46 |
| dansmith | edleafe: see the note in that code.. it's temporary until RT is passing us a proper compute-node | 14:46 |
| dansmith | edleafe: i.e. the abstraction is already broken, we're just not fixing up the compute node object properly | 14:46 |
| sdague | mriedem: all the old samples tests only go through the positive path | 14:47 |
| openstackgerrit | Benjamin GRASSART proposed openstack/nova: Add vnc_keymap support for vmware compute https://review.openstack.org/167258 | 14:47 |
| *** tsekiyama has joined #openstack-nova | 14:47 | |
| mriedem | sdague: hrm, i assumed we had some netagive testing in there | 14:47 |
| edleafe | dansmith: I understand it is a band-aid | 14:47 |
| bauzas | dansmith: reviewed | 14:47 |
| sdague | mriedem: no, it was explicitly forbidden in the model | 14:48 |
| sdague | at least originally | 14:48 |
| *** mjturek1 has joined #openstack-nova | 14:48 | |
| mriedem | sdague: ok, that helps, thanks | 14:48 |
| bauzas | dansmith: I tend to agree with edleafe, that helper can be provided in the ComputeNode object and left temporary still | 14:48 |
| dansmith | bauzas: ah damn, it is a pool list | 14:48 |
| sdague | honestly, I'm trying to establish some new patterns then figure out what can move over, what can factor out | 14:48 |
| bauzas | dansmith: exactly | 14:48 |
| dansmith | moshele: hang on a sec | 14:48 |
| *** mpaolino has quit IRC | 14:49 | |
| *** browne has quit IRC | 14:49 | |
| *** thedodd has joined #openstack-nova | 14:50 | |
| *** jproulx has joined #openstack-nova | 14:50 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Convert pci_device_pools dict to object before passing to scheduler https://review.openstack.org/167256 | 14:50 |
| dansmith | bauzas: edleafe: we're post-k3, so moving things around for arbitrary cleanup isn't worth the churn, IMHO. This band-aid is already there, it's just wrong | 14:51 |
| bauzas | dansmith: agreed | 14:51 |
| *** EmilienM|PTO is now known as EmilienM | 14:51 | |
| bauzas | post K3 is bad | 14:51 |
| edleafe | dansmith: yeah | 14:51 |
| dansmith | moshele: updated, if you can try *that* :) | 14:51 |
| *** annegentle has joined #openstack-nova | 14:51 | |
| edleafe | dansmith: so just a note saying "don't do that until lemming"? | 14:51 |
| dansmith | edleafe: on my note about it being temporary | 14:52 |
| *** ankit_ag has quit IRC | 14:52 | |
| bauzas | edleafe: well, just a note pointing that it will become useless once an English is able to hit a change | 14:52 |
| dansmith | ? | 14:52 |
| moshele | dansmith: sure I just need 15 min to set again the environment | 14:52 |
| dansmith | moshele: thanks | 14:52 |
| edleafe | bauzas: my Franglish is failing me | 14:52 |
| bauzas | edleafe: lol | 14:53 |
| openstackgerrit | Benjamin GRASSART proposed openstack/nova: Add vnc_keymap support for vmware compute https://review.openstack.org/167258 | 14:53 |
| bauzas | edleafe: I was referring to PaulMurray's patches using objects in the RT | 14:53 |
| dansmith | anyway, let's see if this fixes it for moshele and then we can argue about minutia :) | 14:53 |
| edleafe | bauzas: ah | 14:53 |
| bauzas | dansmith: my daughter has nits in her school... | 14:53 |
| bauzas | dansmith: so that's probably the reason behind that :) | 14:54 |
| *** lpetrut has quit IRC | 14:54 | |
| bauzas | alaski: had you time to figure out how to populate the conf flag to the conductor side ? | 14:55 |
| *** hemnafk is now known as hemna | 14:55 | |
| alaski | bauzas: they're populated on the conductor side | 14:55 |
| bauzas | alaski: oh my bad, I totally missed your last point | 14:55 |
| dansmith | bauzas: I don't know what that means :) | 14:55 |
| bauzas | dansmith: about nits N | 14:55 |
| bauzas | ? | 14:55 |
| dansmith | yeah | 14:55 |
| bauzas | man, I should take note from your Saturday Night Live shows | 14:56 |
| * dansmith is so confused :) | 14:57 | |
| bauzas | dansmith: does 'nit' also refer to that small bugs that you can have in your hair ? | 14:57 |
| *** VW_ has quit IRC | 14:57 | |
| dansmith | ah, lice | 14:57 |
| bauzas | eh | 14:57 |
| dansmith | yes, I think 'nit' is another word, but not the first definition most people think of | 14:57 |
| dansmith | (IMHO) | 14:57 |
| * bauzas amending his Encyclopedia | 14:57 | |
| dansmith | so, you were making a comparison to "picking nits"? I got it now :) | 14:57 |
| bauzas | dansmith: it was an unfortunate attempt to make a pun | 14:58 |
| dansmith | it's all good :) | 14:58 |
| *** VW_ has joined #openstack-nova | 14:59 | |
| bauzas | at least, I know I would have a bad career if I would go to Broadway | 14:59 |
| *** neelashah has quit IRC | 14:59 | |
| openstackgerrit | Ed Leafe proposed openstack/nova: Temporary fix for the pci_device_pools handling. https://review.openstack.org/167260 | 15:00 |
| *** neiljerr` has left #openstack-nova | 15:00 | |
| edleafe | dansmith: bauzas: this was my attempt at a fix ^^ | 15:00 |
| edleafe | I just don't type as fast as dansmith :) | 15:00 |
| bauzas | so back to my discussion with alaski, yeah my bad, forgot your last point about the conductor thing | 15:01 |
| *** raildo|away is now known as raildo | 15:01 | |
| bauzas | so the question is : why oslo.messaging is serializing a ServiceProxy object into a Service NovaObject thing ? | 15:01 |
| alaski | bauzas: np. it looks like fixing this within objects is going to take some work | 15:01 |
| bauzas | alaski: right | 15:01 |
| alaski | bauzas: it's delegating to NovaObjectSerializer which does it | 15:02 |
| bauzas | dansmith: I need your voice here | 15:02 |
| bauzas | dansmith: https://review.openstack.org/#/c/160506/8 | 15:02 |
| edleafe | Gantt meeting starting now in #openstack-meeting | 15:02 |
| bauzas | (if you have time of course while moshele is testing... ) | 15:02 |
| bauzas | dammit ! | 15:02 |
| alaski | bauzas: http://git.openstack.org/cgit/openstack/nova/tree/nova/objects/base.py#n879 | 15:02 |
| alaski | that gets called | 15:03 |
| *** mriedem has quit IRC | 15:03 | |
| alaski | because of the __getitem__/__getattr__ on ServiceProxy | 15:03 |
| dansmith | alaski: so we're trying to serialize an object, which we treat like a dict, | 15:03 |
| *** VW_ has quit IRC | 15:03 | |
| bauzas | alaski: oh | 15:03 |
| dansmith | alaski: and when we get to ['service'] we serialize as a NovaObject, yes? | 15:03 |
| *** mriedem has joined #openstack-nova | 15:04 | |
| *** neelashah has joined #openstack-nova | 15:04 | |
| alaski | dansmith: sort of, because hasattr(entity, 'obj_to_primitive') is true it calls obj_to_primitive | 15:04 |
| bauzas | alaski: so indeed the object path seems hard to fix | 15:04 |
| *** VW_ has joined #openstack-nova | 15:04 | |
| dansmith | well, right | 15:05 |
| *** jaypipes has quit IRC | 15:05 | |
| dansmith | so what's the problem with this? | 15:05 |
| dansmith | it should get to the other side and look like a dict with ['service'] | 15:05 |
| bauzas | dansmith: http://logs.openstack.org/06/160506/8/check/gate-nova-python27/b3bf259/console.html#_2015-03-24_10_58_06_440 | 15:05 |
| alaski | the issue is that ServiceProxy holds the cell routing path info that Service doesn't have | 15:05 |
| alaski | when it gets to the other side it's just a ServiceObject | 15:06 |
| *** achanda has joined #openstack-nova | 15:06 | |
| dansmith | um, I don't understand that | 15:06 |
| *** achanda has quit IRC | 15:06 | |
| openstackgerrit | Claudiu Belu proposed openstack/nova: Adds cleanup on v2.2 keypair api and tests https://review.openstack.org/161857 | 15:06 |
