| *** lucas-dinner has quit IRC | 00:00 | |
| *** harlowja has quit IRC | 00:02 | |
| *** harlowja has joined #openstack-nova | 00:03 | |
| *** flwang1 has quit IRC | 00:05 | |
| *** ZZelle_ has quit IRC | 00:05 | |
| *** hdd has quit IRC | 00:09 | |
| *** flaper87 has quit IRC | 00:11 | |
| *** asettle has joined #openstack-nova | 00:15 | |
| *** flaper87 has joined #openstack-nova | 00:15 | |
| *** vilobhmm11 has joined #openstack-nova | 00:16 | |
| *** promulo has joined #openstack-nova | 00:17 | |
| *** gokrokve_ has quit IRC | 00:17 | |
| *** salv-orlando has quit IRC | 00:18 | |
| *** emagana has quit IRC | 00:19 | |
| *** sseverson has quit IRC | 00:19 | |
| *** vilobhmm1 has quit IRC | 00:19 | |
| *** annashen has quit IRC | 00:19 | |
| *** hdd has joined #openstack-nova | 00:20 | |
| *** promulo__ has quit IRC | 00:21 | |
| *** iamjarvo has quit IRC | 00:22 | |
| *** yingjun has joined #openstack-nova | 00:23 | |
| *** BadCub is now known as BadCub_Hiding | 00:25 | |
| *** eliqiao has joined #openstack-nova | 00:28 | |
| alex_xu | sdague: mriedem_away I'm ok with revert patch https://review.openstack.org/#/c/169464/ agree with too late to fix oslo. | 00:29 |
|---|---|---|
| alex_xu | mriedem_away: move the v2.1 policy rule to separated just for ease the configuration for operator | 00:29 |
| *** hdd has quit IRC | 00:30 | |
| alex_xu | mriedem_away: and change the prefix to 'os_compute_api' is for we didn't have v3 in the future and the api version also bump by microversion, so include the api version inthe policy rule is confused. And after v2 api removed, the v2.1 is the only openstack api. | 00:31 |
| *** dboik has joined #openstack-nova | 00:31 | |
| *** Sukhdev has joined #openstack-nova | 00:32 | |
| *** vilobhmm11 has quit IRC | 00:32 | |
| *** dboik_ has joined #openstack-nova | 00:33 | |
| *** dboik has quit IRC | 00:36 | |
| *** neelashah has joined #openstack-nova | 00:44 | |
| *** rushil has quit IRC | 00:48 | |
| *** zhangtralon has joined #openstack-nova | 00:50 | |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova: Add parameter_type none for JSON-Schema validation https://review.openstack.org/169573 | 00:52 |
| *** neelashah has quit IRC | 00:57 | |
| *** shaohe_feng has joined #openstack-nova | 01:00 | |
| *** vilobhmm1 has joined #openstack-nova | 01:00 | |
| *** suresh12 has quit IRC | 01:01 | |
| *** baoli has joined #openstack-nova | 01:03 | |
| *** coreywright has quit IRC | 01:04 | |
| *** coreywright has joined #openstack-nova | 01:05 | |
| *** suresh12 has joined #openstack-nova | 01:07 | |
| *** pixelb has quit IRC | 01:11 | |
| *** zhenguo has joined #openstack-nova | 01:12 | |
| *** xyang1 has quit IRC | 01:16 | |
| *** eliqiao has quit IRC | 01:16 | |
| *** devlaps has quit IRC | 01:16 | |
| *** eliqiao has joined #openstack-nova | 01:17 | |
| *** gmann has quit IRC | 01:17 | |
| *** ujjain has quit IRC | 01:17 | |
| *** annegentle has joined #openstack-nova | 01:18 | |
| *** gmann has joined #openstack-nova | 01:18 | |
| *** annashen has joined #openstack-nova | 01:20 | |
| *** baoli has quit IRC | 01:20 | |
| *** artom_ has joined #openstack-nova | 01:21 | |
| *** baoli has joined #openstack-nova | 01:21 | |
| *** artom has quit IRC | 01:21 | |
| *** haigang has joined #openstack-nova | 01:22 | |
| *** slong has quit IRC | 01:24 | |
| *** otter768 has joined #openstack-nova | 01:24 | |
| *** diegows has quit IRC | 01:25 | |
| *** annashen has quit IRC | 01:26 | |
| *** mwagner_lap has quit IRC | 01:27 | |
| *** casusbel_ has joined #openstack-nova | 01:28 | |
| openstackgerrit | Yongli He proposed openstack/nova: resource tracker style pci resource management https://review.openstack.org/131321 | 01:28 |
| *** haomaiwang has quit IRC | 01:29 | |
| *** otter768 has quit IRC | 01:30 | |
| *** kaisers1 has joined #openstack-nova | 01:31 | |
| *** casusbelli has quit IRC | 01:31 | |
| *** kaisers has quit IRC | 01:33 | |
| *** baoli has quit IRC | 01:36 | |
| *** dsanders has quit IRC | 01:36 | |
| *** armax has quit IRC | 01:36 | |
| *** dsanders has joined #openstack-nova | 01:36 | |
| *** dboik_ has quit IRC | 01:38 | |
| *** flwang1 has joined #openstack-nova | 01:38 | |
| *** mwagner_lap has joined #openstack-nova | 01:39 | |
| *** oomichi has joined #openstack-nova | 01:39 | |
| *** dboik has joined #openstack-nova | 01:39 | |
| *** dsanders has quit IRC | 01:41 | |
| *** annegentle has quit IRC | 01:43 | |
| *** salv-orlando has joined #openstack-nova | 01:44 | |
| *** alanf-mc has quit IRC | 01:47 | |
| *** haigang has quit IRC | 01:48 | |
| *** zhangtralon has quit IRC | 01:48 | |
| *** erkules has quit IRC | 01:49 | |
| *** erkules_ has joined #openstack-nova | 01:49 | |
| *** royce has joined #openstack-nova | 01:50 | |
| *** vmtrooper has joined #openstack-nova | 01:52 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge V2 and V2.1 Agent functional tests https://review.openstack.org/167096 | 01:58 |
| *** haigang has joined #openstack-nova | 01:58 | |
| *** iamjarvo has joined #openstack-nova | 02:00 | |
| *** haomaiwang has joined #openstack-nova | 02:00 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge certificates functional tests of v2 and v2.1 https://review.openstack.org/167473 | 02:05 |
| *** mwagner_lap has quit IRC | 02:06 | |
| *** patrickeast has quit IRC | 02:09 | |
| *** vilobhmm1 has quit IRC | 02:10 | |
| *** mtanino has quit IRC | 02:10 | |
| *** royce has quit IRC | 02:11 | |
| eliqiao | hi dims_: | 02:12 |
| *** Sukhdev has quit IRC | 02:14 | |
| *** popw has joined #openstack-nova | 02:14 | |
| *** rfolco has joined #openstack-nova | 02:15 | |
| *** yamahata has quit IRC | 02:15 | |
| *** baoli has joined #openstack-nova | 02:16 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge V2 and V2.1 cloudpipe functional tests https://review.openstack.org/167481 | 02:16 |
| *** asettle has quit IRC | 02:18 | |
| *** unicell has quit IRC | 02:18 | |
| *** annegentle has joined #openstack-nova | 02:19 | |
| *** mwagner_lap has joined #openstack-nova | 02:19 | |
| *** shaohe_feng has quit IRC | 02:21 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Share V2 and V2.1 hosts functional tests https://review.openstack.org/167520 | 02:21 |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Share V2 and V2.1 images functional tests https://review.openstack.org/167854 | 02:22 |
| *** annegentle has quit IRC | 02:24 | |
| *** vmtrooper has quit IRC | 02:25 | |
| dims_ | eliqiao: thanks for the quick reviews, we are talking about it now on #openstack-keystone if you want to hop in there | 02:28 |
| *** harlowja is now known as harlowja_away | 02:29 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merging instance_actions tests between V2 and V2.1 https://review.openstack.org/167860 | 02:30 |
| *** yingjun has quit IRC | 02:30 | |
| *** jecarey has joined #openstack-nova | 02:31 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge instance_usage_audit_log tests between V2 and V2.1 https://review.openstack.org/167864 | 02:31 |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Share migrations tests between V2 and V2.1 https://review.openstack.org/167866 | 02:32 |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge sec grp default rules tests between V2 and V2.1 https://review.openstack.org/167874 | 02:32 |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Share V2 and V2.1 tenant-networks functional tests https://review.openstack.org/168259 | 02:32 |
| *** suresh12 has quit IRC | 02:33 | |
| *** yingjun has joined #openstack-nova | 02:35 | |
| *** zhangtralon has joined #openstack-nova | 02:36 | |
| *** suresh12 has joined #openstack-nova | 02:38 | |
| dims_ | eliqiao: you should port the change to nova and get it running and check if it works as designed (look at the logs). though you should keep in as WIP until the one in oslo.policy merges | 02:42 |
| *** jecarey has quit IRC | 02:54 | |
| *** dims_ has quit IRC | 02:55 | |
| *** annegentle has joined #openstack-nova | 02:57 | |
| *** achanda has quit IRC | 02:59 | |
| *** salv-orlando has quit IRC | 02:59 | |
| *** Longgeek_ has joined #openstack-nova | 03:03 | |
| *** Longgeek has quit IRC | 03:05 | |
| openstackgerrit | rajiv proposed openstack/python-novaclient: Combine test cases for checking nova limits response https://review.openstack.org/167876 | 03:06 |
| *** hdd has joined #openstack-nova | 03:09 | |
| *** david-lyle has joined #openstack-nova | 03:13 | |
| *** vilobhmm1 has joined #openstack-nova | 03:16 | |
| openstackgerrit | Lingxian Kong proposed openstack/nova: Raise exception when using invalid az for vm creation https://review.openstack.org/163842 | 03:17 |
| *** asettle has joined #openstack-nova | 03:18 | |
| *** hdd has quit IRC | 03:21 | |
| *** suresh12 has quit IRC | 03:23 | |
| *** annashen has joined #openstack-nova | 03:23 | |
| *** dsanders has joined #openstack-nova | 03:23 | |
| openstackgerrit | Rui Chen proposed openstack/nova: Fix scheduler issue when multiple-create failed https://review.openstack.org/147048 | 03:24 |
| *** otter768 has joined #openstack-nova | 03:25 | |
| *** vilobhmm1 has quit IRC | 03:26 | |
| *** annashen has quit IRC | 03:27 | |
| *** armax has joined #openstack-nova | 03:29 | |
| *** otter768 has quit IRC | 03:30 | |
| *** Nikolay_St has joined #openstack-nova | 03:36 | |
| *** changbl_ has joined #openstack-nova | 03:36 | |
| *** changbl has quit IRC | 03:38 | |
| *** dsanders has quit IRC | 03:44 | |
| *** Marga_ has quit IRC | 03:44 | |
| *** Marga_ has joined #openstack-nova | 03:46 | |
| *** dboik has quit IRC | 03:47 | |
| openstackgerrit | zhangtralon proposed openstack/nova: add support to get migrate object in muti statuses https://review.openstack.org/169604 | 03:48 |
| openstackgerrit | Sudipta Biswas proposed openstack/nova: Avoids Error code 500 on negative port quota https://review.openstack.org/169605 | 03:52 |
| *** rushiagr_away is now known as rushiagr | 03:54 | |
| *** baoli has quit IRC | 03:55 | |
| *** salv-orlando has joined #openstack-nova | 03:58 | |
| *** Sukhdev has joined #openstack-nova | 04:00 | |
| *** vmtrooper has joined #openstack-nova | 04:00 | |
| *** yingjun has quit IRC | 04:00 | |
| *** vmtrooper has quit IRC | 04:01 | |
| *** flwang1 has quit IRC | 04:01 | |
| *** vmtrooper has joined #openstack-nova | 04:01 | |
| *** iamjarvo has quit IRC | 04:02 | |
| *** rajesht has joined #openstack-nova | 04:03 | |
| *** emagana has joined #openstack-nova | 04:07 | |
| *** Nic1 has quit IRC | 04:12 | |
| *** achanda has joined #openstack-nova | 04:12 | |
| *** lpetrut has joined #openstack-nova | 04:13 | |
| *** dsanders has joined #openstack-nova | 04:14 | |
| openstackgerrit | zhangtralon proposed openstack/nova: add support to get migrate object in muti statuses https://review.openstack.org/169604 | 04:16 |
| *** rushiagr is now known as rushiagr_away | 04:22 | |
| *** annegentle has quit IRC | 04:26 | |
| *** signed8bit_ZZZzz is now known as signed8b_ | 04:27 | |
| *** salv-orlando has quit IRC | 04:32 | |
| *** yingjun has joined #openstack-nova | 04:33 | |
| *** yingjun has quit IRC | 04:35 | |
| *** yingjun has joined #openstack-nova | 04:36 | |
| *** haigang has quit IRC | 04:38 | |
| *** Marga_ has quit IRC | 04:42 | |
| *** dave-mccowan has quit IRC | 04:42 | |
| *** haigang has joined #openstack-nova | 04:44 | |
| *** eharney has quit IRC | 04:47 | |
| *** garyk has quit IRC | 04:48 | |
| *** krtaylor has quit IRC | 04:48 | |
| *** krtaylor has joined #openstack-nova | 04:49 | |
| *** eharney has joined #openstack-nova | 04:49 | |
| *** haigang has quit IRC | 04:50 | |
| *** vmtrooper has quit IRC | 04:50 | |
| *** zhangtralon has quit IRC | 04:50 | |
| *** hdd has joined #openstack-nova | 04:59 | |
| *** Marga_ has joined #openstack-nova | 05:03 | |
| *** Marga_ has quit IRC | 05:03 | |
| *** Marga_ has joined #openstack-nova | 05:04 | |
| *** emagana has quit IRC | 05:04 | |
| *** emagana has joined #openstack-nova | 05:05 | |
| *** yingjun has quit IRC | 05:06 | |
| *** emagana has quit IRC | 05:10 | |
| *** signed8b_ has quit IRC | 05:10 | |
| *** rushiagr_away is now known as rushiagr | 05:14 | |
| *** rushiagr is now known as rushiagr_away | 05:23 | |
| *** otter768 has joined #openstack-nova | 05:26 | |
| *** lpetrut has quit IRC | 05:29 | |
| *** yamahata has joined #openstack-nova | 05:30 | |
| *** otter768 has quit IRC | 05:31 | |
| *** haigang has joined #openstack-nova | 05:35 | |
| *** haigang has quit IRC | 05:35 | |
| *** Mike_D_laptop has joined #openstack-nova | 05:37 | |
| *** Longgeek_ has quit IRC | 05:38 | |
| *** Longgeek has joined #openstack-nova | 05:39 | |
| *** garyk has joined #openstack-nova | 05:39 | |
| *** vilobhmm1 has joined #openstack-nova | 05:40 | |
| *** yingjun has joined #openstack-nova | 05:41 | |
| *** _gryf has quit IRC | 05:42 | |
| *** haigang has joined #openstack-nova | 05:45 | |
| *** Maike has joined #openstack-nova | 05:48 | |
| *** Maike_ has joined #openstack-nova | 05:50 | |
| openstackgerrit | Eli Qiao proposed openstack/nova: API: Avoid reloading policy files in policy.d for every call https://review.openstack.org/168784 | 05:50 |
| *** Maike has quit IRC | 05:53 | |
| *** unicell has joined #openstack-nova | 05:53 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge baremetal_nodes func tests between V2 and V2.1 https://review.openstack.org/168225 | 05:53 |
| *** asettle has quit IRC | 05:58 | |
| *** sudipto has joined #openstack-nova | 05:59 | |
| *** Sukhdev has quit IRC | 06:00 | |
| sudipto | bauzas: Hi | 06:01 |
| *** hdd has quit IRC | 06:04 | |
| openstackgerrit | heijlong proposed openstack/nova: Merge quota_sets functional tests of v2 and v2.1 https://review.openstack.org/167899 | 06:04 |
| *** park_heijlong has joined #openstack-nova | 06:04 | |
| *** pradipta has joined #openstack-nova | 06:05 | |
| *** zhangtralon has joined #openstack-nova | 06:06 | |
| openstackgerrit | zhangtralon proposed openstack/nova: add support for muti_statuses_query from migration https://review.openstack.org/169631 | 06:06 |
| *** hblixt has quit IRC | 06:07 | |
| *** haigang has quit IRC | 06:07 | |
| *** packet has quit IRC | 06:07 | |
| *** haigang has joined #openstack-nova | 06:09 | |
| *** haigang has quit IRC | 06:10 | |
| openstackgerrit | heijlong proposed openstack/nova: Merge user_data functional tests of v2 and v2.1 https://review.openstack.org/168265 | 06:11 |
| *** bfic has joined #openstack-nova | 06:12 | |
| openstackgerrit | zhangtralon proposed openstack/nova: add support to get migrate object in muti statuses https://review.openstack.org/169604 | 06:12 |