| dansmith | the ServiceProxy doesn't have obj_to_primitive, right? so we should still send the container, right? | 15:07 |
| alaski | dansmith: it's due to how ServiceProxy is defined. it's not really a dict, it proxies getitem/getattr to the object | 15:07 |
| bauzas | dansmith: we're calling a remotable method which passes thru RPC a ServiceProxy object which turns into a Service object | 15:07 |
| dansmith | sorry for being not caught up, but when is this happening? on return from some remotable classmethod? | 15:07 |
| alaski | yes | 15:07 |
| bauzas | dansmith: my bad https://review.openstack.org/#/c/160506/8/nova/objects/service.py,cm | 15:08 |
| alaski | ServiceProxy doesn't have obj_to_primitive, but getattr passes that through to the object, and we lose the wrapper | 15:08 |
| bauzas | alaski: yeah... | 15:09 |
| *** xuhanp_ has quit IRC | 15:09 | |
| *** xuhanp has quit IRC | 15:09 | |
| *** atuvenie has joined #openstack-nova | 15:09 | |
| garyk | mdbooth: you around? | 15:10 |
| dansmith | alaski: ah, getattr not getitem I see | 15:10 |
| alaski | dansmith: bauzas if ServiceProxy was a dict with an object in it, the serializer should work. So maybe ServiceProxy needs to not expose obj_to_primitive | 15:10 |
| bauzas | dansmith: so due to that, the alternative is to make calls to the host api instead in https://review.openstack.org/#/c/160506/8/nova/api/openstack/compute/contrib/hypervisors.py,cm | 15:10 |
| bauzas | alaski: that's doable | 15:11 |
| bauzas | alaski: but that just sounds hacky | 15:11 |
| dansmith | well, | 15:11 |
| dansmith | serviceproxy is hacky :) | 15:11 |
| *** mpaolino has joined #openstack-nova | 15:11 | |
| bauzas | lol | 15:11 |
| alaski | true :) | 15:11 |
| dansmith | if we can just exclude obj_to_primitive() there, that seems like an easy fix for now | 15:11 |
| bauzas | dansmith: ack, let's go playing with this | 15:12 |
| alaski | dansmith: another idea I had was to add a cell_routing field to Service that isn't synced with the db | 15:12 |
| alaski | and avoid the ServiceProxy stuff | 15:12 |
| bauzas | alaski: that would be far cleaner | 15:12 |
| dansmith | that sounds like the right long-term fix | 15:12 |
| dansmith | but | 15:12 |
| dansmith | I think I'd go with the hack for the moment to keep change down | 15:12 |
| dansmith | unless it solves a lot of problems or something | 15:12 |
| bauzas | dansmith: yeah, plus we're post K3 | 15:13 |
| dansmith | right | 15:13 |
| dansmith | a new field is a new object version, new RPC, and potentially upgrade concerns, which I haven't thought through yet | 15:13 |
| bauzas | ok, writing that change now once the gantt meeting is done | 15:13 |
| bauzas | by 'that', meaning the proxy fix | 15:13 |
| dansmith | okay | 15:14 |
| bauzas | dansmith: yeah, version bumps post-K3 sound a bad idea | 15:14 |
| dansmith | not the end of the world, but I'd rather avoid them for sure | 15:14 |
| alaski | yeah. let's try the obj_to_primitive exclusion first | 15:15 |
| *** hdd has joined #openstack-nova | 15:16 | |
| *** ijw has joined #openstack-nova | 15:19 | |
| *** lpetrut has joined #openstack-nova | 15:19 | |
| sahid | dansmith: can i have a first look from you when you have a moment of https://review.openstack.org/#/c/165838/ ? | 15:20 |
| dansmith | sahid: in a sec | 15:20 |
| sahid | i would like to know if that really make sense or not | 15:20 |
| *** wN has joined #openstack-nova | 15:20 | |
| *** ijw has quit IRC | 15:20 | |
| sahid | dansmith: cool did not expected so quick thanks | 15:20 |
| *** Nikolay_St has quit IRC | 15:21 | |
| *** browne has joined #openstack-nova | 15:21 | |
| *** Marga_ has joined #openstack-nova | 15:22 | |
| *** ijw has joined #openstack-nova | 15:22 | |
| moshele | dansmith: it didn't work see http://pastebin.com/FyiMqdx6 | 15:22 |
| dansmith | moshele: it did work actually :) | 15:23 |
| *** ijw has quit IRC | 15:23 | |
| *** ijw has joined #openstack-nova | 15:24 | |
| *** sdake has joined #openstack-nova | 15:25 | |
| *** garyk has quit IRC | 15:25 | |
| *** ijw has quit IRC | 15:25 | |
| moshele | dansmith: I don't fowllow | 15:26 |
| *** wuhao has joined #openstack-nova | 15:26 | |
| *** gszasz has quit IRC | 15:26 | |
| dansmith | moshele: kind of a joke. It got farther than before.. :) | 15:26 |
| *** ijw has joined #openstack-nova | 15:26 | |
| *** dsanders has joined #openstack-nova | 15:27 | |
| *** emagana has joined #openstack-nova | 15:28 | |
| dansmith | moshele: oh wait, was this during the build of the instance? | 15:28 |
| *** sdake__ has quit IRC | 15:29 | |
| moshele | dansmith: yes | 15:29 |
| dansmith | moshele: before, were you trying to delete an instance that booted successfully, or was it during cleanup of a failed boot? | 15:29 |
| *** otter768 has joined #openstack-nova | 15:29 | |
| moshele | dansmith: it was cleanup failed, the error for the build was in the DEUBG log level | 15:30 |
| dansmith | okay | 15:30 |
| dansmith | moshele: is there anything in the n-sch log? | 15:30 |
| *** exploreshaifali has quit IRC | 15:30 | |
| *** bkopilov has quit IRC | 15:31 | |
| *** dsanders has quit IRC | 15:31 | |
| dansmith | moshele: or n-cond | 15:31 |
| *** rushil_ has quit IRC | 15:32 | |
| *** annegentle has quit IRC | 15:33 | |
| dansmith | moshele: also, could you capture a bunch of the lines above this n-cpu traceback so I can try to figure out where we were before we hit this? | 15:33 |
| *** otter768 has quit IRC | 15:33 | |
| moshele | dansmith: sure in a few mintues | 15:35 |
| dansmith | moshele: also, it would be helpful to see your whitelist config so I can try to repro | 15:35 |
| *** gszasz has joined #openstack-nova | 15:36 | |
| *** browne has quit IRC | 15:39 | |
| *** annegentle has joined #openstack-nova | 15:43 | |
| moshele | dansmith: n-cond.log http://pastebin.com/XxT3azpp | 15:44 |
| moshele | dansmith: n-cpu.log http://pastebin.com/1ZnrwJSN | 15:46 |
| *** changbl has joined #openstack-nova | 15:46 | |
| moshele | dansmth: n-sch.log http://pastebin.com/hgJfsWyS | 15:47 |
| *** Mike_D_laptop has quit IRC | 15:47 | |
| moshele | dansmith: pci_passthrough_whitelist = {"address":"*:02:00.*","physical_network":"physnet1"} | 15:47 |
| dansmith | moshele: okay, the conductor log is interesting for sure | 15:48 |
| *** rmoe has quit IRC | 15:48 | |
| *** bkopilov has joined #openstack-nova | 15:49 | |
| *** lpetrut has quit IRC | 15:50 | |
| *** annegentle has quit IRC | 15:50 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Fix how service objects are looked up for Cells https://review.openstack.org/160506 | 15:50 |
| bauzas | dansmith: n-cpu logs are better, because conductor just says that it failed to boot | 15:51 |
| dansmith | bauzas: the conductor log tells us that it's happening on the conductor side though | 15:52 |
| dansmith | i.e. as a result of the save() | 15:52 |
| bauzas | dansmith: well, that's a ValidHost exception coming from a RescheduleException right ? | 15:52 |
| *** lpetrut has joined #openstack-nova | 15:52 | |
| bauzas | *NoValidHost of course... | 15:53 |
| *** wN has left #openstack-nova | 15:53 | |
| dansmith | bauzas: how can you tell? | 15:54 |
| dansmith | I feel like we've lost some rpc debugging recently | 15:54 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/nova: Recover instance state after nova-compute service failure during spawn https://review.openstack.org/167281 | 15:54 |
| *** markus_z has quit IRC | 15:55 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Add DB2 support https://review.openstack.org/69047 | 15:55 |
| bauzas | dansmith: http://pastebin.com/XxT3azpp L192 points to L172 | 15:55 |
| dansmith | bauzas: I think the conductor trace is a result of a call from compute, not that it's logging that compute threw an exception while it was making a call from cond->compute | 15:55 |
| dansmith | bauzas: sure, but that is after the failure | 15:56 |
| bauzas | dansmith: mmm, L172 is the result of the n-cpu exception right | 15:57 |
| dansmith | I don't think so | 15:57 |
| bauzas | dansmith: mmm | 15:57 |
| bauzas | File "/.autodirect/mtrswgwork/moshele/openstack/nova/nova/compute/manager.py", line 2318, in _build_and_run_instance\n   instance_uuid=instance.uuid, reason=six.text_type(e))\n', u'RescheduledException: Build of instance 24974567-637b-4ada-a724-648bd8812e9f was re-scheduled: A string is required here, not dict\nTraceback (most recent call last):\n\n | 15:57 |
| dansmith | look at L163 "returning exception to caller" | 15:57 |
| dansmith | I think what we're seeing is actually api->cond->compute->cond* | 15:58 |
| dansmith | * == failure | 15:58 |
| *** rushiagr is now known as rushiagr_away | 15:58 | |
| *** sdake has quit IRC | 15:58 | |
| bauzas | dansmith: yeah, that's what I'm saying :) | 15:58 |
| *** ildikov has joined #openstack-nova | 15:59 | |
| bauzas | dansmith: L163 is the same exception, but not yet raising a RescheduledException | 15:59 |
| dansmith | okay, but the conductor log is useful because it shows that the conductor is having a hard time reconstructing the compute_node object that we're doing the save on, I think | 15:59 |
| bauzas | what's hard is reading those logs :D | 15:59 |
| bauzas | (kidding) | 15:59 |
| dansmith | more hard is trying to fix this without being able to reproduce :( | 15:59 |
| *** emagana has quit IRC | 16:00 | |
| *** salv-orlando has quit IRC | 16:00 | |
| dansmith | hmm | 16:01 |
| moshele | dansmith: this is the value {u'numa_node': None, u'physical_network': u'physnet1'} that expected to be string | 16:02 |
| dansmith | uh | 16:02 |
| *** EmilienM is now known as EmilienM|PTO | 16:02 | |
| dansmith | I dunno what that is.. bauzas is that PciDevicePool.tags ? | 16:02 |
| *** emagana has joined #openstack-nova | 16:03 | |
| *** ildikov is now known as ildikov_afk | 16:03 | |
| dansmith | moshele: can you tell what field it's trying to set when it does that? | 16:04 |
| bauzas | dansmith: I'm trying to map the coerce line to see the field type | 16:04 |
| *** melwitt has joined #openstack-nova | 16:04 | |
| *** amotoki has quit IRC | 16:04 | |
| *** rmoe has joined #openstack-nova | 16:05 | |
| moshele | dansmith: I am checking | 16:05 |
| *** amotoki has joined #openstack-nova | 16:05 | |
| dansmith | moshele: in nova/objects/base.py, in setter(), you can print/log the name | 16:05 |
| bauzas | dansmith: yeah, that sounds like tags | 16:05 |
| *** Marga_ has quit IRC | 16:05 | |
| *** jistr has quit IRC | 16:05 | |
| dansmith | bauzas: maybe but tags is a dict, so it would expect a dict | 16:05 |
| bauzas | dansmith: tags are differing, based on if you're getting them from the DB or from the object IIRC | 16:06 |
| bauzas | just because I can read, is this a Numa stuff ? | 16:06 |
| *** salv-orlando has joined #openstack-nova | 16:07 | |
| bauzas | bingo | 16:07 |
| bauzas | objects.PCIDevice | 16:07 |
| dansmith | bauzas: what? | 16:08 |
| openstackgerrit | Brent Eagles proposed openstack/nova: WIP: Refactor Nova VIF model https://review.openstack.org/167287 | 16:08 |
| dansmith | although I do see that we need to handle hvspec too | 16:09 |
| moshele | dansmith: pci_device_stats before converting to obj list: [{'count': 7, 'vendor_id': u'15b3', 'product_id': u'1004', 'tags': {u'numa_node': None, u'physical_network': u'physnet1'}}] | 16:09 |
| dansmith | moshele: okay | 16:09 |
| bauzas | tags is a DictOfNullableStringsField | 16:10 |
| bauzas | so maybe it's trying to convert a tag which is not a string, but rather a dict ? | 16:11 |
| bauzas | (I mean a nested dict) | 16:11 |
| dansmith | moshele: can you print the compute_node.pci_device_pools after we create that? | 16:11 |
| dansmith | it should show us what tags looks like in the object | 16:11 |
| *** penick has joined #openstack-nova | 16:11 | |
| moshele | dansmith: sure | 16:12 |
| *** unicell has quit IRC | 16:12 | |
| *** iamjarvo has quit IRC | 16:13 | |
| *** garyk has joined #openstack-nova | 16:13 | |
| *** dulek has quit IRC | 16:15 | |
| openstackgerrit | Ed Leafe proposed openstack/nova: Move ComputeNode creation at init stage in ResourceTracker https://review.openstack.org/148904 | 16:15 |
| *** Nikolay_St has joined #openstack-nova | 16:17 | |
| *** melwitt has quit IRC | 16:17 | |
| *** thedodd has quit IRC | 16:19 | |
| *** melwitt has joined #openstack-nova | 16:20 | |
| *** garyk has quit IRC | 16:21 | |
| *** yamahata has quit IRC | 16:21 | |
| moshele | dansmith:list: [PciDevicePool(count=7,product_id='1004',tags={tags='{u'numa_node': None, u'physical_network': u'physnet1'}'},vendor_id='15b3')] | 16:22 |
| dansmith | ah hah | 16:22 |
| moshele | dansmith I think we have duplicate tags | 16:22 |
| *** sdake has joined #openstack-nova | 16:22 | |
| dansmith | I think it's from_dict() is broken | 16:23 |
| dansmith | well, rather, | 16:23 |
| dansmith | it's not the right thing to use | 16:23 |
| bauzas | *facepalm* | 16:24 |
| bauzas | from_pci_stats | 16:24 |
| *** sdake__ has joined #openstack-nova | 16:24 | |
| bauzas | dansmith: ^ | 16:24 |
| dansmith | bauzas: yeah, but that's still not right | 16:24 |
| openstackgerrit | Dan Smith proposed openstack/nova: Convert pci_device_pools dict to object before passing to scheduler https://review.openstack.org/167256 | 16:24 |
| dansmith | moshele: ^ | 16:24 |
| *** bkopilov has quit IRC | 16:25 | |
| bauzas | dansmith: hence the problem of reversibility | 16:25 |
| dansmith | aye | 16:25 |
| bauzas | that's just horrible | 16:25 |
| bauzas | and as the PCI stuff is not really covered, then we have high chances to break something | 16:25 |
| *** browne has joined #openstack-nova | 16:26 | |
| *** mestery has quit IRC | 16:26 | |
| dansmith | yep | 16:26 |
| *** mestery has joined #openstack-nova | 16:26 | |
| bauzas | dansmith: MHO is that we should prevent the use of obj_to_primitive() | 16:26 |
| dansmith | well, it's a hack | 16:26 |
| dansmith | when we don't need it anymore, it should go away for sure | 16:27 |
| *** sdake has quit IRC | 16:27 | |
| bauzas | dansmith: that's just because it's just taking each field and converting it to something which is *not* a DB primitive | 16:27 |
| dansmith | yep | 16:27 |
| *** lan has quit IRC | 16:28 | |
| moshele | dansmith: it is working :) | 16:28 |
| bauzas | \o/ | 16:28 |
| bauzas | at least for pci :) | 16:28 |
| dansmith | cool | 16:28 |
| dansmith | I will update this as closes-bug | 16:29 |
| openstackgerrit | Dan Smith proposed openstack/nova: Convert pci_device_pools dict to object before passing to scheduler https://review.openstack.org/167256 | 16:29 |
| *** Bsony has quit IRC | 16:30 | |
| bauzas | dansmith: shall we not cover this ? :) | 16:30 |
| moshele | dansmith: do you know in the resource I see pcit_device_pools and pci_stats it seem that they contain the same information? | 16:30 |
| bauzas | moshele: that's 2 ways of presenting the same information | 16:30 |
| dansmith | bauzas: cover this? | 16:30 |
| *** Nikolay_St has quit IRC | 16:30 | |