| *** haigang has joined #openstack-nova | 06:13 | |
| openstackgerrit | Rui Chen proposed openstack/nova: Fix failure of stopping instances during init host https://review.openstack.org/167926 | 06:15 |
| *** oro has joined #openstack-nova | 06:17 | |
| *** xylan has joined #openstack-nova | 06:17 | |
| *** KrustyB has joined #openstack-nova | 06:17 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Imported Translations from Transifex https://review.openstack.org/169637 | 06:18 |
| *** salv-orlando has joined #openstack-nova | 06:19 | |
| *** _gryf has joined #openstack-nova | 06:19 | |
| *** afazekas has joined #openstack-nova | 06:21 | |
| openstackgerrit | Rui Chen proposed openstack/nova-specs: CreateVM supports subnet specified https://review.openstack.org/169638 | 06:23 |
| *** haigang has quit IRC | 06:23 | |
| openstackgerrit | Ramaraja proposed openstack/python-novaclient: Report better error message --ephemeral poor usage https://review.openstack.org/169639 | 06:24 |
| *** annashen has joined #openstack-nova | 06:25 | |
| *** haigang has joined #openstack-nova | 06:25 | |
| *** ildikov_ has quit IRC | 06:25 | |
| *** ankit_ag has joined #openstack-nova | 06:25 | |
| openstackgerrit | zhangtralon proposed openstack/nova: add support to get migrate object in muti statuses https://review.openstack.org/169604 | 06:26 |
| openstackgerrit | Ramaraja proposed openstack/python-novaclient: Report better error message --ephemeral poor usage https://review.openstack.org/169639 | 06:27 |
| *** annashen has quit IRC | 06:29 | |
| *** wverdugo500 has joined #openstack-nova | 06:30 | |
| openstackgerrit | zhangtralon proposed openstack/nova: add support to get migrate object in muti statuses https://review.openstack.org/169604 | 06:30 |
| *** zhangtralon has quit IRC | 06:31 | |
| openstackgerrit | zhangtralon proposed openstack/nova: resize-revert can break anti-affinity policy https://review.openstack.org/160110 | 06:33 |
| openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge V2 and V2.1 networks functional tests https://review.openstack.org/168242 | 06:34 |
| *** Mike_D_laptop has quit IRC | 06:37 | |
| openstackgerrit | zhangtralon proposed openstack/nova: add support for muti_statuses_query from migration https://review.openstack.org/169631 | 06:38 |
| *** vmtrooper has joined #openstack-nova | 06:38 | |
| *** wverdugo5001 has joined #openstack-nova | 06:39 | |
| *** ishant has joined #openstack-nova | 06:41 | |
| *** flaper87 has quit IRC | 06:45 | |
| *** flaper87 has joined #openstack-nova | 06:45 | |
| *** dulek has joined #openstack-nova | 06:45 | |
| *** vilobhmm1 has quit IRC | 06:45 | |
| openstackgerrit | zhangtralon proposed openstack/nova: add support to get migrate object in muti statuses https://review.openstack.org/169604 | 06:47 |
| *** TobiasE has joined #openstack-nova | 06:48 | |
| openstackgerrit | Deepak C Shetty proposed openstack/nova: libvirt: Use 'relative' flag for online snapshot's commit/rebase operations https://review.openstack.org/168805 | 06:51 |
| *** moshele has joined #openstack-nova | 06:51 | |
| *** vmtrooper has quit IRC | 06:53 | |
| *** vmtrooper has joined #openstack-nova | 06:53 | |
| *** salv-orlando has quit IRC | 06:54 | |
| openstackgerrit | Eli Qiao proposed openstack/nova: API: Avoid reloading policy files in policy.d for every call https://review.openstack.org/168784 | 06:55 |
| *** fawadkhaliq has joined #openstack-nova | 06:55 | |
| openstackgerrit | zhangtralon proposed openstack/nova: resize-revert can break anti-affinity policy https://review.openstack.org/160110 | 06:55 |
| *** vmtrooper has quit IRC | 06:58 | |
| *** zhangtralon has joined #openstack-nova | 06:59 | |
| *** park_heijlong has quit IRC | 07:00 | |
| *** doude has joined #openstack-nova | 07:01 | |
| *** jcoufal has joined #openstack-nova | 07:01 | |
| *** cfriesen has quit IRC | 07:02 | |
| *** erkules_ is now known as erkules | 07:03 | |
| *** fawadkhaliq has quit IRC | 07:08 | |
| *** markvoelker has quit IRC | 07:09 | |
| *** fawadkhaliq has joined #openstack-nova | 07:10 | |
| *** KrustyB has quit IRC | 07:10 | |
| *** sbfox has joined #openstack-nova | 07:11 | |
| *** ildikov_ has joined #openstack-nova | 07:11 | |
| *** pask81 has joined #openstack-nova | 07:12 | |
| openstackgerrit | Yongli He proposed openstack/nova: resource tracker style pci resource management https://review.openstack.org/131321 | 07:13 |
| *** jistr has joined #openstack-nova | 07:14 | |
| *** park_heijlong has joined #openstack-nova | 07:15 | |
| *** fawadkhaliq is now known as fawadkhaliq_ | 07:15 | |
| *** fawadkhaliq_ is now known as fawadkhaliq | 07:15 | |
| *** KrustyB has joined #openstack-nova | 07:16 | |
| *** fawadkhaliq has quit IRC | 07:16 | |
| *** fawadkhaliq has joined #openstack-nova | 07:17 | |
| *** fawadkhaliq has quit IRC | 07:21 | |
| *** sbfox has quit IRC | 07:22 | |
| *** fawadkhaliq has joined #openstack-nova | 07:22 | |
| *** romainh has joined #openstack-nova | 07:23 | |
| *** casusbel_ has quit IRC | 07:25 | |
| *** casusbelli has joined #openstack-nova | 07:26 | |
| *** dsanders has quit IRC | 07:26 | |
| *** otter768 has joined #openstack-nova | 07:27 | |
| *** dsanders has joined #openstack-nova | 07:28 | |
| *** rushiagr_away is now known as rushiagr | 07:30 | |
| *** jcoufal has quit IRC | 07:32 | |
| *** otter768 has quit IRC | 07:32 | |
| *** mrda is now known as mrda-away | 07:33 | |
| *** matrohon has joined #openstack-nova | 07:34 | |
| *** achanda has quit IRC | 07:36 | |
| *** rushiagr is now known as rushiagr_away | 07:37 | |
| *** claudiub has joined #openstack-nova | 07:39 | |
| openstackgerrit | Ramaraja proposed openstack/python-novaclient: Report better error message --ephemeral poor usage https://review.openstack.org/169639 | 07:39 |
| *** park_heijlong has quit IRC | 07:40 | |
| *** doude has quit IRC | 07:40 | |
| *** doude has joined #openstack-nova | 07:41 | |
| *** achanda has joined #openstack-nova | 07:42 | |
| *** achanda has quit IRC | 07:43 | |
| bauzas | sudipto: hi (please avoid naked pings https://blogs.gnome.org/markmc/2014/02/20/naked-pings/ ) | 07:44 |
| bauzas | sudipto: what do you want to know ? | 07:44 |
| *** markus_z has joined #openstack-nova | 07:44 | |
| *** markus_z has quit IRC | 07:47 | |
| *** dims_ has joined #openstack-nova | 07:50 | |
| *** Marga_ has quit IRC | 07:50 | |
| *** eglynn has joined #openstack-nova | 07:52 | |
| *** mpavone has joined #openstack-nova | 07:53 | |
| *** dims_ has quit IRC | 07:55 | |
| *** jcoufal has joined #openstack-nova | 07:56 | |
| *** sahid has joined #openstack-nova | 08:00 | |
| *** yamahata has quit IRC | 08:02 | |
| *** dsanders has quit IRC | 08:02 | |
| *** salv-orlando has joined #openstack-nova | 08:03 | |
| *** yingjun has quit IRC | 08:06 | |
| openstackgerrit | Romain Hardouin proposed openstack/nova-specs: Refresh quotas usage https://review.openstack.org/161782 | 08:06 |
| *** ndipanov has joined #openstack-nova | 08:07 | |
| *** park_heijlong has joined #openstack-nova | 08:07 | |
| *** Mike_D_laptop has joined #openstack-nova | 08:08 | |
| *** mpaolino has joined #openstack-nova | 08:08 | |
| *** park_heijlong has quit IRC | 08:12 | |
| *** Longgeek has quit IRC | 08:13 | |
| *** park_heijlong has joined #openstack-nova | 08:13 | |
| *** Longgeek has joined #openstack-nova | 08:13 | |
| *** zz_johnthetubagu is now known as johnthetubaguy | 08:15 | |
| *** yingjun has joined #openstack-nova | 08:15 | |
| sudipto | bauzas: Point taken :-) Thanks! I wanted to know if we need to expose some specific metrics from a specific set of h/w (KVM based) - for openstack users. What's the approach one needs to take? | 08:16 |
| bauzas | sudipto: I would tend to say that's the virt driver which is exposing what he wants | 08:17 |
| bauzas | sudipto: and the RT will commit them to the DB | 08:17 |
| bauzas | sudipto: look at the NUMA specific stuff | 08:18 |
| *** lpetrut has joined #openstack-nova | 08:18 | |
| *** park_heijlong has quit IRC | 08:18 | |
| sudipto | bauzas, Ok - but since the virt driver is very libvirt specific and this probably is not a part of the libvirt APIs - would it still be a part of the driver code for libvirt? Can there be exceptions made? | 08:18 |
| bauzas | sudipto: I don't get your point, we have many virt drivers that are not libvirt | 08:19 |
| bauzas | sudipto: and libvirt itself exposes many backends (KVM, Xen) | 08:19 |
| sudipto | bauzas, my point was - that the libvirt driver is essential for everything else. So the driver will be the libvirt KVM driver - but the metrics cannot be gathered via Libvirt. | 08:20 |
| bauzas | sudipto: I now remember our discussion | 08:20 |
| sudipto | bauzas, as in - we need the libvirt driver with the KVM backend as well. | 08:20 |
| bauzas | sudipto: honestly, that's really hard to say what kind of technical direction you should take without having more details | 08:21 |
| sudipto | bauzas, yeah - i actually tested the code out via the libvirt driver - the API is written inside it at the moment but I am not sure if that's the way to go for upstreaming. | 08:21 |
| bauzas | sudipto: you should probably write a spec | 08:21 |
| bauzas | sudipto: because I can also put you on the wrong direction | 08:21 |
| sudipto | bauzas, Ok, let me do write a spec on what we want to do then add you as a reviewer... | 08:22 |
| bauzas | sudipto: and there are many things to consider (are you claiming those resources when a compute is starting ? do those resources needing to be checked when migrating ? etc.) | 08:22 |
| bauzas | sudipto: sure thing | 08:22 |
| sudipto | bauzas, Ok. Just to answer that, I have exposed it via the monitors - like you had suggested the last time. So it's now in the Nova DB as a part of the host states. | 08:23 |
| bauzas | sudipto: okay | 08:23 |
| sudipto | bauzas, It could be used for a scheduler filter or consumed by ceilometer for some additional heuristics etc. | 08:23 |
| *** zhangtralon has quit IRC | 08:23 | |
| bauzas | sudipto: so a new metric plugin ? | 08:24 |
| bauzas | s/metric/monitor | 08:24 |
| sudipto | bauzas, I didn't sort of write a plugin but extended the RT | 08:24 |
| *** zhangtralon has joined #openstack-nova | 08:24 | |
| bauzas | mmm | 08:24 |
| *** RuiChen has quit IRC | 08:24 | |
| *** lucasagomes has joined #openstack-nova | 08:24 | |
| *** RuiChen has joined #openstack-nova | 08:25 | |
| bauzas | as I probably already explained to you, we have some tech debt on the RT related to the extensibility | 08:25 |
| bauzas | so I don't know what you meant by "extended the RT" but the only way atm is to write a monitor AFAIK | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: fixed tests in test_iptables_network to work with random PYTHONHASHSEED https://review.openstack.org/163352 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: refactored tests to pass in test_inject to pass with random PYTHONHASHSEED https://review.openstack.org/163354 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: fix for vmware test_driver_api to pass with random PYTHONHASHSEED https://review.openstack.org/163357 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: fixed tests in test_matchers to pass with random PYTHONHASHSEED https://review.openstack.org/163358 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: fixed tests in test_instance to pass with random PYTHONHASHSEED https://review.openstack.org/161135 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: refactored test in vmware test_read_write_util to pass with random PYTHONHASHSEED https://review.openstack.org/163359 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: Refactored tests in test_neutron_driver to pass with random PYTHONHASHSEED https://review.openstack.org/163361 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: refactored tests in test_objects to pass with random PYTHONHASHSEED https://review.openstack.org/161313 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: Removed 'PYTHONHASHSEED=0' from tox.ini https://review.openstack.org/163877 | 08:26 |
| openstackgerrit | Timofey Durakov proposed openstack/nova: Fixed tests in test_glance to pass with random PYTHONHASHSEED https://review.openstack.org/163364 | 08:26 |
| bauzas | sudipto: because the resources plugins are a dead-end | 08:26 |
| *** zhangtralon has quit IRC | 08:27 | |
| sudipto | bauzas, yeah wrote a monitor by extending the base class. That's what you had recommended. | 08:27 |
| *** RuiChen has quit IRC | 08:27 | |
| bauzas | sudipto: ok, then that's technically a plugin :) | 08:27 |
| sudipto | bauzas, opps! Ok :) | 08:27 |
| bauzas | sudipto: provided you added an entrypoint | 08:27 |
| sudipto | bauzas, yeah have the entry point as well. | 08:27 |
| bauzas | cool | 08:27 |
| sudipto | bauzas, do you reckon it's the right way for now? I will write the spec. | 08:28 |
| bauzas | well, as I said, I probably need more details - and I guess others will want too :) | 08:28 |
| sudipto | bauzas, yeah - will add all the details. Thanks. | 08:28 |
| bauzas | so a Liberty spec is a good thing IMHO | 08:28 |
| sudipto | bauzas, ok | 08:29 |
| sudipto | bauzas, also if i could ask you for a very minor review here: https://review.openstack.org/#/c/169605/ | 08:29 |
| bauzas | sudipto: I can only +1 but sure I can do | 08:29 |
| sudipto | bauzas, that's still good :) | 08:29 |
| *** RuiChen has joined #openstack-nova | 08:31 | |
| *** RuiChen has quit IRC | 08:34 | |
| *** RuiChen has joined #openstack-nova | 08:34 | |
| *** park_heijlong has joined #openstack-nova | 08:34 | |
| *** RuiChen has quit IRC | 08:35 | |
| *** RuiChen has joined #openstack-nova | 08:35 | |
| openstackgerrit | Alex Xu proposed openstack/nova: Add json-schema for v2.1 fixed-ips https://review.openstack.org/169139 | 08:36 |
| bauzas | sudipto: I'm all good with this one, and I will shed the light on it | 08:38 |
| *** garyk1 has joined #openstack-nova | 08:40 | |
| *** garyk has quit IRC | 08:41 | |
| *** atuvenie has joined #openstack-nova | 08:41 | |
| *** vmtrooper has joined #openstack-nova | 08:42 | |
| bauzas | mikal: still up ? | 08:44 |
| bauzas | mikal: I was reviewing the trivial bug list | 08:44 |
| *** garyk1 has quit IRC | 08:44 | |
| bauzas | mikal: and I wonder if some good enough are worth being targeted for RC1 | 08:44 |
| bauzas | mikal: it would increase our merge momentum | 08:44 |
| *** garyk has joined #openstack-nova | 08:45 | |
| *** park_heijlong has quit IRC | 08:45 | |