| bauzas | dansmith: I mean, testing it | 16:31 |
| dansmith | oh | 16:31 |
| dansmith | yeah, I suppose so :) | 16:31 |
| *** Nikolay_St has joined #openstack-nova | 16:32 | |
| *** thedodd has joined #openstack-nova | 16:32 | |
| bauzas | dansmith: I triaged the bug | 16:32 |
| openstackgerrit | Benjamin GRASSART proposed openstack/nova: Add vnc_keymap support for vmware compute https://review.openstack.org/167258 | 16:34 |
| *** amotoki has quit IRC | 16:35 | |
| *** nkrinner has quit IRC | 16:37 | |
| *** annegentle has joined #openstack-nova | 16:37 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Convert pci_device_pools dict to object before passing to scheduler https://review.openstack.org/167256 | 16:37 |
| dansmith | bauzas: covered ^ | 16:38 |
| *** annegent_ has joined #openstack-nova | 16:38 | |
| *** rushil has joined #openstack-nova | 16:39 | |
| *** lan has joined #openstack-nova | 16:39 | |
| *** unicell has joined #openstack-nova | 16:39 | |
| *** gszasz has quit IRC | 16:40 | |
| *** jaypipes_ has joined #openstack-nova | 16:41 | |
| dims | dansmith: bauzas: dumb question _from_db_object seems to get a field 'pci_stats' and turn that into compute.pci_device_pools, is there a possibility that the updates in the patch may have 'pci_stats' too? does it matter? | 16:42 |
| dansmith | no | 16:42 |
| dansmith | it's pci_stats in the db, pci_device_pools in the object | 16:42 |
| *** yamahata has joined #openstack-nova | 16:42 | |
| dansmith | the problem is that we're going from object -> dict -> object here, | 16:42 |
| dansmith | not taking a trip through the db representation | 16:42 |
| *** annegent_ has quit IRC | 16:42 | |
| *** annegentle has quit IRC | 16:43 | |
| dims | dansmith: ah ok. thanks | 16:43 |
| *** armax has quit IRC | 16:44 | |
| *** hdd has quit IRC | 16:44 | |
| *** sdake has joined #openstack-nova | 16:44 | |
| *** romainh has left #openstack-nova | 16:45 | |
| mriedem | dansmith: that's a no-nonsense commit message :) | 16:45 |
| *** ijw has quit IRC | 16:46 | |
| dansmith | more? | 16:46 |
| mriedem | shrug | 16:46 |
| mriedem | i'm just tailing her | 16:46 |
| mriedem | *here | 16:46 |
| mriedem | maybe a note in the code would be helpful | 16:47 |
| *** TobiasE has quit IRC | 16:48 | |
| *** dsanders has joined #openstack-nova | 16:48 | |
| *** sdake__ has quit IRC | 16:49 | |
| *** Nic has joined #openstack-nova | 16:50 | |
| *** matrohon has quit IRC | 16:50 | |
| *** sbfox has joined #openstack-nova | 16:51 | |
| *** sdake__ has joined #openstack-nova | 16:52 | |
| *** exploreshaifali has joined #openstack-nova | 16:54 | |
| *** sdake has quit IRC | 16:56 | |
| *** neelashah has quit IRC | 17:00 | |
| *** baoli has quit IRC | 17:02 | |
| *** emagana has quit IRC | 17:02 | |
| *** abhijeetm has quit IRC | 17:05 | |
| *** mpaolino has quit IRC | 17:05 | |
| *** neelashah has joined #openstack-nova | 17:05 | |
| *** jlanoux_ has quit IRC | 17:07 | |
| *** moshele has quit IRC | 17:08 | |
| *** Marga_ has joined #openstack-nova | 17:08 | |
| *** reed has joined #openstack-nova | 17:10 | |
| *** mikeit has joined #openstack-nova | 17:11 | |
| *** annegentle has joined #openstack-nova | 17:14 | |
| *** wuhao has quit IRC | 17:15 | |
| openstackgerrit | stgleb proposed openstack/python-novaclient: Fix hostname usage for Host resource. https://review.openstack.org/165905 | 17:17 |
| *** harlowja_away is now known as harlowja | 17:18 | |
| *** mpaolino has joined #openstack-nova | 17:18 | |
| *** Nic has quit IRC | 17:20 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Convert pci_device_pools dict to object before passing to scheduler https://review.openstack.org/167256 | 17:20 |
| *** Nic has joined #openstack-nova | 17:21 | |
| *** hdd has joined #openstack-nova | 17:23 | |
| *** baoli has joined #openstack-nova | 17:24 | |
| *** ijw has joined #openstack-nova | 17:26 | |
| *** mjturek1 has quit IRC | 17:26 | |
| *** salv-orlando has quit IRC | 17:26 | |
| *** rushil has quit IRC | 17:28 | |
| *** achanda has joined #openstack-nova | 17:28 | |
| *** annashen has joined #openstack-nova | 17:28 | |
| *** salv-orlando has joined #openstack-nova | 17:28 | |
| *** baoli has quit IRC | 17:28 | |
| *** patrickeast has joined #openstack-nova | 17:29 | |
| *** dboik_ has quit IRC | 17:30 | |
| *** baoli has joined #openstack-nova | 17:30 | |
| *** otter768 has joined #openstack-nova | 17:30 | |
| *** dboik has joined #openstack-nova | 17:30 | |
| openstackgerrit | Sean Dague proposed openstack/nova: don't 500 on invalid security group format https://review.openstack.org/167220 | 17:31 |
| *** sbfox has quit IRC | 17:31 | |
| *** annashen has quit IRC | 17:33 | |
| *** annegentle has quit IRC | 17:34 | |
| *** otter768 has quit IRC | 17:35 | |
| *** Bsony has joined #openstack-nova | 17:36 | |
| jproulx | Hate to ask a usage question in a dev channel, but coming up empty elsewhere | 17:39 |
| jproulx | How do you apply a quota_class to a project? defining them is easy but can't find how to actually use them... | 17:39 |
| mriedem | jproulx: i'm pretty sure quota classes aren't supported / don't work - besides the default quota clsas | 17:41 |
| mriedem | *class | 17:41 |
| *** Bsony has quit IRC | 17:41 | |
| russellb | yep, was never fully implemented | 17:41 |
| jproulx | that would explain my issues :) | 17:41 |
| russellb | sorry that it's confusing | 17:42 |
| mriedem | fwiw we tried removing it :) | 17:42 |
| mriedem | but you know, can't remove apis kind of | 17:42 |
| *** oro has quit IRC | 17:42 | |
| jproulx | sigh, but at least I can stop trying to figure it out and just loop through the projects I want to quota... | 17:42 |
| *** oro_ has quit IRC | 17:42 | |
| *** annegentle has joined #openstack-nova | 17:43 | |
| *** lpetrut has quit IRC | 17:44 | |
| *** dboik_ has joined #openstack-nova | 17:44 | |
| *** lpetrut has joined #openstack-nova | 17:44 | |
| *** derekh has quit IRC | 17:45 | |
| *** mikeit has quit IRC | 17:46 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/nova: Ironic: Remove passing Flavor's deploy_{kernel, ramdisk} https://review.openstack.org/167341 | 17:47 |
| *** dboik has quit IRC | 17:47 | |
| sdague | dansmith: are we emitting deprecation messages on the json pci format? or did that not even get to deprecation? | 17:48 |
| *** tonytan4ever has joined #openstack-nova | 17:48 | |
| *** zz_jgrimm is now known as jgrimm | 17:48 | |
| dansmith | sdague: I don't think we've done any of the things necessary to handle any of the broken stuff, unfortunately | 17:48 |
| dansmith | but I could be wrong | 17:48 |
| sdague | dansmith: ok | 17:49 |
| *** ijw has quit IRC | 17:50 | |
| *** armax has joined #openstack-nova | 17:52 | |
| *** exploreshaifali has quit IRC | 17:53 | |
| *** Mike_D_laptop has joined #openstack-nova | 17:54 | |
| claudiub | hello. if anyone has a little bit of time to take a look at the v2.2 cleanup, please do. :) https://review.openstack.org/#/c/161857/ | 17:55 |
| sahid | eglynn: the hotfix is done http://brewweb.devel.redhat.com/brew/taskinfo?taskID=8899182 | 17:56 |
| eglynn | sahid: nice :) | 17:56 |
| sdague | dims: - https://bugs.launchpad.net/nova/+bug/1433142 you have comments in there, is that a confirmed state? | 17:56 |
| openstack | Launchpad bug 1433142 in oslo-incubator "ProcessLauncher should support reloading config file for parent process on receiving SIGHUP" [Undecided,In progress] - Assigned to Elena Ezhova (eezhova) | 17:56 |