| openstackgerrit | Flavio Percoco proposed openstack/nova: Refactor image's code to support both versions https://review.openstack.org/144875 | 08:45 |
| *** garyk has quit IRC | 08:45 | |
| bauzas | mikal: hell, 7.45pm your time, just checked, my bad | 08:45 |
| *** garyk has joined #openstack-nova | 08:46 | |
| bauzas | Austrilia is still DST... | 08:46 |
| *** vmtrooper has quit IRC | 08:47 | |
| *** claudiub has quit IRC | 08:48 | |
| sudipto | bauzas, Thanks!! | 08:49 |
| openstackgerrit | Rui Chen proposed openstack/nova: Fix scheduler issue when multiple-create failed https://review.openstack.org/147048 | 08:49 |
| *** garyk has quit IRC | 08:50 | |
| openstackgerrit | Rui Chen proposed openstack/nova: Fix scheduler issue when multiple-create failed https://review.openstack.org/147048 | 08:53 |
| *** sudipto has quit IRC | 08:54 | |
| openstackgerrit | heijlong proposed openstack/nova: Merge quota_sets functional tests of v2 and v2.1 https://review.openstack.org/167899 | 08:56 |
| *** oro has quit IRC | 08:56 | |
| *** sudipto has joined #openstack-nova | 08:57 | |
| *** park_heijlong has joined #openstack-nova | 08:57 | |
| *** Longgeek has quit IRC | 08:59 | |
| *** sudipto has quit IRC | 08:59 | |
| *** andreykurilin_ has joined #openstack-nova | 08:59 | |
| *** sudipto has joined #openstack-nova | 09:00 | |
| *** rushiagr_away is now known as rushiagr | 09:01 | |
| *** park_heijlong has quit IRC | 09:08 | |
| *** garyk has joined #openstack-nova | 09:11 | |
| *** markus_z has joined #openstack-nova | 09:15 | |
| *** sdake has joined #openstack-nova | 09:18 | |
| *** salv-orl_ has joined #openstack-nova | 09:19 | |
| openstackgerrit | Alexey I. Froloff proposed openstack/nova: Store user-requested ephemeral size in instance object https://review.openstack.org/163393 | 09:19 |
| *** sdake_ has quit IRC | 09:22 | |
| *** sdake has quit IRC | 09:22 | |
| *** salv-orlando has quit IRC | 09:22 | |
| *** Longgeek has joined #openstack-nova | 09:27 | |
| *** links has joined #openstack-nova | 09:27 | |
| *** annashen has joined #openstack-nova | 09:27 | |
| *** otter768 has joined #openstack-nova | 09:28 | |
| *** ociuhandu has quit IRC | 09:28 | |
| *** yingjun has quit IRC | 09:32 | |
| *** annashen has quit IRC | 09:33 | |
| *** otter768 has quit IRC | 09:33 | |
| *** pixelb has joined #openstack-nova | 09:36 | |
| openstackgerrit | Romain Hardouin proposed openstack/nova-specs: Refresh quotas usage https://review.openstack.org/161782 | 09:38 |
| *** KrustyB has quit IRC | 09:38 | |
| *** lpetrut has quit IRC | 09:39 | |
| *** bfic has quit IRC | 09:39 | |
| *** claudiub has joined #openstack-nova | 09:47 | |
| *** haomaiwang has quit IRC | 09:51 | |
| *** dims_ has joined #openstack-nova | 10:00 | |
| *** dims__ has joined #openstack-nova | 10:01 | |
| *** alexpilotti has joined #openstack-nova | 10:03 | |
| *** mpavone has quit IRC | 10:03 | |
| *** claudiub has quit IRC | 10:03 | |
| *** dims_ has quit IRC | 10:04 | |
| *** oro has joined #openstack-nova | 10:07 | |
| *** dbite has joined #openstack-nova | 10:14 | |
| *** andreykurilin_ has quit IRC | 10:14 | |
| *** pixelb has quit IRC | 10:16 | |
| *** obondarev has joined #openstack-nova | 10:19 | |
| *** rushiagr is now known as rushiagr_away | 10:19 | |
| *** lpetrut has joined #openstack-nova | 10:21 | |
| *** rushiagr_away is now known as rushiagr | 10:22 | |
| *** jistr has quit IRC | 10:23 | |
| *** haigang has quit IRC | 10:24 | |
| *** pixelb has joined #openstack-nova | 10:27 | |
| kashyap | mdbooth: When you get a sec, in your live migration tests do you boot a Nova instance with "--block-device-mapping"? | 10:30 |
| *** vmtrooper has joined #openstack-nova | 10:31 | |
| *** sudipto has quit IRC | 10:32 | |
| openstackgerrit | Alex Xu proposed openstack/nova: Add json-schema for v2.1 fixed-ips https://review.openstack.org/169139 | 10:34 |
| *** claudiub has joined #openstack-nova | 10:35 | |
| *** vmtrooper has quit IRC | 10:36 | |
| *** jistr has joined #openstack-nova | 10:36 | |
| *** IanGovett has joined #openstack-nova | 10:41 | |
| *** matrohon has quit IRC | 10:42 | |
| *** achanda has joined #openstack-nova | 10:44 | |
| *** ociuhandu has joined #openstack-nova | 10:44 | |
| *** ishant has quit IRC | 10:44 | |
| *** achanda has quit IRC | 10:48 | |
| *** mpaolino has quit IRC | 10:49 | |
| *** nellysmitt has joined #openstack-nova | 10:50 | |
| openstackgerrit | zhangtralon proposed openstack/nova: creating ordinary numa vm fails with large_page_vm https://review.openstack.org/167917 | 10:51 |
| *** derekh has joined #openstack-nova | 10:52 | |
| *** haigang has joined #openstack-nova | 10:54 | |
| *** oomichi has quit IRC | 10:58 | |
| *** haigang has quit IRC | 11:10 | |
| *** jistr is now known as jistr|demo | 11:11 | |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: Deprecate v1.1 and remove v3 https://review.openstack.org/169378 | 11:18 |
| *** dbite has quit IRC | 11:20 | |
| *** mwagner_lap has quit IRC | 11:20 | |
| *** dguitarbite has joined #openstack-nova | 11:20 | |
| *** eliqiao has quit IRC | 11:21 | |
| *** mpavone has joined #openstack-nova | 11:21 | |
| *** claudiub has quit IRC | 11:26 | |
| *** rushiagr is now known as rushiagr_away | 11:28 | |
| *** rushiagr_away is now known as rushiagr | 11:28 | |
| *** otter768 has joined #openstack-nova | 11:29 | |
| *** annashen has joined #openstack-nova | 11:29 | |
| *** otter768 has quit IRC | 11:34 | |
| *** annashen has quit IRC | 11:34 | |
| *** dulek has quit IRC | 11:40 | |
| *** pradipta has quit IRC | 11:45 | |
| *** popw has quit IRC | 11:47 | |
| *** dprince has joined #openstack-nova | 11:51 | |
| *** lpetrut has quit IRC | 11:53 | |
| *** gszasz has joined #openstack-nova | 11:57 | |
| *** eliqiao has joined #openstack-nova | 12:00 | |
| *** eliqiao has quit IRC | 12:00 | |
| *** matrohon has joined #openstack-nova | 12:01 | |
| *** lucasagomes is now known as lucas-hungry | 12:07 | |
| *** mwagner_lap has joined #openstack-nova | 12:09 | |
| *** mpaolino has joined #openstack-nova | 12:09 | |
| *** chinmay_ has joined #openstack-nova | 12:12 | |
| *** Mike_D_laptop has quit IRC | 12:14 | |
| *** markvoelker has joined #openstack-nova | 12:14 | |
| *** busterswt has joined #openstack-nova | 12:16 | |
| raorn | sahid: ayt? | 12:17 |
| *** vladikr has joined #openstack-nova | 12:18 | |
| *** matrohon has quit IRC | 12:19 | |
| *** vmtrooper has joined #openstack-nova | 12:20 | |
| sdague | johnthetubaguy: https://review.openstack.org/#/c/169464/ has a +1 from alex, that prevents the policy reread regression | 12:20 |
| sdague | jaypipes: you're up as well ^^^ | 12:20 |
| sdague | dansmith: once you get up - https://review.openstack.org/#/c/149746/ has a stack above it which seems like it could all go in, but you should review the bottom | 12:20 |
| rajesht | hi cores, | 12:23 |
| rajesht | could you please review https://review.openstack.org/#/c/135202/ | 12:23 |
| rajesht | thanks a lot | 12:23 |
| *** Mike_D_laptop has joined #openstack-nova | 12:23 | |
| *** smccully_ has joined #openstack-nova | 12:23 | |
| *** smccully has joined #openstack-nova | 12:23 | |
| *** smccully has quit IRC | 12:24 | |
| *** vmtrooper has quit IRC | 12:25 | |
| *** smccully_ has quit IRC | 12:25 | |
| garyk | sdague: any idea what the microsoft help is doing on gerrit? | 12:25 |
| *** dave-mccowan has joined #openstack-nova | 12:25 | |
| sdague | being helpful? | 12:27 |
| *** smccully has joined #openstack-nova | 12:27 | |
| garyk | sdague: no, it is just an eyesore | 12:27 |
| garyk | do you not see it on gerrit (right hand side on the bottom) | 12:28 |
| openstackgerrit | Mikhail Feoktistov proposed openstack/nova: Add config drive support for PCS containers https://review.openstack.org/149253 | 12:29 |
| *** links has quit IRC | 12:30 | |
| *** salv-orl_ has quit IRC | 12:32 | |
| *** claudiub has joined #openstack-nova | 12:34 | |
| lxsli | garyk: ts1 lirpA | 12:36 |
| garyk | lxsli: thanks. turns out that i am the fool | 12:37 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: Deprecate v1.1 and remove v3 https://review.openstack.org/169378 | 12:37 |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: Adds os-compute-api-type flag https://review.openstack.org/169726 | 12:37 |
| alex_xu | sdague: dims__ and eliqiao works on this https://review.openstack.org/#/c/169535/, looks like pretty close, should we consider that? | 12:39 |
| *** dulek has joined #openstack-nova | 12:39 | |
| sdague | alex_xu: honestly, it still adds a bunch of extra fs calls | 12:39 |
| *** aysyd has joined #openstack-nova | 12:40 | |
| sdague | and that's in oslo.policy, which we don't use | 12:40 |
| sdague | so for kilo I think we need to revert this | 12:40 |
| sdague | and I think we need a real conversation about whether this is useful at all | 12:40 |
| sdague | because best practice seems that you shouldn't use it, as it's going to add a ton of extra file system calls | 12:41 |
| alex_xu | sdague: ok, got it. I'm ok with revert, then we can find out better fix later | 12:41 |
| sdague | yep | 12:42 |
| *** matrohon has joined #openstack-nova | 12:42 | |
| *** hdd has joined #openstack-nova | 12:43 | |
| alex_xu | sdague: so I can stop this for now also https://review.openstack.org/#/c/166811/ | 12:43 |
| sdague | alex_xu: I think we should still land that because we're probably bringing this back in some form next cycle | 12:45 |
| sdague | and we're going to have to land that patch anyway | 12:45 |
| alex_xu | sdague: at that time, we probably already switch to oslo.policy, and the option for policy dir already removed | 12:46 |
| *** diegows has joined #openstack-nova | 12:46 | |
| *** mpaolino has quit IRC | 12:46 | |
| alex_xu | sdague: then need another fix for that | 12:46 |
| sdague | alex_xu: true | 12:47 |
| *** pixelb has quit IRC | 12:51 | |
| *** diegows has quit IRC | 12:51 | |
| *** otter768 has joined #openstack-nova | 12:52 | |
| *** dboik has joined #openstack-nova | 12:52 | |
| bauzas | jaypipes: morning, by discussing with jogo, he would appreciate your thoughts on https://review.openstack.org/#/c/169365/ | 12:53 |
| bauzas | jaypipes: I know you're not gantt-core but a vote is worth it :) | 12:53 |
| *** salv-orlando has joined #openstack-nova | 12:53 | |
| *** leong has joined #openstack-nova | 12:54 | |
| *** dboik_ has joined #openstack-nova | 12:55 | |
| leong | is it possible to configure nova so that all the vm instance name can have a default naming convention? eg. vmxxx-yyy00001 | 12:55 |
| *** baoli has joined #openstack-nova | 12:56 | |
| *** raildo|away is now known as raildo | 12:57 | |
| *** zul has quit IRC | 12:57 | |
| *** zul has joined #openstack-nova | 12:57 | |
| bauzas | leong: instance_name_template is what you need | 12:57 |
| *** dboik has quit IRC | 12:57 | |
| *** neelashah has joined #openstack-nova | 12:59 | |
| *** baoli has quit IRC | 12:59 | |
| *** baoli has joined #openstack-nova | 13:00 | |
| leong | bauzas: should that be set at controller or compute node? If i have two Availability Zone, can the instance name be generated as e.g. vm-az1-yyy0001 and vm-az2-yyy0002 | 13:00 |
| jaypipes | bauzas: done | 13:01 |
| *** mpaolino has joined #openstack-nova | 13:02 | |
| *** suresh12 has joined #openstack-nova | 13:02 | |
| bauzas | leong: mmm, excellent question, it was originally done at the DB level so it was only set at the controller | 13:02 |
| bauzas | leong: but IIRC this is now done at the object level | 13:02 |
| jaypipes | sdague: done | 13:02 |
| *** hdd has quit IRC | 13:02 | |
| sdague | jaypipes: thanks! | 13:03 |
| jaypipes | sdague: 33197 times. wow. | 13:03 |
| bauzas | leong: by thinking about it, that should be still at the controller level | 13:03 |
| sdague | man, looking at year old incomplete bug in nova bug tracker ... so much cruft in there | 13:03 |
| bauzas | sdague: I was pinging mikal this morning but had no reply, so bugging you about trivial bugs | 13:04 |
| sdague | jaypipes: yeh, I think the policy.d support was not well thought out and didn't consider the caching issue | 13:04 |
| bauzas | jaypipes: thanks | 13:04 |
| sdague | bauzas: about what? | 13:04 |
| bauzas | sdague: about the possibility to tag those as RC1 | 13:04 |
| *** adalbas has joined #openstack-nova | 13:04 | |
| bauzas | in order to have a good view for them | 13:04 |
| *** annegentle has joined #openstack-nova | 13:05 | |
| bauzas | I made a few passes on the list, and some are waiting some core support, so I was thinking that RC1 could give them a better visibility than an etherpad | 13:05 |
| sdague | bauzas: no, please do not | 13:05 |
| sdague | rc1 is supposed to be about must fix | 13:06 |
| bauzas | sdague: okay, it was only a question | 13:06 |
| bauzas | :) | 13:06 |
| bauzas | I see, untargeted bugs are acceptable to be merged for RC1, but targeted bugs are only those we consider as must have | 13:06 |
| sahid | hum.. it's only me or there is something like a "pin" in the UI of review.openstack.org | 13:07 |
| johnthetubaguy | +1 stuff on the rc1 list blocks the release of rc1 | 13:07 |
| sahid | raorn: i have replied on the review | 13:07 |
| bauzas | sahid: did you say fish ? | 13:07 |
| *** xuhanp has joined #openstack-nova | 13:07 | |
| johnthetubaguy | mr clippy is back… the end of the world is near | 13:08 |
| bauzas | johnthetubaguy: okay, then my related question is when we are planning to triage those as per mikal's email :) | 13:08 |
| sahid | bauzas: no it's not a fish | 13:08 |
| *** lucas-hungry is now known as lucasagomes | 13:08 | |
| bauzas | sahid: eh, clippy is a good fish | 13:08 |
| bauzas | :) | 13:09 |
| bauzas | as smart as a dolphin | 13:09 |
| sahid | johnthetubaguy: ah yes it's a mr clippy ! | 13:09 |
| *** leong has quit IRC | 13:09 | |
| sahid | but did you see it? | 13:09 |
| dims__ | sdague: so, even after https://review.openstack.org/#/c/169464/ merges, the support for policy.d is still there in the code and if a deployer wants to use it, they will still see the problem for reloads. | 13:09 |
| *** ferest has joined #openstack-nova | 13:09 | |
| sahid | i mean do you see it? | 13:09 |
| johnthetubaguy | bauzas: we need a list of release blockers on there ASAP, basically, the tag is a way to suggest possible blockers | 13:09 |
| bauzas | sahid: since yesterday evening yes, and that's bugging my speakers | 13:10 |
| dims__ | sdague: all we are doing in that patch is fix the reloads we see in our gate | 13:10 |
| sdague | dims__: and changing the nova defaults so a default config doesn't trigger it | 13:10 |
| bauzas | johnthetubaguy: ok, as said to mikal, I offer my help | 13:11 |
| sdague | that's the more important thing for me, making sure our default config isn't a regression | 13:11 |
| *** xuhanp has quit IRC | 13:11 | |
| dims__ | sdague: fwiw, https://review.openstack.org/#/c/169535/ is merging now. So Eli's port can be merged - https://review.openstack.org/#/c/168784/ to fix the actual problem | 13:12 |
| sdague | dims__: it still adds a bunch of fs load | 13:12 |
| sdague | because you are reading the full directory contents now on every API | 13:13 |
| sdague | previously this was just a stat on a single file | 13:13 |
| *** tonytan4ever has joined #openstack-nova | 13:14 | |
| dims__ | sdague: so it's better than status quo which is if a deployer has a file in policy.d then there are tons of reloads of the policies | 13:14 |
| sdague | now this turns the cache invalidation into reading actual contents | 13:14 |
| sdague | dims__: there will still be half a dozen gratuitous extra file reads (of the directory) on every api call | 13:14 |
| dims__ | sdague: reading contents of files in policy.d? | 13:14 |
| *** ferest has quit IRC | 13:14 | |
| sdague | no, reading the directory | 13:14 |
| sdague | because you are getting the file list | 13:14 |
| johnthetubaguy | bauzas: feel free to trim down the RC1 list to only things you think we need to block the release on, and add kilo-rc-potential thingy tag to the maybe's, also looking through the critical and high bug list to spot major bloopers and add them to the RC1 list, not sure if others have started that processes yet | 13:14 |
| dims__ | i am not arguing that we revert the changes in default | 13:15 |
| sdague | dims__: I'm saying I don't actually think the base bug is fixed | 13:15 |
| *** kaufer has joined #openstack-nova | 13:15 | |
| bauzas | johnthetubaguy: ack | 13:15 |
| sdague | because this is still the add of a lot of fs load | 13:15 |
| *** dboik_ has quit IRC | 13:15 | |
| dims__ | sdague: so if a deployer chooses to do policy.d, you would prefer the policy reloads to just fs mtime checks? | 13:16 |
| dims__ | which is what we will have | 13:16 |
| sdague | dims__: no you don't | 13:17 |
| *** pixelb has joined #openstack-nova | 13:17 | |
| dims__ | sdague: am perfectly ok with changing our defaults https://review.openstack.org/#/c/169464/. but worried about what a deployer will see if they turn policy.d on by adding a file there | 13:17 |
| sdague | dims__: os.listdir | 13:18 |
| sdague | that creates a real filesystem read | 13:18 |
| *** otter768 has quit IRC | 13:18 | |
| sdague | of the directory | 13:18 |
| *** pask81 has quit IRC | 13:18 | |
| sdague | you are no longer just doing stats | 13:18 |
| dims__ | sdague: i know | 13:19 |
| sdague | which means it's a regression | 13:19 |
| sdague | and the new fix remains not actually a fix | 13:19 |
| *** pask81 has joined #openstack-nova | 13:20 | |
| dims__ | sdague: can you forget about the fs mtime listdir for a minute. then look at the status of nova after https://review.openstack.org/#/c/169464/ merges | 13:20 |
| dims__ | sdague: after that merges, what will happen if a deployer decides to try the policy.d config option and drop a new file for their own environment | 13:20 |
| sdague | dims__: right, all the same issues as before | 13:21 |
| dims__ | exactly | 13:21 |
| sdague | but the oslo fix, still isn't a fix | 13:21 |
| dims__ | i am totally with you | 13:21 |
| sdague | however, it means the shipped version of nova makes no reference to policy.d | 13:21 |
| sdague | which makes it unlikely people will use it | 13:22 |
| sdague | and I think that in liberty this needs to be revisited, because the current model of policy.json being able to be change without service restart is only minimally impactful when it's a single file in a well known location | 13:23 |
| *** suresh12 has quit IRC | 13:23 | |
| *** mriedem_away is now known as mriedem | 13:24 | |
| dims__ | fair enough, with that fix and given the fact that policy.d won't exist, the fs calls are not more than what it is now and it makes it slightly better for the cases where someone may turn it on accidentally | 13:24 |
| sdague | so I'm not convinced that merging an oslo patch which masks out the issue, because now we don't put debug info in there around the listdir is a good direction, that would just make it a hidden performance hit, vs. seeing the debug reload messages which makes you realize that we're not caching | 13:24 |
| *** aysyd has quit IRC | 13:25 | |
| sdague | basically the fix is a regression with no extra debug information to make you aware of it | 13:25 |
| sdague | so it's actually worse | 13:25 |
| *** jecarey has joined #openstack-nova | 13:26 | |
| sdague | at least with the current code you get some indication it's doing something suboptimal at debug level | 13:26 |
| dims__ | sdague: reloading message was when the file was getting reloaded. | 13:26 |
| dims__ | there's a os.path.exists before the listdir as well | 13:26 |
| dims__ | it's fine, no swaet | 13:27 |
| sdague | dims__: we seem to be going around in circles here, and I'm not sure why | 13:27 |
| sdague | in old code) | 13:27 |
| dims__ | sdague: that's fine. i'll stop arguing :) | 13:27 |
| sdague | if you only have policy.json, it's freshness checked with 1 stat call | 13:27 |
| sdague | if you have a directory, real system reads are done, and DEBUG about files being reloaded | 13:28 |
| sdague | in oslo "fix") | 13:28 |
| sdague | if you only have policy.json, it's freshness checked with 1 stat call | 13:28 |
| -openstackstatus- NOTICE: gerrit has been restarted to restore event streaming. any change events missed by zuul (between 12:48 and 13:28 utc) will need to be rechecked or have new approval votes set | 13:29 | |
| sdague | if you have policy.d the directory is fully read on every policy check with *no* DEBUG warning that you are now doing more fs io <===== regression | 13:29 |
| dims__ | sdague: that's easy add a DEBUG warning, but it's ok. i give up on the idea | 13:29 |
| *** pixelb has quit IRC | 13:30 | |
| *** haigang has joined #openstack-nova | 13:31 | |
| rajesht | dims, sdague: could you please review https://review.openstack.org/#/c/135202 | 13:31 |
| sdague | dims__: right, but it seems to be really layers of work around. It seems like if this is a desired feature we basically need to turn off automatic cache punch through | 13:31 |
| sdague | because moving beyond a single well known file makes that non sensible | 13:31 |
| dims__ | sdague: that i agree with as well | 13:31 |
| dansmith | sdague: I've been avoiding that because I figured it'd wait | 13:31 |
| *** artom_ has quit IRC | 13:32 | |
| sdague | dansmith: ok, seemed like sane cleanups pre release, but if you want to wait, that's cool | 13:32 |
| *** erkules has quit IRC | 13:32 | |
| *** erkules has joined #openstack-nova | 13:32 | |
| dansmith | sdague: it's a little more than cleanups | 13:32 |
| sdague | dansmith: ok | 13:32 |
| dansmith | sdague: I mean, I'd like lots of stuff to get in, but I've been holding off on this and other such things | 13:33 |
| bauzas | sdague: quick question, is https://review.openstack.org/#/c/162746/8/nova/compute/manager.py,cm not breaking String Freeze ? | 13:33 |
| sdague | so in instances like that, it would be handy to actually -1 the base patch | 13:33 |
| sdague | because you end up with this giant +2ed stack | 13:33 |
| dansmith | I think I said something about it on an earlier set, or maybe one below this, but sure | 13:33 |
| *** aysyd has joined #openstack-nova | 13:34 | |
| sdague | bauzas: yeh, probably. | 13:34 |
| *** dboik has joined #openstack-nova | 13:34 | |
| *** doron has joined #openstack-nova | 13:35 | |
| *** dkliban_afk is now known as dkliban | 13:35 | |
| *** alexpilotti has quit IRC | 13:35 | |
| bauzas | sdague: that comes to the bug I'm working on atm | 13:35 |
| *** signed8bit has joined #openstack-nova | 13:35 | |
| bauzas | sdague: https://bugs.launchpad.net/nova/+bug/1431291 | 13:36 |
| openstack | Launchpad bug 1431291 in OpenStack Compute (nova) "Scheduler Failures are no longer logged with enough detail for a site admin to do problem determination" [High,Confirmed] | 13:36 |
| bauzas | sdague: unless we ask for a StringFreeze exception once I'm uploading the change, there are few chances to get it in Kilo - unless we backport it | 13:36 |
| sdague | bauzas: that bug is worth breaking string freeze | 13:36 |
| sdague | garyk: https://bugs.launchpad.net/nova/+bug/1289205 - comments on whether this remains a valid bu? | 13:37 |
| openstack | Launchpad bug 1289205 in OpenStack Compute (nova) "VMware: 404 Not Found error when create snapshot" [Undecided,Incomplete] | 13:37 |
| bauzas | sdague: oki doki, continuing my work | 13:37 |
| *** VW_ has joined #openstack-nova | 13:37 | |
| garyk | sdague: at the moment it is pretty vague. i think that we need to ask them ti try and reproduce with the latest code in master. the snapshot has changed in K | 13:39 |
| garyk | that is we now upload a stream optimized image and what ever they are checking or using is no longer valid | 13:39 |
| *** doron_ has joined #openstack-nova | 13:39 | |
| garyk | sdague: in addition to that the last iteration on it was over a 8 months ago | 13:39 |
| *** doron has quit IRC | 13:40 | |
| *** burt has joined #openstack-nova | 13:40 | |
| sdague | garyk: ok, so in that case it's worth probably just marking as Invalid and say reopen if it's still an issue | 13:40 |
| garyk | sdague: ok, will do | 13:40 |
| *** hdd has joined #openstack-nova | 13:41 | |
| garyk | done | 13:41 |
| *** claudiub has quit IRC | 13:41 | |
| *** lpetrut has joined #openstack-nova | 13:42 | |
| sdague | thanks | 13:43 |
| mriedem | gdi, google | 13:43 |
| mriedem | clippy on my browser | 13:43 |
| dansmith | mriedem: why are you blaming google? | 13:44 |
| *** hdd has quit IRC | 13:44 | |
| mriedem | april fools day | 13:44 |
| dansmith | mriedem: why are you blaming google? | 13:44 |
| mriedem | b/c i love window | 13:44 |
| mriedem | *windows | 13:44 |
| mriedem | and eclipse | 13:45 |
| *** pask81 has quit IRC | 13:45 | |
| dansmith | heh | 13:45 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: consoleuth: remove an instance of mutation while iterating https://review.openstack.org/169751 | 13:45 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: consoleauth: Store access_url on token authorization https://review.openstack.org/169752 | 13:45 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: websocketproxy: Make protocol validation use connection_info https://review.openstack.org/169753 | 13:45 |
| *** kaufer1 has joined #openstack-nova | 13:46 | |
| *** diegows has joined #openstack-nova | 13:46 | |
| sdague | definitely blame eclipse | 13:46 |
| mriedem | jogo: so far the aiocpu job hasn't failed with this https://review.openstack.org/#/c/169520/ - but i'm also not seeing the debug message hit so we're not hitting the retry loop | 13:46 |
| * bauzas says hello to Pac-Man | 13:46 | |
| dims__ | mriedem: i threw in a css to hide it :) | 13:47 |
| *** kaufer has quit IRC | 13:47 | |
| *** mjturek1 has joined #openstack-nova | 13:47 | |
| dims__ | sdague: do we need to fix references to policy.d here as well? https://github.com/openstack/nova/blob/master/doc/source/devref/policy_enforcement.rst | 13:48 |
| sdague | dims__: maybe, it's devref | 13:49 |
| *** matrohon has quit IRC | 13:49 | |
| dims__ | k leaving it alone then | 13:49 |
| mriedem | dims__: sdague: naw | 13:49 |
| sdague | so I'm less concerned at the moment | 13:49 |
| mriedem | i saw that yesterday, that's more of a manifesto doc | 13:49 |
| mriedem | note the 'here is a vision' at the top | 13:50 |
| dims__ | mriedem: cool | 13:50 |
| dims__ | just crossing the t's and dotting the i's :) | 13:50 |
| openstackgerrit | Matt Riedemann proposed openstack/nova-specs: Show 'reserved' status in os-fixed-ips API https://review.openstack.org/168996 | 13:50 |
| *** haigang has quit IRC | 13:52 | |
| sdague | garyk: how about this one - https://bugs.launchpad.net/nova/+bug/1353977 | 13:53 |
| openstack | Launchpad bug 1353977 in OpenStack Compute (nova) "VMware vCenter Driver copy_virtual_disk failed with fileType error" [Undecided,Incomplete] | 13:53 |
| sdague | or - https://bugs.launchpad.net/nova/+bug/1346549 | 13:54 |
| openstack | Launchpad bug 1346549 in OpenStack Compute (nova) "VMware: "Storage error: Unable to find iSCSI Target"" [Undecided,Incomplete] | 13:54 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: consoleauth: remove an instance of mutation while iterating https://review.openstack.org/169751 | 13:54 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: consoleauth: Store access_url on token authorization https://review.openstack.org/169752 | 13:54 |
| openstackgerrit | Nikola Dipanov proposed openstack/nova: websocketproxy: Make protocol validation use connection_info https://review.openstack.org/169753 | 13:54 |
| ndipanov | dansmith, sdague ^^ | 13:54 |
| ndipanov | my thinking is that if we agree with that approach - getting it in for kilo would save some work on the deployment side of things | 13:54 |
| dims__ | mriedem: sdague: there was one more context bug that showed up. fix is here if you get a chance - https://review.openstack.org/#/c/168153/ | 13:55 |
| ndipanov | (pinging you guys since you approved the original bugfix. | 13:55 |
| *** jcoufal has quit IRC | 13:55 | |
| sdague | dims__: thanks | 13:56 |
| mriedem | ndipanov: i'm +2 on https://review.openstack.org/#/c/169751/ at least :) | 13:56 |
| garyk | sdague: looking | 13:57 |
| ndipanov | mriedem, yeah | 13:57 |
| dansmith | yeah | 13:57 |
| dansmith | the next one is an RPCAPI change | 13:57 |
| mriedem | https://bugs.launchpad.net/nova/+bug/1346549 seems like a dupe of a thing with a fix already | 13:57 |
| openstack | Launchpad bug 1346549 in OpenStack Compute (nova) "VMware: "Storage error: Unable to find iSCSI Target"" [Undecided,Incomplete] | 13:57 |
| mriedem | garyk: ^ | 13:57 |
| dansmith | and doesn't have a bug attached, so I'm not sure what the impact is | 13:57 |
| dansmith | I guess it's the bug attached to the last one | 13:58 |
| garyk | mriedem: there are a number of iscsi bugs that have been addressed by ian | 13:58 |
| dansmith | having the config in sync is a solved problem, right? | 13:58 |
| mriedem | dansmith: the bug ref is at the top in the middle patch | 13:58 |
| ndipanov | dansmith, right | 13:58 |