| sahid | i need to go, i reconnect myself tonight to see if all is OK | 17:58 |
| sahid | see you | 17:58 |
| * sahid afk | 17:58 | |
| *** jamie_h has joined #openstack-nova | 17:58 | |
| *** baoli has quit IRC | 17:59 | |
| *** ijw has joined #openstack-nova | 18:00 | |
| *** baoli_ has joined #openstack-nova | 18:00 | |
| *** emagana has joined #openstack-nova | 18:03 | |
| dims | sdague: one more needed to merge - https://review.openstack.org/#/c/165104/ in oslo-inc | 18:04 |
| *** jcoufal has quit IRC | 18:04 | |
| *** jamielennox|away is now known as jamielennox | 18:04 | |
| *** lucasagomes is now known as lucas-dinner | 18:04 | |
| sdague | dims: ok | 18:05 |
| *** mpaolino has quit IRC | 18:05 | |
| sdague | can you categorize the severity on that bug? | 18:05 |
| sdague | trying to get things out of in progress | 18:06 |
| sdague | sorry, trying to get things out of Undecided,New | 18:06 |
| dims | thanks sdague | 18:06 |
| sdague | dims: you going to own the oslo sync once that is done? | 18:07 |
| dims | sdague: yep | 18:07 |
| sdague | ok, assigning you to the nova bug | 18:07 |
| dims | sdague: fyi, i am running tracy's nova-bugs util on another server if you want to use - http://198.11.213.110/nova_bugs/ | 18:07 |
| *** emagana has quit IRC | 18:08 | |
| sdague | ok, I was just running the new bugs list to get that down to 0 again | 18:08 |
| *** exploreshaifali has joined #openstack-nova | 18:09 | |
| *** mpavone has quit IRC | 18:12 | |
| *** PaulMurray is now known as ptm_away | 18:13 | |
| *** baoli_ has quit IRC | 18:15 | |
| *** annegentle has quit IRC | 18:16 | |
| openstackgerrit | Josh Gachnang proposed openstack/nova: Adjust resource tracker for new Ironic states https://review.openstack.org/164313 | 18:20 |
| *** reed has quit IRC | 18:24 | |
| *** Marga_ has quit IRC | 18:24 | |
| *** sbfox has joined #openstack-nova | 18:24 | |
| *** ZZelle_ has joined #openstack-nova | 18:26 | |
| *** yassine has quit IRC | 18:26 | |
| *** achanda has quit IRC | 18:28 | |
| *** emagana has joined #openstack-nova | 18:29 | |
| *** annegentle has joined #openstack-nova | 18:30 | |
| *** apuimedo has quit IRC | 18:31 | |
| *** lpetrut has quit IRC | 18:32 | |
| *** Maike has joined #openstack-nova | 18:33 | |
| *** emagana has quit IRC | 18:33 | |
| *** neelashah has quit IRC | 18:33 | |
| *** VW_ has quit IRC | 18:35 | |
| *** reed has joined #openstack-nova | 18:35 | |
| *** eglynn has quit IRC | 18:37 | |
| *** oro has joined #openstack-nova | 18:37 | |
| *** oro_ has joined #openstack-nova | 18:37 | |
| *** Mike_D_laptop has quit IRC | 18:41 | |
| *** annegentle has quit IRC | 18:45 | |
| *** ritesh has joined #openstack-nova | 18:47 | |
| *** ritesh has quit IRC | 18:49 | |
| openstackgerrit | Josh Gachnang proposed openstack/nova: Add support for cleaning in Ironic driver https://review.openstack.org/161474 | 18:49 |
| *** hdd has quit IRC | 18:55 | |
| *** achanda has joined #openstack-nova | 18:57 | |
| *** VW_ has joined #openstack-nova | 18:59 | |
| *** achanda has quit IRC | 19:00 | |
| *** achanda has joined #openstack-nova | 19:00 | |
| dansmith | sdague: so on this: https://review.openstack.org/#/c/131321/ | 19:01 |
| dansmith | sdague: it sounds like there is more that needs to be done, but that this *is* incremental improvement | 19:01 |
| *** jistr has joined #openstack-nova | 19:02 | |
| dansmith | sdague: given how little progress has been made after so long, I feel like putting this in to at least avoid the crash on start is probably a good idea | 19:02 |
| dansmith | if either of the main pci guys were around, I would ask them about next steps and timing, but alas... | 19:02 |
| *** ijw has quit IRC | 19:08 | |
| *** jcoufal has joined #openstack-nova | 19:11 | |
| *** ajayaa has quit IRC | 19:12 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Convert pci_device_pools dict to object before passing to scheduler https://review.openstack.org/167256 | 19:12 |
| dansmith | sdague: mriedem: had to handle an =None case ^^ | 19:12 |
| mriedem | looking | 19:13 |
| dansmith | not really sure why we failed something in grenade for that and not anywhere else | 19:14 |
| *** johnthetubaguy is now known as zz_johnthetubagu | 19:16 | |
| *** eharney has quit IRC | 19:20 | |
| *** Marga_ has joined #openstack-nova | 19:21 | |
| mriedem | sdague: melwitt: remember during the novaclient 2.21 debacle with volumes where some discussion happened around super ugly code in novaclient and how it deals with volumes? i think it's talking directly to cinder API or something? | 19:24 |
| mriedem | is something happening to clean that up? | 19:25 |
| mriedem | i lost track | 19:25 |
| *** edmondsw has joined #openstack-nova | 19:25 | |
| melwitt | mriedem: yeah, the issue was that the cli creates a client object with service_type='volume' which will use the volume endpoint from the service catalog, whereas the api relied on the caller to pass the same in order to call cinder, else it would call the nova proxy with a wrong path. recently we merged a patch that makes the api use the volume endpoint too, so that the cli and api do the same thing | 19:27 |
| andreykurilin | claudiub: hi! | 19:28 |
| claudiub | andreykurilin: hello | 19:29 |
| *** tellesnobrega has quit IRC | 19:29 | |
| andreykurilin | claudiub: can I use(modify and send new patch-set) your patch https://review.openstack.org/#/c/136458/ ? | 19:29 |
| *** emagana has joined #openstack-nova | 19:30 | |
| *** dboik_ has quit IRC | 19:30 | |
| claudiub | andreykurilin: sure, go ahead. :) but I'm curios about why. :) | 19:30 |
| *** dboik has joined #openstack-nova | 19:31 | |
| *** otter768 has joined #openstack-nova | 19:31 | |
| *** jistr has quit IRC | 19:31 | |
| andreykurilin | claudiub: so I'm working on new patch-set of https://review.openstack.org/#/c/152569 and another patch which add additional decorators(add version info to functions). Your patch looks like the best one for testing it:) | 19:32 |
| *** doude has quit IRC | 19:32 | |
| *** tellesnobrega has joined #openstack-nova | 19:33 | |
| *** emagana has quit IRC | 19:34 | |
| *** dboik_ has joined #openstack-nova | 19:35 | |
| claudiub | andreykurilin: i see. Hm, I still have to send my patchset up, to fix the current pep8 and tests. I will do that now, if that's ok with you. :) | 19:35 |
| *** otter768 has quit IRC | 19:36 | |
| andreykurilin | claudiub: great, thanks) I'll use your patch tomorrow. Today is too late at my timezone:) | 19:37 |
| claudiub | andreykurilin: cool, no problem. :) sending it in 2 mins | 19:37 |
| *** emagana has joined #openstack-nova | 19:37 | |
| mriedem | melwitt: that would be this change right? e0f5072907a00d48a183dd8fc91a6cf6038ca279 | 19:38 |
| edmondsw | melwitt, volume or volumev2? And can you point me to that commit? | 19:38 |
| mriedem | oops https://review.openstack.org/#/c/164321/ | 19:38 |
| *** dboik has quit IRC | 19:38 | |
| mriedem | edmondsw: ^ | 19:38 |
| openstackgerrit | Claudiu Belu proposed openstack/python-novaclient: Adds support for x509 certificates as keypairs https://review.openstack.org/136458 | 19:39 |
| *** morganfainberg is now known as needslesscoffee | 19:39 | |
| *** achanda has quit IRC | 19:42 | |
| melwitt | edmondsw, mriedem: volume, to match the cli. I'm guessing this broke someone using service_type='volumev2' :( if so, it has to be changed to only use that only if it's not already set to 'volume' or 'volumev2' | 19:42 |
| mriedem | melwitt: i think volumev2 is already broken | 19:42 |
| mriedem | https://bugs.launchpad.net/python-novaclient/+bug/1392846 | 19:43 |
| openstack | Launchpad bug 1392846 in python-novaclient "nova volume-list does not work when using cinder v2 API" [High,Confirmed] - Assigned to yatin (yatinkarel) | 19:43 |
| edmondsw | melwitt mriedem, volumev2 is an abomination... they need to get rid of it... version has nothing to do with service type | 19:43 |
| edmondsw | was hoping that was underway here | 19:43 |
| *** exploreshaifali has quit IRC | 19:44 | |
| melwitt | edmondsw, mriedem: okay. I use the word "service_type" because that's how novaclient looks up the things in service catalog and the cinder v2 endpoint is named 'volumev2'. I think. | 19:44 |
| edmondsw | melwitt, yeah, what I mean is that should be just "volume", not "volumev2", in the service catalog | 19:45 |
| mriedem | needslesscoffee: ^ did anything happen with changing that? | 19:46 |
| melwitt | edmondsw: oh, okay | 19:46 |
| needslesscoffee | mriedem, hmm? | 19:46 |
| needslesscoffee | mriedem, reading backlog | 19:47 |
| *** andrewbogott has joined #openstack-nova | 19:48 | |
| needslesscoffee | oh god yeah don't encode versions in the type!! :( | 19:48 |
| *** VW_ has quit IRC | 19:49 | |
| *** Mike_D_laptop has joined #openstack-nova | 19:49 | |
| needslesscoffee | and service type is the right approach in the SC afaik | 19:49 |
| mriedem | needslesscoffee: btw, it's not casual nick friday :) | 19:49 |
| kashyap | mriedem: When you get a moment, since you're involved in block layer stuff in Nova, | 19:49 |
| needslesscoffee | we should summon a thingee and disucss this with him. | 19:49 |
| mriedem | -dev? | 19:50 |
| needslesscoffee | mriedem, i know it isn't... i just need less coffee :( | 19:50 |
| kashyap | Have you seen this bug lately -- https://bugs.launchpad.net/nova/+bug/1353939 | 19:50 |
| openstack | Launchpad bug 1353939 in OpenStack Compute (nova) "Rescue fails with 'Failed to terminate process: Device or resource busy' in the n-cpu log" [High,Confirmed] | 19:50 |
| needslesscoffee | mriedem, -dev or -cinder | 19:50 |
| *** needslesscoffee is now known as morganfainberg | 19:50 | |
| mriedem | kashyap: that's a known gate race yeah | 19:50 |
| kashyap | mriedem: afazekas gave me a reproducer this afternoon, and I'm trying to chase down the QEMU issue. | 19:50 |
| mriedem | edmondsw: moving to #openstack-dev | 19:50 |
| *** doude has joined #openstack-nova | 19:50 | |
| kashyap | mriedem: Oh, Gate race only? | 19:52 |
| kashyap | mriedem: If you take OpenStack out of the equation, you can't reproduce it? | 19:52 |
| kashyap | Current error I see is -- "main-loop: WARNING: I/O thread spun for 1000 iterations" | 19:52 |
| kashyap | When you invoke: $ tox -eall ServerRescueNegativeTestJSON | 19:53 |
| *** eharney has joined #openstack-nova | 19:53 | |
| mriedem | kashyap: i haven't tried to reproduce it outside of openstack, or in generally honestly | 19:53 |
| mriedem | *general | 19:53 |
| kashyap | mriedem: Not that specific bug (it's a 'related' one to what afazekas pointed to me on -qa this evening). | 19:54 |
| kashyap | Anyway, I'll see if I'll muster up enough energy tongiht to find _some_ cause of the I/O error above. | 19:54 |
| *** mpaolino has joined #openstack-nova | 19:54 | |
| mriedem | fwiw i always lumped that bug in with the other gate bugs related to lvm intermittently getting locked up for 2 minutes | 19:54 |
| mriedem | causing timeouts | 19:54 |
| *** doude has quit IRC | 19:54 | |
| mriedem | kashyap: https://bugs.launchpad.net/nova/+bug/1373513 | 19:55 |
| openstack | Launchpad bug 1373513 in Cinder "Lvm hang during tempest tests" [Critical,Fix committed] - Assigned to John Griffith (john-griffith) | 19:55 |
| kashyap | mriedem: Hmm, thanks for this context. I'll just try to reproduce this w/o Tempest. | 19:56 |
| andrewbogott | Is the tenant_id exposed to an instance in any way? If not, would there be support for that to be added to the instance metadata? | 19:57 |
| *** jcoufal has quit IRC | 19:58 | |
| *** flwang1 has joined #openstack-nova | 20:00 | |
| flwang1 | mriedem: ping | 20:01 |
| mriedem | melwitt: isn't this wrong now? https://github.com/openstack/python-novaclient/blob/master/novaclient/shell.py#L68 | 20:03 |
| mriedem | given https://github.com/openstack-dev/devstack/blob/master/lib/nova#L368 | 20:04 |
| mriedem | shouldn't that be 21 / computev21? | 20:04 |
| *** matrohon has joined #openstack-nova | 20:04 | |
| openstackgerrit | Josh Gachnang proposed openstack/nova-specs: API: Proxy neutron configuration to guest instance https://review.openstack.org/167385 | 20:04 |
| mriedem | flwang: pong | 20:04 |
| *** emagana has quit IRC | 20:04 | |
| *** agentle has joined #openstack-nova | 20:06 | |
| *** baoli has joined #openstack-nova | 20:06 | |
| *** neelashah has joined #openstack-nova | 20:06 | |
| melwitt | mriedem: yes. I have seen the fix as part of a much bigger patch that's adding microversion capability on the client https://review.openstack.org/#/c/152569/20/novaclient/shell.py,cm | 20:07 |
| *** doude has joined #openstack-nova | 20:07 | |
| *** baoli has quit IRC | 20:07 | |
| *** baoli has joined #openstack-nova | 20:08 | |
| flwang1 | mriedem: could you pls revisit https://review.openstack.org/164494 ? | 20:09 |
| flwang1 | mriedem: i'm not sure if i fully understand your comments as for bumping the rpc version | 20:09 |
| mriedem | flwang: so we aren't going to be backporting the rpc api changes | 20:09 |
| flwang1 | mriedem: so did you mean I should split current patch into 2? | 20:10 |
| mriedem | flwang: the tactical, backportable fix is detect if the instance is volume-backed in the backup api and raise an exception | 20:10 |
| mriedem | flwang: yeah | 20:10 |
| flwang1 | i mean put the rpc version bumping to another patch? | 20:10 |
| mriedem | yes | 20:10 |
| mriedem | fail if they try to backup a volume-backed api today since it's not supported | 20:10 |
| *** promulo has joined #openstack-nova | 20:10 | |
| mriedem | then do another patch that adds the support with the rpc api version change | 20:10 |
| mriedem | which is really more of a feature | 20:11 |
| flwang1 | mriedem: TBH, i'm not really understand "how" to fail if they try to backup a volume-backed api today since it's not supported | 20:12 |
| *** Mike_D_laptop has quit IRC | 20:12 | |
| mriedem | flwang: in https://review.openstack.org/#/c/164494/9/nova/compute/api.py - if if self.is_volume_backed_instance(context, instance, bdms): is True, raise an exception | 20:12 |
| mriedem | the backup api extension can handle the exception and return a 400 | 20:13 |
| *** xyang1 has joined #openstack-nova | 20:13 | |
| flwang1 | does that mean even though we backport this to juno and icehouse, the two releases still can't support the volume-backed backup? | 20:13 |
| mriedem | flwang: correct | 20:13 |
| mriedem | it makes the failure more obvious | 20:13 |
| flwang1 | nope!!=-O | 20:13 |
| mriedem | since we fail fast | 20:13 |
| flwang1 | so we will let the user can't use this function until Kilo? | 20:14 |
| mriedem | assuming the rpc change to allow it makes it into kilo at this point | 20:14 |
| *** apuimedo has joined #openstack-nova | 20:15 | |
| mriedem | but yes, kilo would be the earliest possible | 20:15 |
| mriedem | since we wouldn't be backporting the rpc api changes | 20:15 |
| mriedem | especially not stable/icehouse, since it's not a security fix | 20:15 |
| flwang1 | mriedem: so is there any way we can get the function in juno and icehouse? | 20:16 |