| ndipanov | unless you actually have to debug envornments controlled by puppet | 13:58 |
| *** iamjarvo has joined #openstack-nova | 13:59 | |
| mriedem | garyk: i was thinking this https://review.openstack.org/#/c/144600/ | 13:59 |
| *** signed8b_ has joined #openstack-nova | 13:59 | |
| mriedem | but i guess not the same | 13:59 |
| *** fawadkhaliq has quit IRC | 13:59 | |
| garyk | yes, that is the on. on irc his nickname is ian | 13:59 |
| sdague | the bottom patch is easy, I'm kind of wondering about the other ones | 13:59 |
| sdague | unrelated, the pci ci system seems mega face plant | 14:00 |
| garyk | mriedem: as mentioned onthe reviews the CI that we run uses the cinder vmdk driver. no testing has been done withthe iscsi driver | 14:00 |
| mriedem | garyk: yeah, remember that | 14:01 |
| garyk | i recall we asked that he paste a tempest run with the iscsi driver running. | 14:01 |
| ndipanov | that paperclip offers no useful advice whatsoever | 14:01 |
| garyk | but i am ok with this as the code is kosher | 14:01 |
| mriedem | ndipanov: so i'm not the only one | 14:02 |
| ndipanov | mriedem, broken if you ask me | 14:02 |
| ndipanov | UI has regressed a lot since 1998 | 14:02 |
| ndipanov | er AI | 14:02 |
| *** signed8bit has quit IRC | 14:03 | |
| sdague | ndipanov: :) | 14:03 |
| sdague | yeh, I don't think nibalizer had enough time to make it fully robust | 14:04 |
| ndipanov | nice touch I have to say | 14:04 |
| sdague | does solly ever hang out in irc any more? | 14:04 |
| ndipanov | if flaper87 hadn't spoiled for me on twitter I would have probably enjoyed it even more | 14:04 |
| sdague | :) | 14:04 |
| ndipanov | directxman12, you mean | 14:04 |
| ndipanov | ? | 14:05 |
| ndipanov | which is solly's real name | 14:05 |
| flaper87 | ndipanov: i'mmmmmmmmm sorrrrryyyyyyyy | 14:05 |
| garyk | sdague: there were comments added by vui in #2. The person opening the bug did not address them. I suggest that we close this one until there is another reproducet (https://bugs.launchpad.net/nova/+bug/1353977). The driver did undergo a refactor since this bug was opened | 14:05 |
| openstack | Launchpad bug 1353977 in OpenStack Compute (nova) "VMware vCenter Driver copy_virtual_disk failed with fileType error" [Undecided,Incomplete] | 14:05 |
| sdague | directxman12: hey - https://bugs.launchpad.net/nova/+bug/1278342 still a thing we need to worry about | 14:05 |
| openstack | Launchpad bug 1278342 in OpenStack Compute (nova) "novncproxy accepts un-masked client websocket frames" [Undecided,Incomplete] | 14:05 |
| sdague | garyk: sounds good to me | 14:05 |
| *** zz_jgrimm is now known as jgrimm | 14:05 | |
| dansmith | sdague: I'm pretty sure this was just a "drop clippy.js in the gerrit header" :) | 14:05 |
| sdague | dansmith: yep | 14:05 |
| dansmith | sdague: no robustification to be done, I mean :) | 14:06 |
| * kashyap was reading the history on "Critisim" about Clippy | 14:06 | |
| sdague | https://review.openstack.org/#/c/169010/ | 14:06 |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Show 'reserved' status in os-fixed-ips https://review.openstack.org/168966 | 14:06 |
| *** pask81 has joined #openstack-nova | 14:06 | |
| sdague | johnthetubaguy: you are the reporter on this - https://bugs.launchpad.net/nova/+bug/1214858 - 2 years ago, is this still an actionable kind of thing? | 14:07 |
| openstack | Launchpad bug 1214858 in OpenStack Compute (nova) "XenAPIBridgeDriver has annoying bugs when you have a VLAN misconfiguration" [Low,Incomplete] | 14:07 |
| *** Maike_ has quit IRC | 14:08 | |
| *** ijw has joined #openstack-nova | 14:08 | |
| kashyap | [OT] "Several days later, an Easter egg was found in the then-preview version of Windows Phone 8.1. When asked if she likes Clippy, the personal assistant Cortana would answer: "What's not to like? That guy took a heck of a beating and he's still smiling."" | 14:08 |
| *** signed8b_ has quit IRC | 14:08 | |
| *** vmtrooper has joined #openstack-nova | 14:08 | |
| johnthetubaguy | sdague: technically yes, if you use nova-network VLAN mode, but its kinda super low priority at this point | 14:09 |
| sdague | johnthetubaguy: is it a thing that's a bug or a feature? | 14:09 |
| *** hdd has joined #openstack-nova | 14:09 | |
| johnthetubaguy | sdague: its a bug, basically if you try to use it, it breaks in horrible ways | 14:09 |
| *** annegentle has quit IRC | 14:09 | |
| johnthetubaguy | (if you hit any problems, it errors out in a very messy way) | 14:10 |
| sdague | is it a thing that's ever going to be fixed by anyone? | 14:10 |
| *** zul has quit IRC | 14:10 | |
| raorn | please... look at https://review.openstack.org/#/c/163393/ | 14:11 |
| raorn | there's some discussion, need your opinion... | 14:11 |
| johnthetubaguy | sdague: hmm, let just nuke it I guess | 14:11 |
| *** zul has joined #openstack-nova | 14:11 | |
| openstackgerrit | Andrey Kurilin proposed openstack/python-novaclient: Adds os-compute-api-type flag https://review.openstack.org/167577 | 14:12 |
| sdague | ok, I'll mark as opinion | 14:12 |
| *** atuvenie has quit IRC | 14:13 | |
| *** vmtrooper has quit IRC | 14:13 | |
| *** rushiagr is now known as rushiagr_away | 14:14 | |
| *** nelsnelson has joined #openstack-nova | 14:15 | |
| *** thangp has joined #openstack-nova | 14:16 | |
| *** sdake has joined #openstack-nova | 14:16 | |
| *** mtanino has joined #openstack-nova | 14:18 | |
| openstackgerrit | jichenjc proposed openstack/nova: Do not delete volumes during reschedule https://review.openstack.org/169097 | 14:18 |
| *** eglynn has quit IRC | 14:19 | |
| *** ZZelle has quit IRC | 14:19 | |
| *** kaufer1 has quit IRC | 14:20 | |
| *** ZZelle has joined #openstack-nova | 14:20 | |
| *** eglynn has joined #openstack-nova | 14:20 | |
| *** BadCub_Hiding is now known as BadCub | 14:22 | |
| sdague | mriedem:any idea if this but is still a thing - https://bugs.launchpad.net/nova/+bug/1379511 - you are the reporter | 14:23 |
| openstack | Launchpad bug 1379511 in OpenStack Compute (nova) "v2 API: os-attach-interfaces:show will return 500 error if InstanceNotFound" [Undecided,Incomplete] | 14:23 |
| *** rajesht has quit IRC | 14:23 | |
| mriedem | sdague: i'll have to dig into it after finishing something else quick | 14:24 |
| *** kaufer has joined #openstack-nova | 14:24 | |
| *** pradipta has joined #openstack-nova | 14:24 | |
| *** bkopilov has quit IRC | 14:27 | |
| *** bkopilov has joined #openstack-nova | 14:28 | |
| directxman12 | sdague: ack re https://bugs.launchpad.net/nova/+bug/1278342 | 14:29 |
| openstack | Launchpad bug 1278342 in OpenStack Compute (nova) "novncproxy accepts un-masked client websocket frames" [Undecided,Incomplete] | 14:29 |
| directxman12 | sdague: do you want to file a bug against noVNC, or shall I do so? | 14:30 |
| sdague | directxman12: if you can, that would be great | 14:31 |
| sdague | directxman12: so is that really a novnc bug, or should nova be actively doing a thing as well? | 14:31 |
| sdague | we are doing the mask translation all the time | 14:31 |
| *** dave-mccowan has quit IRC | 14:32 | |
| directxman12 | sdague: sure thing. At a glance it looks like websockify itself should be a bit stricter (or provide an option to do so). I'll do a bit of investigating | 14:33 |
| *** mpaolino has quit IRC | 14:33 | |
| *** suresh12 has joined #openstack-nova | 14:34 | |
| sdague | directxman12: ok, cool. If it turns out there is no nova change needed would you mind marking that bug Invalid? | 14:35 |
| sdague | or if there is, say what is needed in nova | 14:35 |
| directxman12 | sdague: yep, no problem | 14:35 |
| sdague | directxman12: thank you | 14:35 |
| openstackgerrit | Alexey I. Froloff proposed openstack/nova: Store user-requested ephemeral size in instance object https://review.openstack.org/163393 | 14:36 |
| *** annegentle has joined #openstack-nova | 14:36 | |
| directxman12 | sdague: FWIW, I'm usually on IRC from ~10:30 EST until ~18:30 EST | 14:36 |
| *** alexpilotti has joined #openstack-nova | 14:38 | |
| sdague | directxman12: cool, thanks | 14:38 |
| *** suresh12 has quit IRC | 14:39 | |
| *** exploreshaifali has joined #openstack-nova | 14:39 | |
| *** cfriesen has joined #openstack-nova | 14:39 | |
| *** diegows has quit IRC | 14:43 | |
| sdague | garyk: more incomplete vmware bugs - https://bugs.launchpad.net/nova/+bug/1244752 | 14:43 |
| openstack | Launchpad bug 1244752 in OpenStack Compute (nova) "VMwareVCDriver: VSAN datastores cannot be used" [Wishlist,Incomplete] | 14:43 |
| *** artom has joined #openstack-nova | 14:43 | |
| openstackgerrit | Ed Leafe proposed openstack/nova: Optimize periodic call to get_by_host https://review.openstack.org/165849 | 14:44 |
| garyk | sdague: implemented in K | 14:44 |
| edleafe | Cores: quick bug fix suggested by johnthetubaguy ^^ | 14:45 |
| *** patrickeast has joined #openstack-nova | 14:45 | |
| *** ujjain has joined #openstack-nova | 14:45 | |
| sdague | garyk: can you set to Fix Committed then? | 14:46 |
| garyk | sdague: already did that :) | 14:46 |
| sdague | garyk: thanks | 14:46 |
| *** atuvenie has joined #openstack-nova | 14:47 | |
| *** yamahata has joined #openstack-nova | 14:48 | |
| openstackgerrit | garyk proposed openstack/nova: VMware: detach cinder volume when instance detroyed https://review.openstack.org/169732 | 14:49 |
| openstackgerrit | Mikhail Feoktistov proposed openstack/nova: Add config drive support for PCS containers https://review.openstack.org/149253 | 14:49 |
| *** unicell1 has joined #openstack-nova | 14:49 | |
| *** unicell has quit IRC | 14:50 | |
| *** Marga_ has joined #openstack-nova | 14:50 | |
| dansmith | edleafe: I echo the question about unit tests | 14:51 |
| *** mpaolino has joined #openstack-nova | 14:51 | |
| garyk | would it be possible that someone please look at https://review.openstack.org/169318 - it just adds instance=instance to debug log message inthe vmop code. this is hepful for debugging and trouble shooting | 14:53 |
| *** fawadkhaliq has joined #openstack-nova | 14:53 | |
| *** dguitarbite has quit IRC | 14:54 | |
| *** tsekiyama has joined #openstack-nova | 14:54 | |
| sdague | who is responsible for keeping an eye on the ceph job? | 14:57 |
| mriedem | no one really | 14:58 |
| *** emagana has joined #openstack-nova | 14:58 | |
| mriedem | i believe in paris pixelb was tagged, but that was awhile ago and kind of informal | 14:58 |
| *** kaufer has quit IRC | 14:58 | |
| mriedem | sdague: are you seeing new errors? | 14:59 |
| *** kaufer has joined #openstack-nova | 14:59 | |
| sdague | it was green until last week some time | 14:59 |
| sdague | now it's failing | 14:59 |
| dansmith | I'm seeing an ssh failure | 14:59 |
| sdague | if no one is keeping up on it, I'd rather just delete it and no use the nodes | 14:59 |
| *** Bsony has joined #openstack-nova | 15:00 | |
| mriedem | passed 2 days ago http://logs.openstack.org/66/168966/2/check/check-tempest-dsvm-full-ceph/b2caa16/ | 15:00 |
| mriedem | i'd rather make it voting so we don't break it once it's green | 15:00 |
| dansmith | it's not voting now | 15:00 |
| mriedem | which i think was the plan coming out of paris | 15:01 |
| mriedem | i know | 15:01 |
| *** ildikov_ has quit IRC | 15:01 | |
| dansmith | oh, rather make it voting, I see | 15:01 |
| dansmith | I agree, it seems to be a thing we want to keep working | 15:01 |
| mriedem | there was a push for some patches a few months ago that got it green | 15:01 |
| mriedem | then we didn't make it voting and it's regressed | 15:01 |
| dansmith | the fails I'm seeing are ssh-to-the-gues though, and only two in this run | 15:01 |
| mriedem | can someone link me to recent failures? | 15:01 |
| dansmith | are there others? or is this broken because the guest is not started? | 15:01 |
| sdague | mriedem: I'm only pro making it voting is someone is signed up to own it | 15:01 |
| mriedem | i'd like to logstash fu it | 15:01 |
| directxman12 | sdague: https://github.com/kanaka/websockify/issues/164 | 15:02 |
| *** annegent_ has joined #openstack-nova | 15:02 | |
| sdague | because making contributors have to go debug a ceph environment that have no ceph experience seems pretty bad form | 15:02 |
| mriedem | sdague: yeah, probably a nova meeting topic or ML then | 15:02 |
| mriedem | that's why pixelb was signed up in paris | 15:03 |
| dansmith | I'm asking about an owner | 15:03 |
| mriedem | btw, i think he was volunteered :) | 15:03 |
| dansmith | right | 15:03 |
| sdague | dansmith: no idea, it's been faceplant 100% of the time for a week from glancing at it | 15:03 |
| *** nellysmi_ has joined #openstack-nova | 15:03 | |
| dansmith | mriedem says two days | 15:03 |
| *** mpavone has quit IRC | 15:03 | |
| mriedem | i had a green run on 3/30 | 15:03 |
| *** nellysmitt has quit IRC | 15:04 | |
| *** annegentle has quit IRC | 15:04 | |
| mriedem | let me dig into logstash a bit | 15:04 |
| *** nellysmitt has joined #openstack-nova | 15:04 | |
| dansmith | my feeling is that ceph is becoming more popular and relevant to us, and that it's something we as a team should aim to keep working | 15:04 |
| dansmith | but I also understand the desire for an owner | 15:04 |
| *** packet has joined #openstack-nova | 15:04 | |
| mriedem | it is spiking http://goo.gl/9LKujL | 15:05 |
| sdague | dansmith: so, my issue with that is that if we keep expanding the scope of everything we as a team always need to keep working, basically everyone burns out and gives up | 15:05 |
| mriedem | b/c test_volume_boot_pattern was re-enabled 2 days ago | 15:05 |
| mriedem | sdague: jogo: dansmith: ^ | 15:05 |
| mriedem | dansmith: that's the ssh problem you're seeing | 15:05 |
| dansmith | mriedem: okay, that's a cinder thing, right? | 15:05 |
| *** nellysmitt has quit IRC | 15:05 | |
| dansmith | bfv? | 15:05 |
| sdague | dansmith: yeh | 15:05 |
| mriedem | yes | 15:06 |
| mriedem | it was busted forveer | 15:06 |
| mriedem | let me see how that's doing in the gate now | 15:06 |
| dansmith | sdague: I hear you, but if we ignore the cases that people really want to run, it all doesn't matter in the long run anyway :) | 15:06 |
| mriedem | i was surprised it was enabled again | 15:06 |
| dansmith | so if this test was voting, we wouldn't have enabled that test? | 15:06 |
| *** garyk has quit IRC | 15:06 | |
| sdague | it was a tempest enable, so I don't know | 15:07 |