| mriedem | flwang: you can certainly carry it out of tree at your own risk :) | 20:16 |
| flwang1 | mriedem: lol, so you mean keep it in private repo? | 20:16 |
| mriedem | yeah | 20:17 |
| *** ijw has joined #openstack-nova | 20:18 | |
| flwang1 | mriedem: ah, I see. thank you so much for your clarification. | 20:20 |
| *** doude has quit IRC | 20:20 | |
| *** eglynn has joined #openstack-nova | 20:21 | |
| flwang1 | mriedem: so the first patch will raise an exception if it's volume backed, and the current patch could be the 2nd patch, is it? | 20:23 |
| mriedem | flwang: yeah | 20:24 |
| flwang1 | mriedem: cool, thanks for the confirmation | 20:24 |
| flwang1 | eglynn: ping | 20:24 |
| eglynn | flwang1: hello | 20:24 |
| flwang1 | eglynn: i'm working on a nova backup bug and I saw some code your wrote long time ago, see https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/servers.py#L1075-L1081 | 20:25 |
| flwang1 | eglynn: i opened a bug https://bugs.launchpad.net/nova/+bug/1435596 to refactor | 20:26 |
| openstack | Launchpad bug 1435596 in OpenStack Compute (nova) "[RFE] Refactor the image-create code" [Low,In progress] - Assigned to Fei Long Wang (flwang) | 20:26 |
| flwang1 | IMHO, any volume backed instance doesn't have the image ref | 20:26 |
| * eglynn looking | 20:27 | |
| *** jamie_h has quit IRC | 20:28 | |
| flwang1 | but when you fixed bug https://bugs.launchpad.net/nova/+bug/1034730 , the patch gets the image ref from instance and to do the image create | 20:29 |
| openstack | Launchpad bug 1034730 in OpenStack Compute (nova) "qemu-img receives 'None' as source image to snapshot" [Medium,Fix released] - Assigned to Eoghan Glynn (eglynn) | 20:29 |
| flwang1 | so am i missing anything? thanks | 20:29 |
| *** ijw_ has joined #openstack-nova | 20:30 | |
| *** baoli has quit IRC | 20:31 | |
| *** mwagner_lap has quit IRC | 20:32 | |
| *** baoli has joined #openstack-nova | 20:32 | |
| *** dkliban is now known as dkliban_afk | 20:32 | |
| *** mpaolino has quit IRC | 20:32 | |
| *** ijw has quit IRC | 20:33 | |
| *** vladikr has quit IRC | 20:34 | |
| *** jcru has quit IRC | 20:35 | |
| *** rushil has joined #openstack-nova | 20:35 | |
| *** achanda has joined #openstack-nova | 20:35 | |
| *** andrewbogott is now known as andrewbogott_afk | 20:37 | |
| eglynn | flwang1: here's my original change https://github.com/openstack/nova/commit/c3476b5ca7ab5237d3cb8a84fcb7a3292237b764 | 20:37 |
| mriedem | yeah https://review.openstack.org/#/c/12583/ | 20:38 |
| flwang1 | eglynn: yep, i saw that | 20:38 |
| eglynn | flwang1: and here's the subsequent change that I believe you're confused about https://github.com/openstack/nova/commit/4493c279 | 20:38 |
| flwang1 | it's changed by someone else | 20:38 |
| mriedem | seems we should have made the libvirt driver smarter in the case that source.file isn't in the xml? | 20:38 |
| flwang1 | but what's i don't understand is why it tried to get the image ref from a volume backed instance | 20:38 |
| eglynn | flwang1: yes, exactly | 20:39 |
| mriedem | flwang: b/c the libvirt driver was blindly passing source.file to qemu assuming it was populated | 20:39 |
| mriedem | which isn't not when booting from a volume | 20:39 |
| eglynn | flwang1: see https://bugs.launchpad.net/nova/+bug/1191069 | 20:39 |
| openstack | Launchpad bug 1191069 in OpenStack Compute (nova) "image-create fails on a boot from volume when no image ref is specified" [High,Fix released] - Assigned to Vincent Hou (houshengbo) | 20:39 |
| mriedem | seems the libvirt driver should have tried source.dev if source.file wasn't there | 20:39 |
| *** sahid has quit IRC | 20:40 | |
| mriedem | i'll admit i don't know what qemu will or will not allow here | 20:41 |
| flwang1 | eglynn: yep, i saw that either. so do we still need that logic? since i think for now there is no volume-backed instance with image-ref | 20:41 |
| eglynn | flwang1: yeah the confusion here around the fact the ability to boot from volume *without* an image ref was added subsequent to the original patch you reference above | 20:42 |
| eglynn | flwang1: ... and at that point the image-ref became optional | 20:42 |
| flwang1 | eglynn: 'optional'? | 20:42 |
| *** reed has quit IRC | 20:42 | |
| flwang1 | eglynn: does that mean there is a scenario that a volume-backed instance has an image ref? | 20:43 |
| *** baoli has quit IRC | 20:43 | |
| eglynn | flwang1: TBH I don't remember off-hand the exact sequence of changes from several years ago ... at one point image-ref was needed for boot-for-volume, later it wasn't | 20:43 |
| eglynn | boot *from volume | 20:44 |
| *** baoli has joined #openstack-nova | 20:44 | |
| *** neelashah has quit IRC | 20:47 | |
| *** nellysmitt has joined #openstack-nova | 20:47 | |
| *** nellysmitt has quit IRC | 20:47 | |
| eglynn | flwang1: this change to devstack demonstrates how boot-from-volume changed https://review.openstack.org/#/c/14579/3/exercises/boot_from_volume.sh | 20:50 |
| *** baoli has quit IRC | 20:50 | |
| *** __TheDodd__ has joined #openstack-nova | 20:50 | |
| flwang1 | eglynn: ah, i can see your point, I will give it a try | 20:51 |
| flwang1 | eglynn: thank you so much for the explanation | 20:51 |
| eglynn | flwang1: np! | 20:51 |
| *** baoli has joined #openstack-nova | 20:52 | |
| *** baoli has quit IRC | 20:53 | |
| *** thedodd has quit IRC | 20:53 | |
| *** baoli has joined #openstack-nova | 20:54 | |
| *** Maike has quit IRC | 20:56 | |
| *** andrewbogott_afk is now known as andrewbogott | 20:58 | |
| *** apuimedo has quit IRC | 20:59 | |
| *** melwitt has quit IRC | 21:05 | |
| *** melwitt_ has joined #openstack-nova | 21:05 | |
| *** melwitt_ has quit IRC | 21:05 | |
| *** emagana has joined #openstack-nova | 21:05 | |
| *** melwitt has joined #openstack-nova | 21:05 | |
| *** openstackgerrit has quit IRC | 21:07 | |
| *** openstackgerrit has joined #openstack-nova | 21:07 | |
| *** apuimedo has joined #openstack-nova | 21:07 | |
| *** aysyd has quit IRC | 21:08 | |
| *** doude has joined #openstack-nova | 21:09 | |
| *** emagana has quit IRC | 21:10 | |
| *** VW_ has joined #openstack-nova | 21:12 | |
| *** gokrokve has joined #openstack-nova | 21:13 | |
| *** jcru has joined #openstack-nova | 21:13 | |
| *** jcru has quit IRC | 21:13 | |
| *** neelashah has joined #openstack-nova | 21:20 | |
| *** reed has joined #openstack-nova | 21:20 | |
| mriedem | is it bad that we don't have any boot from volume api samples tests? | 21:20 |
| mriedem | we do have a unit test | 21:20 |
| *** ndipanov has quit IRC | 21:22 | |
| melwitt | mriedem: I think we don't have a tempest test either (being skipped right now) test_volume_boot_pattern | 21:27 |
| *** lennyb has quit IRC | 21:28 | |
| mriedem | ah that's right, bug 1373513 | 21:28 |
| openstack | bug 1373513 in Cinder "Lvm hang during tempest tests" [Critical,Fix committed] https://launchpad.net/bugs/1373513 - Assigned to John Griffith (john-griffith) | 21:28 |
| *** pixelbeat has quit IRC | 21:28 | |
| mriedem | api sample test would fake out the volume api though | 21:29 |
| mriedem | we've had an out of tree test since grizzly but it's been busted since havana :P so i'm debating how much i care about trying to fix and upstream the damn thing | 21:29 |
| *** emagana has joined #openstack-nova | 21:29 | |
| melwitt | mriedem: yeah, sorry. was just thinking on the general subject of testing boot from volume | 21:31 |
| *** andreykurilin_ has joined #openstack-nova | 21:31 | |