| dansmith | okay | 15:07 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Fix serialization for Cells Responses https://review.openstack.org/169400 | 15:07 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Fix how service objects are looked up for Cells https://review.openstack.org/160506 | 15:07 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Fix how the Cells API is returning ComputeNode objects https://review.openstack.org/157185 | 15:07 |
| openstackgerrit | Sylvain Bauza proposed openstack/nova: Remove old relation in Cells for ComputeNode and Service https://review.openstack.org/168294 | 15:07 |
| sdague | I don't think this config is on tempest, again for the reasons that we can't run every by every | 15:07 |
| mriedem | i'm checking fail rates for test_volume_boot_pattern | 15:07 |
| *** claudiub has joined #openstack-nova | 15:07 | |
| *** nellysmi_ has quit IRC | 15:08 | |
| mriedem | so, the logstash for http://status.openstack.org/elastic-recheck/index.html#1373513 isn't showing a lot of hits, and some of those are juno since the devstack change probably wasn't backported yet | 15:08 |
| mriedem | checking that | 15:08 |
| mriedem | but that test must be doing something bad for ceph | 15:08 |
| *** ankit_ag has quit IRC | 15:08 | |
| *** Marga_ has quit IRC | 15:08 | |
| dansmith | I don't see fails that make it look like it's even on our end, but still digging | 15:10 |
| sdague | right, which puts us back to my concern, now you have to become a ceph expert to figure that out | 15:10 |
| mriedem | https://review.openstack.org/#/c/165281/ is the devstack change, not on stable/juno, i can work on that - but it's not exploding hard in the gate from what i can tell | 15:10 |
| *** rushiagr_away is now known as rushiagr | 15:11 | |
| sdague | mriedem: right, so that was the change to make the lvm case more robust | 15:11 |
| sahid | ndipanov: hello, i would be happy to get you return on this https://review.openstack.org/#/c/163393/ if you have moment. | 15:11 |
| *** dsanders has joined #openstack-nova | 15:13 | |
| dansmith | so this is supposed to be a config that uses ceph for instance images, right? | 15:16 |
| mriedem | well i'll at least get a bug opened for the ceph job choking on test_volume_boot_pattern | 15:16 |
| dansmith | and it's doing a boot from volume off of cinder.. is cinder backed to ceph? | 15:16 |
| mriedem | i think it's supposed to be in the job config | 15:16 |
| mriedem | http://logs.openstack.org/53/169753/2/check/check-tempest-dsvm-full-ceph/07dcae0/logs/etc/cinder/cinder.conf.txt.gz | 15:16 |
| *** ildikov_ has joined #openstack-nova | 15:16 | |
| mriedem | default_volume_type = ceph enabled_backends = ceph | 15:16 |
| dansmith | ah, yep | 15:16 |
| mriedem | glance is also using rbd backend: http://logs.openstack.org/53/169753/2/check/check-tempest-dsvm-full-ceph/07dcae0/logs/etc/glance/glance-api.conf.txt.gz | 15:17 |
| mriedem | stores = file, http, rbd | 15:17 |
| dansmith | thar she blows | 15:17 |
| dansmith | http://logs.openstack.org/46/169446/2/check/check-tempest-dsvm-full-ceph/7d30778/logs/screen-g-api.txt.gz?level=TRACE | 15:18 |
| dansmith | maybe | 15:18 |
| dansmith | maybe devstack actually failed to upload the image? | 15:18 |
| *** chinmay_ has quit IRC | 15:18 | |
| mriedem | http://logs.openstack.org/46/169446/2/check/check-tempest-dsvm-full-ceph/7d30778/logs/devstacklog.txt.gz#_2015-03-31_20_19_16_367 | 15:19 |
| ndipanov | sahid, hmmm I haven't reviewed that one before | 15:19 |
| ndipanov | let me see | 15:19 |
| dansmith | mriedem: looks successful to me | 15:20 |
| mriedem | yeah | 15:20 |
| mriedem | devstack would bomb out way before tempest if that failed | 15:20 |
| dansmith | well, should, I know | 15:20 |
| mriedem | cinder logs don't show any errors | 15:21 |
| kashyap | sdague: With libvirtd logging enabled, the log file size is 2MB, is that acceptable? -- http://logs.openstack.org/79/169679/1/check/check-tempest-dsvm-full/bb78b79/logs/libvirt/ | 15:21 |
| raorn | sahid: we can discuss it here if you want to | 15:21 |
| kashyap | sdague: With specific filters, i.e. | 15:21 |
| *** kaufer has quit IRC | 15:21 | |
| *** kaufer1 has joined #openstack-nova | 15:21 | |
| sdague | kashyap: can you ask on -infra, I don't know where we stand on log runs | 15:23 |
| sdague | that's 2x the n-api log, for instance | 15:23 |
| kashyap | sdague: I just added Clark to that review and asked him there. | 15:24 |
| raorn | sahid: you are saying, we should always use BDM fo calculate ephemeral size. then what's the point of having instance.ephemeral_gb in this case? it contains wrong number and not used at all | 15:24 |
| kashyap | sdague: Yeah, I realize, hence I was trying to get opinions from the Infra folks. | 15:24 |
| openstackgerrit | Mikhail Feoktistov proposed openstack/nova: Add config drive support for PCS containers https://review.openstack.org/149253 | 15:24 |
| dansmith | mriedem: surely looks like that's the first request for that image | 15:24 |
| sdague | kashyap: cool | 15:26 |
| mriedem | https://bugs.launchpad.net/glance/+bug/1439273 | 15:27 |
| openstack | Launchpad bug 1439273 in OpenStack Compute (nova) "check-tempest-dsvm-full-ceph failing on test_volume_boot_pattern scenario after 3/30" [Undecided,New] | 15:27 |
| mriedem | dansmith: ^ for tracking | 15:27 |
| *** bkopilov has quit IRC | 15:27 | |
| *** hemnafk is now known as hemna | 15:28 | |
| dansmith | mriedem: cool | 15:28 |
| dansmith | flaper87: around? | 15:28 |
| *** Marga_ has joined #openstack-nova | 15:28 | |
| *** TobiasE has quit IRC | 15:28 | |
| mriedem | dansmith: the g-api errors aren't specific to the ceph job http://goo.gl/YRh9e4 | 15:29 |
| dansmith | mriedem: wtf? | 15:29 |
| mriedem | dansmith: lots of services trace in non-error cases for tempest runs :( | 15:30 |
| mriedem | e.g. neutron logs in a successful gate run | 15:30 |
| dansmith | this is a lot of red for "normal" | 15:30 |
| mriedem | heh http://logs.openstack.org/53/169753/2/check/check-tempest-dsvm-full-ceph/07dcae0/logs/screen-ceilometer-acompute.txt.gz?level=TRACE | 15:30 |
| dansmith | sheesh | 15:31 |
| sahid | raorn: yes that is why i have asked ndipanov to take a look at the review, do we have to use it to denormalize block_device_mapping or not? | 15:31 |
| *** exploreshaifali has quit IRC | 15:31 | |
| *** Marga_ has quit IRC | 15:31 | |
| sahid | raorn: in the link i have sahed with you http://git.openstack.org/cgit/openstack/nova/tree/nova/compute/manager.py#n3692 | 15:31 |
| sahid | you can see line 3733 your change will be reverted | 15:32 |
| sahid | in this case | 15:32 |
| raorn | ah, finish_revert_resize... | 15:32 |
| *** annashen has joined #openstack-nova | 15:33 | |
| sahid | so we need a real consensus to know if we have to use block_device_mapping each time we need this information | 15:33 |
| sahid | or if we want to denormalize this information with this attribute | 15:33 |
| sahid | when we know that we can go on a good way to fix that bug | 15:34 |
| sahid | also i do not believe we want to neglect the swap case :) | 15:34 |
| mriedem | dansmith: these also show up in a lot of jobs http://logs.openstack.org/53/169753/2/check/check-tempest-dsvm-full-ceph/07dcae0/logs/screen-n-cpu.txt.gz?level=TRACE | 15:34 |
| kashyap | sdague: Actually, I stand corrected: It's 2 MB of _compressed. So, uncompressed: it's 21MB. But, don't worry: it won't be indexed until a future solution for that is figured out. | 15:34 |
| mriedem | maybe we shouldn't be trying to find a fixed ip with no address :) | 15:35 |
| mriedem | checking that n-net code | 15:35 |
| dansmith | mriedem: I saw one request or response in the tempest trace that had instance_uuid: null for a network thing... | 15:35 |
| *** hemna has quit IRC | 15:36 | |
| dansmith | mriedem: so the libvirt xml for the guest looks right to me | 15:37 |
| dansmith | mriedem: and it seems like we're doing all the right things there | 15:37 |
| *** annashen has quit IRC | 15:37 | |
| dansmith | sdague: does this mean that the guest had no console log output? http://logs.openstack.org/46/169446/2/check/check-tempest-dsvm-full-ceph/7d30778/console.html#_2015-03-31_20_46_35_266 | 15:38 |
| sdague | dansmith: yes, I think so | 15:38 |
| dansmith | given everything else that looks right, | 15:40 |
| *** Marga_ has joined #openstack-nova | 15:40 | |
| dansmith | I assume that means our image on the volume is borked | 15:40 |
| dansmith | I wonder if cinder is responsible for that in the rbd case like it is for the normal bfv case... | 15:40 |
| *** hdd has quit IRC | 15:41 | |
| raorn | sahid: usually swap is hundred times smaller (gigabytes vs. tens or hundreds gigabytes) than ephemeral disk, it could be neglected. doesn't mean it should, but swap potentially creates less problems than ephemeral and is missing from instance object... | 15:41 |
| *** hemna has joined #openstack-nova | 15:42 | |
| mriedem | dansmith: bingo http://logs.openstack.org/53/169753/2/check/check-tempest-dsvm-full-ceph/07dcae0/logs/screen-n-cpu.txt.gz#_2015-04-01_14_34_35_696 | 15:43 |
| dansmith | awesome that that is DEBUG level | 15:44 |
| mriedem | i'm bumping it up now | 15:44 |
| dansmith | well, | 15:44 |
| dansmith | the actual thing should probably be debug, and we should log at error in driver wherever we're callign that | 15:44 |
| *** hemna has quit IRC | 15:45 | |
| dansmith | I think it might be used in other places to test presence | 15:45 |
| *** rushil has joined #openstack-nova | 15:45 | |
| mriedem | ah yeah there is an exists check | 15:45 |
| dansmith | also | 15:45 |
| dansmith | I don't think I see that in the exact case I'm looking at | 15:46 |
| *** bkopilov has joined #openstack-nova | 15:46 | |
| *** bkopilov has quit IRC | 15:46 | |
| *** marun has joined #openstack-nova | 15:47 | |
| dansmith | so libvirt does not do the image write on rbd BFV, just like the normal case | 15:48 |
| *** aysyd has quit IRC | 15:49 | |
| *** xyang1 has joined #openstack-nova | 15:49 | |
| *** hemna has joined #openstack-nova | 15:49 | |
| *** yamahata has quit IRC | 15:50 | |
| *** yamahata has joined #openstack-nova | 15:50 | |
| *** mpaolino has quit IRC | 15:50 | |
| dansmith | no errors from qemu about failure to hit the root disk either | 15:51 |
| *** Marga_ has quit IRC | 15:51 | |
| *** garyk has joined #openstack-nova | 15:52 | |
| dansmith | nor libvirt | 15:52 |
| *** Marga_ has joined #openstack-nova | 15:52 | |
| dansmith | which kinda seems to indicate that all the ceph bits are working | 15:52 |
| dansmith | I mean, presumably | 15:52 |
| *** iamjarvo has quit IRC | 15:53 | |
| dansmith | and yeah, cinder is doing the image extract onto the volume | 15:54 |
| mriedem | dansmith: somewhat related, question in your rbd retry patach https://review.openstack.org/#/c/169446/ | 15:55 |
| mriedem | *patch | 15:55 |
| *** hdd has joined #openstack-nova | 15:55 | |
| *** unicell1 has quit IRC | 15:55 | |
| *** Marga_ has quit IRC | 15:55 | |
| *** Marga_ has joined #openstack-nova | 15:55 | |
| dansmith | mriedem: I think the hassnapshots might be waiting for the snapshots to delete so the main image can delete | 15:56 |
| dansmith | mriedem: this isn't really my patch, I just cleaned it up and added tests | 15:56 |
| dansmith | the logic came from ceph-knowing people | 15:56 |
| dansmith | mriedem: but also, I need to bump that up to 10s it sounds like | 15:56 |
| dansmith | they had it loop forever | 15:56 |
| dansmith | and I chose 3 seconds somewhat randomly, which doesn't seem to be long enough if you're hammering the crap out of it | 15:56 |
| dansmith | rook is trying 10s now to see if that's a better number | 15:57 |
| mriedem | are there any pre-existing retry/timeout config options that could be re-used? | 15:57 |
| *** vmtrooper has joined #openstack-nova | 15:57 | |
| dansmith | I avoided adding one because of the impact and timing, but if there is, that'd be cool | 15:58 |
| mriedem | looking | 15:58 |
| *** dboik_ has joined #openstack-nova | 16:00 | |
| *** doron_ is now known as doron_afk | 16:00 | |
| dansmith | all the cinder bits seem to be happening in the log too | 16:01 |
| dansmith | creating the volume, cloning the image, converting it from ami to raw I think, etc | 16:01 |
| dansmith | and all that happens before attach | 16:01 |
| *** tonytan_brb has joined #openstack-nova | 16:02 | |
| *** vmtrooper has quit IRC | 16:02 | |
| *** tonytan4ever has quit IRC | 16:02 | |
| mriedem | dansmith: want me to hold off on +W for https://review.openstack.org/#/c/169446/ on that timeout number? | 16:02 |
| dansmith | mriedem: might as well.. 3s fixes it most of the time, but when they're hammering it like crazy, they can make it fail even with 3s.. so it's an improvement that we could tweak later, but I think we could wait too | 16:03 |
| dansmith | I think we should hear from rook soonish anyway | 16:03 |
| mriedem | would be nice if this ceph job logged the ceph.conf | 16:03 |
| dansmith | and it's possible that under heavy load, 10s isn't enough either | 16:03 |
| *** mpaolino has joined #openstack-nova | 16:03 | |
| dansmith | mriedem: yeah, was noticing that | 16:04 |
| *** dboik has quit IRC | 16:04 | |
| mriedem | it can be added | 16:04 |
| rook | hope so... should be close -- FIO is running on all of them :) | 16:04 |
| *** aysyd has joined #openstack-nova | 16:04 | |
| *** arnaud___ has joined #openstack-nova | 16:05 | |
| *** whenry has joined #openstack-nova | 16:06 | |
| *** pixelb has joined #openstack-nova | 16:07 | |
| *** dboik_ has quit IRC | 16:08 | |
| *** dboik has joined #openstack-nova | 16:08 | |
| *** patrickeast has quit IRC | 16:09 | |
| *** dave-mccowan has joined #openstack-nova | 16:10 | |
| openstackgerrit | Sean Dague proposed openstack/nova: fix virtual machine evacuate under neutron https://review.openstack.org/169827 | 16:10 |
| mriedem | well the network teardown thing with no fixed ips is obviously a separate issue | 16:11 |
| mriedem | that code could be tightened up so we don't trace all over hell though | 16:12 |
| *** jbernard has joined #openstack-nova | 16:12 | |
| mriedem | i'll open a separate bug for that | 16:12 |
| dansmith | sdague: mriedem: jbernard was already looking at these failures | 16:12 |
| dansmith | and sounds like he should be the current owner | 16:12 |
| mriedem | for rbd? | 16:12 |
| dansmith | ceph testing yeah | 16:12 |
| jbernard | at least test_volume_boot_pattern seems to break consistently, again | 16:12 |
| mriedem | sounds good to me, he was the one that fixed the stuff a few months back with resize | 16:12 |
| dansmith | yeah | 16:13 |
| jbernard | but it's not always the same | 16:13 |
| *** alanf-mc has joined #openstack-nova | 16:13 | |
| jbernard | mriedem: have you noticed anything in addition? | 16:13 |
| sdague | dansmith: good to know | 16:13 |
| *** devlaps has joined #openstack-nova | 16:13 | |
| mriedem | jbernard: not really, my notes are in the bug report | 16:13 |
| dansmith | jbernard: the stuff I was looking at looked like everything was happy, but the guest started with a borken disk and just never...booted | 16:13 |
| jbernard | mriedem: what's the #? | 16:14 |
| mriedem | https://bugs.launchpad.net/glance/+bug/1439273 | 16:14 |
| openstack | Launchpad bug 1439273 in OpenStack Compute (nova) "check-tempest-dsvm-full-ceph failing on test_volume_boot_pattern scenario after 3/30" [Undecided,New] | 16:14 |
| *** Mike_D_laptop has quit IRC | 16:14 | |
| dansmith | jbernard: cinder did the volume create, image lay, we booted the guest with the right libvirt xml, but the guest console never showed any bootiness | 16:14 |
| *** doron_afk has quit IRC | 16:15 | |
| jbernard | dansmith, mriedem: ok, ill take a look and see what I can come up with | 16:15 |
| *** ferest has joined #openstack-nova | 16:15 | |
| dansmith | jbernard: thanks, let me know if I can help at all | 16:15 |
| jbernard | will do | 16:15 |
| *** rushil has quit IRC | 16:15 | |
| jbernard | we see it locally as well, or only in the gate? | 16:15 |
| dansmith | jbernard: I was going to try to get a devstack running with ceph to try to reproduce | 16:15 |
| *** Sukhdev has joined #openstack-nova | 16:16 | |
| jbernard | dansmith: ill do the same | 16:16 |
| dansmith | cool | 16:16 |
| *** alexpilotti has quit IRC | 16:16 | |
| mriedem | i'll see what i can do to get ceph.conf in the ceph job archive | 16:16 |
| dansmith | sweet | 16:18 |
| *** dulek has quit IRC | 16:19 | |
| sdague | sahid: can you dup your massive repeating bug filing? | 16:19 |
| sdague | https://bugs.launchpad.net/nova/+bugs?search=Search&field.status=New | 16:19 |
| sdague | 8 bugs, same title | 16:19 |
| *** annashen has joined #openstack-nova | 16:20 | |
| openstackgerrit | Ed Leafe proposed openstack/nova: Optimize periodic call to get_by_host https://review.openstack.org/165849 | 16:20 |
| *** andreykurilin_ has joined #openstack-nova | 16:21 | |
| sahid | sdague: well i think one bug repported was not enough so... | 16:21 |
| edleafe | johnthetubaguy: dansmith: bauzas: Added the test ^^ | 16:21 |
| sahid | sdague: it was a bug with launchpad.net not my fault | 16:21 |
| sdague | mriedem: https://bugs.launchpad.net/nova/+bug/1439302 | 16:21 |
| openstack | Launchpad bug 1439302 in OpenStack Compute (nova) ""FixedIpNotFoundForAddress: Fixed ip not found for address None." traces in gate runs" [Undecided,New] | 16:21 |
| winston-d | dims__: ping | 16:21 |
| sdague | sahid: sure, can you clean it up though? | 16:21 |
| mriedem | sdague: i just opened that, like 1 second ago :) | 16:22 |
| sahid | sdague: yes i will do | 16:22 |
| sdague | mriedem: yeh, well I'm processing new bugs :) | 16:22 |
| sdague | that should be tagged ceph right? | 16:22 |
| mriedem | i noticed | 16:22 |
| mriedem | no | 16:22 |
| mriedem | it's not specific to the ceph job | 16:22 |
| *** annegent_ has quit IRC | 16:22 | |
| mriedem | the instance gets deleted while it's being built | 16:22 |
| sdague | do you have a non ceph set of logs to link? | 16:22 |
| dansmith | lol | 16:22 |
| dansmith | tempest.scenario.test_volume_boot_pattern.TestVolumeBootPatternV2.test_volume_boot_pattern ... SKIPPED: Skipped until Bug: 1373513 is resolved. | 16:22 |
| openstack | bug 1373513 in Cinder "Lvm hang during tempest tests" [Critical,Fix committed] https://launchpad.net/bugs/1373513 - Assigned to John Griffith (john-griffith) | 16:22 |
| mriedem | sdague: http://goo.gl/ktyzeP | 16:23 |
| sdague | mriedem: ok, can you post it into the bug? | 16:23 |
| mriedem | the logstash query is in the bug report | 16:23 |
| mriedem | here is a non-ceph failure http://logs.openstack.org/70/167370/6/check/check-tempest-dsvm-full/9924145/logs/screen-n-cpu.txt | 16:24 |
| jgriffith | sdague: I thought we re-enabled that test? | 16:24 |
| mriedem | it's not really a failure, it's ugly trace | 16:24 |
| mriedem | the test actually passes | 16:24 |
| mriedem | jgriffith: yes, we did | 16:24 |
| jgriffith | sdague: as the bug has been fixed | 16:24 |
| jgriffith | mriedem: pheww.. ok | 16:24 |
| mriedem | jgriffith: something about it is making the ceph job puke | 16:24 |
| dansmith | 100% of the time | 16:24 |
| mriedem | which is non-voting, but still | 16:24 |
| jgriffith | I haven't lost my mind *completely* | 16:24 |
| dansmith | I don't know why it's still skipping locally though | 16:24 |
| jgriffith | got ya | 16:24 |
| mriedem | dansmith: do you have latest tempest? | 16:25 |
| dansmith | I might not've recloned | 16:25 |
| jgriffith | dansmith: I've noticed lately a lot of folks are using static tempest checkout... or adding their own skips in :( | 16:25 |
| *** gszasz has quit IRC | 16:25 | |
| dansmith | cd | 16:25 |
| dansmith | oops | 16:25 |
| jgriffith | dansmith: oh.. .it's your setup... sorry then, never mind | 16:25 |
| dansmith | yar | 16:25 |
| *** exploreshaifali has joined #openstack-nova | 16:25 | |
| *** ildikov_ has quit IRC | 16:26 | |
| *** alanf-mc_ has joined #openstack-nova | 16:33 | |
| *** mdbooth has quit IRC | 16:35 | |
| rook | dansmith: hm, not sure if the issue is related to the 10s retry : https://gist.github.com/jtaleric/aff908889e22276f7e8b | 16:36 |
| *** alanf-mc has quit IRC | 16:36 | |
| *** jaypipes has quit IRC | 16:37 | |
| dansmith | rook: did I fat finger that? | 16:37 |
| dansmith | I put imagenotfound, but it's supposed to be imagebusy on the except: | 16:38 |
| rook | i just caught it. | 16:38 |
| dansmith | right? | 16:38 |
| rook | yeah | 16:38 |
| rook | ImageBusy | 16:38 |
| *** arnaud___ has quit IRC | 16:38 | |
| dansmith | sonofa | 16:38 |
| rook | i was like, why the ef is this not being caught | 16:38 |
| dansmith | I'm sure that was my bad when changing the patch | 16:38 |
| rook | ok | 16:38 |
| dansmith | will fix | 16:38 |
| rook | rgr, going to patch here locally and go back to 6 | 16:38 |
| *** pask81 has quit IRC | 16:38 | |
| *** andreykurilin_ has quit IRC | 16:39 | |
| openstackgerrit | Dan Smith proposed openstack/nova: Fixes _cleanup_rbd code to capture ImageBusy exception https://review.openstack.org/169446 | 16:39 |
| dansmith | rook: ^ | 16:39 |
| *** markus_z has quit IRC | 16:40 | |
| *** unicell has joined #openstack-nova | 16:41 | |
| *** alanf-mc_ has quit IRC | 16:41 | |
| *** alanf-mc has joined #openstack-nova | 16:42 | |
| *** mdbooth has joined #openstack-nova | 16:42 | |
| rook | dansmith: so, idle Instances would never hit this... now how "busy" do they have to be, not sure. I am running FIO, so possibly worse case right now. | 16:43 |
| sahid | sdague: i can see that you are triaging bugs; aobut this one: https://bugs.launchpad.net/nova/+bug/1300775 | 16:43 |
| openstack | Launchpad bug 1300775 in OpenStack Compute (nova) "Scheduler, performance impact when dealing with aggregates" [Medium,Confirmed] | 16:43 |
| dansmith | rook: yep | 16:43 |
| sahid | you have confirmed it but i think we can consider it as fixed since https://review.openstack.org/#/c/159904/ | 16:43 |
| openstackgerrit | Tomi Juvonen proposed openstack/nova-specs: State of compute node and instances running on it should be changed down by new API when host fault detected by external tool. https://review.openstack.org/169836 | 16:43 |
| sahid | sdague: do you have any feedbacks or can i close it? | 16:44 |
| *** jistr|demo has quit IRC | 16:47 | |
| dansmith | jbernard: I see it locally | 16:48 |
| jbernard | dansmith: me too | 16:48 |
| dansmith | jbernard: that's "good" :) | 16:48 |
| jbernard | dansmith: do you see a traceback in your g-reg log? | 16:48 |
| *** winston-d is now known as winston-d_zZZ | 16:48 | |
| dansmith | jbernard: I see lots of them, but they're supposedly unrelated per mriedem | 16:49 |
| jbernard | ok, good to know | 16:49 |
| *** derekh has quit IRC | 16:49 | |
| *** ZZelle is now known as ZZelle__ | 16:51 | |
| jbernard | dansmith: aside from that, i dont see any other log errors | 16:51 |
| bauzas | man, I hate the dependency merge issue | 16:52 |
| *** hblixt has joined #openstack-nova | 16:52 | |
| *** dims__ has quit IRC | 16:53 | |
| jbernard | dansmith: nova appears to believe the instance has booted, i wonder if the cinder volume created from the image contains the correct bootable data | 16:53 |
| *** jlvillal has quit IRC | 16:53 | |
| *** dims_ has joined #openstack-nova | 16:53 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/nova: Ironic: Extra configdrive metadata from Nodes https://review.openstack.org/145302 | 16:54 |
| *** Marga_ has quit IRC | 16:54 | |
| *** Marga_ has joined #openstack-nova | 16:55 | |
| dansmith | jbernard: that's what I was saying above | 16:55 |
| *** ferest has quit IRC | 16:55 | |
| dansmith | jbernard: either the image seems broken, or we're connected but not actually able to read from it or something like that | 16:55 |
| jbernard | dansmith: i hit this last week: https://bugs.launchpad.net/cinder/+bug/1430751 | 16:56 |
| openstack | Launchpad bug 1430751 in Cinder "Launching instance that creates a new volume fails" [Undecided,Incomplete] | 16:56 |
| jbernard | dansmith: maybe unrelated, but it makes me question | 16:56 |
| jbernard | dansmith: is there a way to prevent temptest from tearing down after a failed test? | 16:56 |
| *** dboik has quit IRC | 16:57 | |
| *** moshele has quit IRC | 16:57 | |
| dansmith | jbernard: I think you have to hack cleanUp | 16:57 |
| mriedem | right | 16:57 |
| dansmith | jbernard: well, we don't see it fail that way, but yeah | 16:57 |
| *** dboik has joined #openstack-nova | 16:57 | |
| mriedem | otherwise you'd go over quota | 16:57 |
| *** hdd has quit IRC | 16:57 | |
| jbernard | ok, i would like to verify the contents for the cinder volume | 16:58 |
| jbernard | so ill give it a try, unless someone beats me to it | 16:58 |
| dansmith | that's what I'd do next | 16:58 |
| dansmith | (if I were you) :) | 16:58 |
| *** lpetrut has quit IRC | 16:59 | |
| *** alexpilotti has joined #openstack-nova | 17:00 | |
| *** whenry has quit IRC | 17:00 | |
| *** jlvillal has joined #openstack-nova | 17:00 | |
| sdague | sahid: sounds good if you can move it to Fixed | 17:01 |
| *** harlowja_away is now known as harlowja | 17:02 | |
| *** tonytan_brb has quit IRC | 17:02 | |
| *** sdake_ has joined #openstack-nova | 17:02 | |
| *** kaufer1 has quit IRC | 17:02 | |
| *** kaufer has joined #openstack-nova | 17:03 | |
| *** emagana has quit IRC | 17:04 | |
| *** ildikov_ has joined #openstack-nova | 17:05 | |
| *** jlvillal has quit IRC | 17:05 | |
| *** tjones1 has joined #openstack-nova | 17:05 | |
| *** sdake has quit IRC | 17:05 | |
| *** VW_ has quit IRC | 17:06 | |
| *** aysyd has quit IRC | 17:07 | |
| *** achanda has joined #openstack-nova | 17:08 | |
| *** tonytan4ever has joined #openstack-nova | 17:09 | |
| sahid | done | 17:09 |
| *** sahid has quit IRC | 17:09 | |
| *** achanda has quit IRC | 17:10 | |
| *** romainh has quit IRC | 17:10 | |
| *** lucasagomes is now known as lucas-dinner | 17:10 | |
| *** jlvillal has joined #openstack-nova | 17:11 | |
| *** annegentle has joined #openstack-nova | 17:12 | |
| *** oro has quit IRC | 17:13 | |
| mriedem | sdague: given the timing of that fixedipnotfoundforinstance trace on deallocate_for_instance, going back to 3/22 in logstash, it started showing up on 3/26 and i'm not seeing much in nova changing around then which seems related, | 17:13 |
| mriedem | sdague: but this merged in tempest on 3/26 which looks possibly related https://github.com/openstack/tempest/commit/9ce97dfa353e3659aa0ec0a7f62ed3a7e54c4bdf | 17:13 |
| dansmith | jbernard: I have a recipe to avoid killing the server if you want/need it | 17:13 |
| *** achanda has joined #openstack-nova | 17:14 | |
| *** Sukhdev has quit IRC | 17:14 | |
| *** patrickeast has joined #openstack-nova | 17:14 | |
| *** penick has joined #openstack-nova | 17:14 | |
| dansmith | jbernard: http://pastebin.com/pncPjZXH | 17:15 |
| sdague | easy fix up - https://review.openstack.org/#/c/168745 | 17:15 |
| sdague | doesnt' change nova code, just fixes conf generator | 17:15 |
| sdague | mriedem: could be | 17:15 |
| *** yamahata has quit IRC | 17:16 | |
| dansmith | sdague: got it | 17:16 |
| mriedem | sdague: i'm adding some logging in that flow to see what's coming in | 17:16 |
| *** Nic has joined #openstack-nova | 17:17 | |
| mriedem | i think requested_networks list of tuples is getting built with a network id and some invalid value for ip | 17:17 |
| jbernard | dansmith: ah, thanks; i just added a 20 minute sleep to the test ;) | 17:17 |
| dansmith | jbernard: lol | 17:17 |
| *** Marga_ has quit IRC | 17:18 | |
| *** iamjarvo has joined #openstack-nova | 17:19 | |
| *** aysyd has joined #openstack-nova | 17:23 | |
| *** pradipta has quit IRC | 17:24 | |
| *** welldannit has joined #openstack-nova | 17:24 | |
| *** atuvenie has quit IRC | 17:26 | |
| *** zzzeek has joined #openstack-nova | 17:27 | |
| *** lucas-dinner has quit IRC | 17:28 | |
| *** alanf-mc has quit IRC | 17:29 | |
| jbernard | dansmith: from rbd export, volume is empty… | 17:29 |
| dansmith | jbernard: orly | 17:29 |
| jbernard | dansmith: well, it's a valid ext3 fs, with no files | 17:29 |
| dansmith | jbernard: so the uec conversion to raw failed or something? | 17:30 |
| *** claudiub has quit IRC | 17:30 | |
| jbernard | would an empty boot volume fail in this way? | 17:30 |
| dansmith | sure | 17:31 |
| dansmith | actually, hang on | 17:31 |
| *** emagana has joined #openstack-nova | 17:31 | |
| dansmith | yea | 17:31 |
| dansmith | it's not doing direct kernel boot, AFAICT | 17:32 |
| dansmith | which means it needs to have a bootloader | 17:32 |
| jbernard | and the instance would move to 'active' with no console? | 17:33 |
| dansmith | yeah | 17:33 |
| dansmith | nova considers it active if it | 17:33 |
| dansmith | is running | 17:33 |
| dansmith | which it is | 17:33 |
| dansmith | after that, nova has no idea | 17:33 |
| *** armax has quit IRC | 17:33 | |
| jbernard | yep :) | 17:33 |
| jbernard | it's succesfully running nothing | 17:33 |
| dansmith | right | 17:34 |
| dansmith | so, what does cinder do to prep the image from a uec? | 17:34 |
| jbernard | so, image convertion or volume creation from an image | 17:34 |
| jbernard | that ill have to look at | 17:34 |