| *** annashen has joined #openstack-nova | 21:32 | |
| *** otter768 has joined #openstack-nova | 21:32 | |
| *** harlowja is now known as harlowja_away | 21:32 | |
| mriedem | maybe i'll just open a bug and punt :) | 21:33 |
| mriedem | i'm satisfied with that level of laziness | 21:33 |
| melwitt | :) | 21:33 |
| *** emagana has quit IRC | 21:33 | |
| *** andreykurilin__ has joined #openstack-nova | 21:36 | |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: Change processing versions https://review.openstack.org/152569 | 21:36 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: WIP: filter functions by supported versions https://review.openstack.org/167408 | 21:36 |
| *** IanGovett1 has joined #openstack-nova | 21:36 | |
| *** andreykurilin_ has quit IRC | 21:36 | |
| *** claudiub has quit IRC | 21:36 | |
| *** annashen has quit IRC | 21:36 | |
| *** otter768 has quit IRC | 21:37 | |
| *** dprince has quit IRC | 21:38 | |
| *** IanGovett has quit IRC | 21:39 | |
| *** thangp has quit IRC | 21:39 | |
| *** edmondsw has quit IRC | 21:39 | |
| flwang1 | mriedem: still around? | 21:45 |
| mriedem | flwang1: leaving in like 5 minutes | 21:46 |
| flwang1 | quick question | 21:46 |
| flwang1 | so what's the exception you prefer for failure? | 21:46 |
| *** VW_ has quit IRC | 21:46 | |
| flwang1 | InvalidRequest? ServiceUnavailable? | 21:46 |
| mriedem | the nova exception? | 21:46 |
| flwang1 | yep | 21:46 |
| mriedem | it'd be a 400 i think | 21:46 |
| mriedem | not sure what the compute api should be raising, probably just invalid with a custom _() message | 21:47 |
| flwang1 | yep, I agree with 400 | 21:47 |
| mriedem | although we're past string freeze... | 21:47 |
| mriedem | i guess look through the existing nova exceptions and see if something fits | 21:47 |
| flwang1 | ok | 21:47 |
| flwang1 | I will submit a patch soon | 21:48 |
| *** agentle has quit IRC | 21:48 | |
| *** mdbooth has quit IRC | 21:48 | |
| *** tlbr has quit IRC | 21:48 | |
| *** wolsen has quit IRC | 21:48 | |
| *** flip214 has quit IRC | 21:48 | |
| *** tristanC has quit IRC | 21:48 | |
| *** agentle has joined #openstack-nova | 21:49 | |
| *** mdbooth has joined #openstack-nova | 21:49 | |
| *** tlbr has joined #openstack-nova | 21:49 | |
| *** wolsen has joined #openstack-nova | 21:49 | |
| *** flip214 has joined #openstack-nova | 21:49 | |
| *** tristanC has joined #openstack-nova | 21:49 | |
| *** agentle has quit IRC | 21:50 | |
| *** kaufer has quit IRC | 21:53 | |
| *** mriedem is now known as mriedem_away | 21:53 | |
| *** emagana has joined #openstack-nova | 21:53 | |
| *** ijw has joined #openstack-nova | 21:54 | |
| *** baoli has quit IRC | 21:56 | |
| *** nelsnelson has quit IRC | 21:56 | |
| *** baoli has joined #openstack-nova | 21:57 | |
| *** ijw_ has quit IRC | 21:57 | |
| *** baoli has quit IRC | 21:58 | |
| *** Marga_ has quit IRC | 21:59 | |
| *** baoli has joined #openstack-nova | 21:59 | |
| *** neelashah has quit IRC | 22:00 | |
| *** jcook has quit IRC | 22:01 | |
| *** s1rp has quit IRC | 22:02 | |
| *** s1rp has joined #openstack-nova | 22:03 | |
| *** lucas-dinner has quit IRC | 22:04 | |
| *** achanda_ has joined #openstack-nova | 22:04 | |
| *** harlowja_away is now known as harlowja | 22:05 | |
| *** VW_ has joined #openstack-nova | 22:06 | |
| *** oro has quit IRC | 22:06 | |
| *** oro_ has quit IRC | 22:07 | |
| *** oro has joined #openstack-nova | 22:07 | |
| *** andrewbogott is now known as andrewbogott_afk | 22:07 | |
| *** achanda has quit IRC | 22:08 | |
| *** baoli has quit IRC | 22:08 | |
| *** sbfox1 has joined #openstack-nova | 22:09 | |
| *** baoli has joined #openstack-nova | 22:10 | |
| *** sbfox has quit IRC | 22:10 | |
| *** VW_ has quit IRC | 22:11 | |
| *** ociuhandu has quit IRC | 22:13 | |
| *** burt has quit IRC | 22:13 | |
| *** changbl has quit IRC | 22:14 | |
| *** VW_ has joined #openstack-nova | 22:14 | |
| *** jcook has joined #openstack-nova | 22:14 | |
| *** Marga_ has joined #openstack-nova | 22:18 | |
| *** andreykurilin__ has quit IRC | 22:23 | |
| *** andreykurilin_ has joined #openstack-nova | 22:23 | |
| *** jcoufal has joined #openstack-nova | 22:23 | |
| *** jecarey has quit IRC | 22:25 | |
| *** VW_ has quit IRC | 22:26 | |
| *** baoli has quit IRC | 22:27 | |
| *** baoli has joined #openstack-nova | 22:27 | |
| *** emagana has quit IRC | 22:29 | |
| *** Marga_ has quit IRC | 22:30 | |
| *** eglynn has quit IRC | 22:34 | |
| openstackgerrit | Fei Long Wang proposed openstack/nova: Raise exception when backup volume-backed instance https://review.openstack.org/167418 | 22:35 |
| *** __TheDodd__ has quit IRC | 22:35 | |
| *** Marga_ has joined #openstack-nova | 22:36 | |
| openstackgerrit | Fei Long Wang proposed openstack/nova: Raise exception when backup volume-backed instance https://review.openstack.org/167418 | 22:37 |
| *** sbfox1 has quit IRC | 22:42 | |
| *** asettle has joined #openstack-nova | 22:42 | |
| *** sbfox has joined #openstack-nova | 22:44 | |
| *** tonytan4ever has quit IRC | 22:48 | |
| *** sbfox has quit IRC | 22:53 | |
| *** baoli_ has joined #openstack-nova | 22:53 | |
| *** armax has quit IRC | 22:54 | |
| *** baoli has quit IRC | 22:55 | |
| *** baoli_ has quit IRC | 22:56 | |
| *** matrohon has quit IRC | 22:56 | |
| *** thomasem is now known as thomasem-away | 22:58 | |
| *** salv-orl_ has joined #openstack-nova | 22:58 | |
| melwitt | dims: did the rules change for the quick hit bug list? I see a lot of strikethroughs for merged things | 23:02 |
| *** salv-orlando has quit IRC | 23:02 | |
| *** penick has quit IRC | 23:02 | |
| *** markvoelker has quit IRC | 23:06 | |
| *** claudiub has joined #openstack-nova | 23:07 | |
| *** penick has joined #openstack-nova | 23:07 | |
| *** andreykurilin_ has quit IRC | 23:07 | |
| *** andreykurilin__ has joined #openstack-nova | 23:07 | |
| *** sdake__ has quit IRC | 23:10 | |
| *** welldannit has joined #openstack-nova | 23:11 | |
| *** dims has quit IRC | 23:14 | |
| *** dims has joined #openstack-nova | 23:16 | |
| *** dims has quit IRC | 23:17 | |
| *** dims has joined #openstack-nova | 23:17 | |
| dims | melwitt: it could use a cleanup | 23:18 |
| dims | melwitt: if you have some time | 23:19 |
| melwitt | dims: okay, I just wanted to check with you before I delete things and add to the counter. thanks :) | 23:19 |
| *** penick has quit IRC | 23:22 | |
| *** ociuhandu has joined #openstack-nova | 23:25 | |
| *** otter768 has joined #openstack-nova | 23:33 | |
| *** annashen has joined #openstack-nova | 23:34 | |
| *** andreykurilin__ has quit IRC | 23:34 | |
| *** flwang1 has quit IRC | 23:35 | |
| *** asettle has quit IRC | 23:35 | |
| *** ijw_ has joined #openstack-nova | 23:37 | |
| *** otter768 has quit IRC | 23:37 | |
| *** annashen has quit IRC | 23:38 | |
| *** mwagner_lap has joined #openstack-nova | 23:39 | |
| *** ijw has quit IRC | 23:40 | |
| *** markvoelker has joined #openstack-nova | 23:47 | |
| *** oomichi has joined #openstack-nova | 23:49 | |
| *** ZZelle_ has quit IRC | 23:52 | |
| *** penick has joined #openstack-nova | 23:52 | |
| *** markvoelker has quit IRC | 23:52 | |
| *** hblixt has joined #openstack-nova | 23:53 | |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova: Fix v2.1 hypervisor servers to return empty list https://review.openstack.org/167100 | 23:54 |
| *** ildikov_afk has quit IRC | 23:54 | |
| *** gyee has joined #openstack-nova | 23:54 | |
| *** oomichi has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!