| dansmith | jbernard: do you know there's no bootloader on the image? | 17:35 |
| dansmith | jbernard: and, what is the easiest way I can look at my local volume? | 17:35 |
| mriedem | heh, it's a good thing that update_dns_entries defaults to False b/c this is totally borked, AttributeError written all over it: https://github.com/openstack/nova/blob/2015.1.0b3/nova/network/manager.py#L561 | 17:35 |
| mriedem | vishy: ^ | 17:35 |
| jbernard | i read the hexdump, it looks like a fresh ext3 fs to me | 17:35 |
| *** annegentle has quit IRC | 17:36 | |
| jbernard | dansmith: with the volume in cinder, you can export it to a file | 17:36 |
| *** claudiub has joined #openstack-nova | 17:36 | |
| jbernard | dansmith: so: rbd -p volumes export <volume-name> <file> | 17:36 |
| *** diegows has joined #openstack-nova | 17:37 | |
| *** armax has joined #openstack-nova | 17:37 | |
| jbernard | dansmith: and that should give you a file that contains the same bytes as the image in glance | 17:37 |
| jbernard | dansmith: if im understanding correctly | 17:37 |
| mriedem | dims_: you beat me to it https://review.openstack.org/#/c/160858/ :) | 17:37 |
| dansmith | well, it can't I don't think | 17:37 |
| mriedem | i just noticed that in the nova-network code | 17:37 |
| vishy | mriedem: yeah i don’t think anyone actually uses the dns stuff | 17:37 |
| dansmith | because the image in glance isn't directly bootable if it's a uec, AFAIK | 17:37 |
| *** yamahata has joined #openstack-nova | 17:37 | |
| mriedem | vishy: well, there is a fix https://review.openstack.org/#/c/160858/ | 17:37 |
| jbernard | dansmith: ahh, my uec knowledge is lacking | 17:38 |
| jbernard | dansmith: but there should be a root filesystem somewhere | 17:38 |
| jbernard | dansmith: and i can say it's certainly missing from the rbd image | 17:38 |
| vishy | fwiw it is one fo the few parts of nova network i didn’t get right | 17:38 |
| mriedem | vishy: no unit tests either :( | 17:39 |
| vishy | omg the clippy on review is genius | 17:39 |
| dansmith | jbernard: you said there was an empty ext3 filesystem right? | 17:39 |
| vishy | * i meant to say I didn’t write | 17:39 |
| dansmith | jbernard: the uec is a kernel, ramdisk, and a tarball of the rootfs, as far as I understand | 17:39 |
| mriedem | vishy: ah, i thought you were just being extremely modest :) | 17:39 |
| vishy | although it is funny that i typed that | 17:40 |
| jbernard | dansmith: right, the result of the export is a file containing an ext3 fs | 17:40 |
| mriedem | slip | 17:40 |
| vishy | multitasking for the lose | 17:40 |
| vishy | yeah right | 17:40 |
| dansmith | jbernard: and creating a qcow or raw image from it requires converting that into a real filesystem with the kernel and ramdisk inside and a bootloader in the MBR | 17:40 |
| vishy | no big ego over here | 17:40 |
| vishy | move along | 17:40 |
| vishy | :( | 17:40 |
| jbernard | dansmith: mounting it loopback, only 'lost+found' is ther | 17:40 |
| jbernard | dansmith: and the hexdump shows mostly 0's | 17:40 |
| dansmith | jbernard: surely seems like image conversion faceplanted | 17:41 |
| jbernard | dansmith: that's my thought as well | 17:41 |
| jbernard | dansmith: but if that were totally broken, wouldn't everyone be seeing this? | 17:41 |
| jbernard | dansmith: why would it only occur in tempest? | 17:41 |
| dansmith | jbernard: well, BFV | 17:41 |
| dansmith | jbernard: but yeah, my initial question was whether the image in glance was okay or not | 17:42 |
| vishy | mriedem: it looks correct to me | 17:42 |
| *** claudiub_ has joined #openstack-nova | 17:42 | |
| mriedem | vishy: yeah, just needs another test | 17:42 |
| *** neelashah has quit IRC | 17:43 | |
| dansmith | jbernard: oh dude | 17:43 |
| *** vilobhmm1 has joined #openstack-nova | 17:43 | |
| dansmith | jbernard: http://pastebin.com/kG7xVW4C | 17:43 |
| dansmith | mriedem: ^ | 17:43 |
| dansmith | I wonder if it really is a devstack issue somehow | 17:44 |
| *** vilobhmm1 has quit IRC | 17:44 | |
| jbernard | dansmith: yeah, i was just trying that :) | 17:44 |
| *** vilobhmm1 has joined #openstack-nova | 17:44 | |
| dansmith | or maybe a glance-backed-by-rbd issue? | 17:44 |
| jbernard | so cinder correctly converts and empty image | 17:44 |
| jbernard | let me export the glance rbd image | 17:44 |
| *** vilobhmm1 has quit IRC | 17:45 | |
| *** vilobhmm1 has joined #openstack-nova | 17:45 | |
| jbernard | yep, empty as well | 17:45 |
| jbernard | the glance image import could contain a problem | 17:46 |
| jbernard | or the downloaded image really is empty | 17:46 |
| *** claudiub has quit IRC | 17:46 | |
| *** vilobhmm11 has joined #openstack-nova | 17:46 | |
| dansmith | cool, this is getting farther and farther from being "my problem" :P | 17:46 |
| jbernard | lol | 17:46 |
| *** vmtrooper has joined #openstack-nova | 17:46 | |
| *** exploreshaifali has quit IRC | 17:47 | |
| *** vilobhmm11 has quit IRC | 17:47 | |
| *** mpaolino has quit IRC | 17:47 | |
| *** vilobhmm11 has joined #openstack-nova | 17:47 | |
| jbernard | the cached image is empty as well (in /opt/stack/data/glance/cache) | 17:48 |
| dims_ | mriedem: give me a few, will update with more tests. | 17:49 |
| *** ndipanov has quit IRC | 17:49 | |
| *** VW_ has joined #openstack-nova | 17:49 | |
| dansmith | hrm | 17:49 |
| *** suresh12 has joined #openstack-nova | 17:49 | |
| dansmith | jbernard: I think we're missing something | 17:49 |
| dansmith | jbernard: I think because of how cirros works, the rootfs is supposed to be empty, and everything is in the ramdisk | 17:50 |
| *** vilobhmm1 has quit IRC | 17:50 | |
| jbernard | dansmith: yes, isn't the ramdisk and kernel supposed to be combined into the boot volume? | 17:50 |
| dansmith | jbernard: not sure.. you can tell libvirt/qemu to keep them separate, but that's not happening here | 17:50 |
| dansmith | it would also be strange to do that in a BFV case I think | 17:51 |
| dansmith | we should ask a cinder person | 17:51 |
| *** vmtrooper has quit IRC | 17:51 | |
| mriedem | sdague: so that tempest change is the culprit, but it just exposed the bug in nova | 17:51 |
| mriedem | sdague: 1 line fix coming up | 17:51 |
| jbernard | dansmith: ah, i see what your saying. yeah, i think my theory falls apart | 17:52 |
| dansmith | jbernard: but what we need to know, | 17:52 |
| dansmith | is if the kernel/ramdisk should be written in there, or if nova should be using direct kernel boot and isn't | 17:53 |
| *** afazekas has quit IRC | 17:53 | |
| jbernard | dansmith: right | 17:53 |
| *** oro has joined #openstack-nova | 17:54 | |
| jbernard | dansmith: how do we do this? :) | 17:54 |
| dansmith | jbernard: asking in -cinder if you want to jump over | 17:55 |
| dansmith | not sure what will come of it | 17:55 |
| *** neelashah has joined #openstack-nova | 17:55 | |
| *** neelashah1 has joined #openstack-nova | 17:56 | |
| *** Marga_ has joined #openstack-nova | 17:56 | |
| *** hdd has joined #openstack-nova | 17:57 | |
| *** ndipanov has joined #openstack-nova | 17:58 | |
| *** unicell has quit IRC | 17:58 | |
| jbernard | dansmith: yeah, im going to retry with lvm | 17:59 |
| dansmith | jbernard: okay | 17:59 |
| *** neelashah has quit IRC | 17:59 | |
| jbernard | dansmith: in the absense of not knowing what's supposed to happen, maybe ill get lucky | 17:59 |
| dansmith | yeah | 17:59 |
| *** mpaolino has joined #openstack-nova | 18:01 | |
| *** Mike_D_laptop has joined #openstack-nova | 18:02 | |
| *** diegows has quit IRC | 18:02 | |
| *** moshele has joined #openstack-nova | 18:03 | |
| *** nkrinner has joined #openstack-nova | 18:08 | |
| *** nkrinner has quit IRC | 18:08 | |
| *** mpaolino has quit IRC | 18:11 | |
| *** sdake has joined #openstack-nova | 18:13 | |
| *** salv-orl_ has joined #openstack-nova | 18:13 | |
| openstackgerrit | Tomi Juvonen proposed openstack/nova-specs: New API to put compute node down. https://review.openstack.org/169836 | 18:14 |
| *** xyang1 has quit IRC | 18:15 | |
| *** rfolco has quit IRC | 18:16 | |
| *** ndipanov has quit IRC | 18:16 | |
| *** sdake_ has quit IRC | 18:16 | |
| *** lpetrut has joined #openstack-nova | 18:17 | |
| *** salv-orlando has quit IRC | 18:17 | |
| *** pixelb has quit IRC | 18:21 | |
| *** diegows has joined #openstack-nova | 18:21 | |
| *** ndipanov has joined #openstack-nova | 18:21 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: instance termination with update_dns_entries set fails https://review.openstack.org/160858 | 18:23 |
| *** bkopilov has joined #openstack-nova | 18:23 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova: Filter fixed IPs from requested_networks in deallocate_for_instance https://review.openstack.org/169868 | 18:23 |
| mriedem | sdague: dansmith: ^ is the fix for that fixedipnotfoundforinstance trace | 18:24 |
| *** isq has quit IRC | 18:24 | |
| *** nkrinner has joined #openstack-nova | 18:24 | |
| dansmith | mriedem: essellent | 18:25 |
| dims_ | mriedem: ^^ new test that you requested :) | 18:26 |
| mriedem | yup, looking | 18:27 |
| *** Bsony_ has joined #openstack-nova | 18:27 | |
| *** Bsony has quit IRC | 18:27 | |
| moshele | dansmith: hi | 18:29 |
| dansmith | moshele: hola | 18:29 |
| moshele | dansmith: I have some questions regarding your comments https://review.openstack.org/#/c/168317/5/nova/objects/pci_device.py | 18:30 |
| *** dboik has quit IRC | 18:30 | |
| dansmith | alright | 18:30 |
| moshele | dansmith: 1) regarding your first comment | 18:31 |
| moshele | attribute can exist in one obj but its value is None and second obj may not have this attribute which in such case I think they are equal | 18:31 |
| *** dboik has joined #openstack-nova | 18:31 | |
| *** rfolco has joined #openstack-nova | 18:31 | |
| *** ndipanov has quit IRC | 18:31 | |
| moshele | dansmith: The reason for this is to avoid a second unnecessary resources update. In the first update self.old_resource is empty and we get the new resources. In the second update the PciDevice in the old resources has some missing attributes like instance_uuid and request_id and these attributes in then new PciDevice exists but with None value. In | 18:31 |
| moshele | such case I don't think we need to update resources. | 18:31 |
| openstackgerrit | Lee Yarwood proposed openstack/nova: Fix multipath device discovery when user_friendly_names are enabled. https://review.openstack.org/169873 | 18:32 |
| *** dboik has quit IRC | 18:32 | |
| moshele | dansmith: do you agree? | 18:32 |
| *** dboik has joined #openstack-nova | 18:33 | |
| dansmith | moshele: I think that you should either (a) fix those attributes so they're always set or (b) exclude them because they're not really part of the resource update anyway | 18:33 |
| mriedem | dims_: a couple of comments inline | 18:33 |
| *** johnthetubaguy is now known as zz_johnthetubagu | 18:33 | |
| moshele | dansmith: ok I will got we (B) | 18:34 |
| *** erlon has joined #openstack-nova | 18:35 | |
| *** Bsony_ has quit IRC | 18:36 | |
| dims_ | mriedem: i'll fix the assert, want me to fix the fake_instance in test_deallocate_for_instance_passes_host_info method too? (that's the one i copied this from) | 18:36 |
| mriedem | dims_: since we want to backport this to stable/juno, let's leave other tests untouched | 18:36 |
| dims_ | ack mriedem | 18:37 |
| moshele | dansmith: regarding your second comment return getattr(a, attr) == getattr(b, attr) will not compare all the attributes ... | 18:37 |
| *** Sukhdev has joined #openstack-nova | 18:38 | |
| *** neelashah1 has quit IRC | 18:38 | |
| bauzas | dansmith: I'm surprised to see you were okay with the __eq__ method for moshele's patch | 18:38 |
| dansmith | bauzas: why? | 18:38 |
| bauzas | dansmith: because I was thinking about it for PaulMurray's patch and he told me you were against the idea | 18:39 |
| dansmith | bauzas: in mountain view, we talked about doing __eq__ for individual objects, just not at the base | 18:39 |
| dansmith | bauzas: against it for the general case | 18:39 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: instance termination with update_dns_entries set fails https://review.openstack.org/160858 | 18:39 |
| bauzas | dansmith: okay, then I missed this convo probably | 18:39 |
| dansmith | bauzas: and he was going to pull out the meat of his into a general base.foo function so others that wanted could use it in their __eq__ | 18:39 |
| dims_ | mriedem: thanks for the review. Done! | 18:39 |
| bauzas | dansmith: okay, sounds fine by me as I said | 18:39 |
| mriedem | dims_: one tiny thing left in there | 18:40 |
| mriedem | dims_: you left the unused mock | 18:40 |
| bauzas | dansmith: re: __eq__ in the base module, I don't want to open a new convo, but I was thinking that if we were not providing __hash__ it was good too | 18:41 |
| bauzas | dansmith: but I can understand that we prefer to leave this up to the direct objects | 18:41 |
| bauzas | so that's ok | 18:41 |
| dansmith | bauzas: the point of not doing it on the base class is because it's not the same for all objects, semantically | 18:41 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: instance termination with update_dns_entries set fails https://review.openstack.org/160858 | 18:41 |
| dansmith | bauzas: the same is probably true for __hash__ | 18:41 |
| dims_ | mriedem: ah. missed that comment | 18:41 |
| bauzas | dansmith: mmm, that's understandable | 18:42 |
| bauzas | dansmith: but some people (like me) could be say that if fields are the same, then 2 objects are identical - but then it goes to the _context attribute | 18:43 |
| dansmith | bauzas: but others would say "what about unset fields? should they be considered?" | 18:44 |
| bauzas | dansmith: yey it would open a can of worms | 18:44 |
| dansmith | yep | 18:44 |
| dansmith | in that convo, | 18:44 |
| bauzas | okay, I understand now better than what my better English guy said to me :) | 18:45 |
| dansmith | we had about three definitions of "equal" that weren't compatible and we didn't even try hard :) | 18:45 |
| bauzas | s/better/best (past 8pm comment) | 18:45 |
| bauzas | yeah agreed... | 18:46 |
| bauzas | ok, I'll change my point to moshele's pathc | 18:46 |
| *** emagana has quit IRC | 18:50 | |
| *** isq has joined #openstack-nova | 18:51 | |
| *** achanda has quit IRC | 18:52 | |
| *** alanf-mc has joined #openstack-nova | 18:54 | |
| *** claudiub_ has quit IRC | 18:56 | |
| *** kudryashova_ has joined #openstack-nova | 18:56 | |
| *** kudryashova_ has quit IRC | 18:56 | |
| *** bkopilov has quit IRC | 18:56 | |
| *** ekudryashova has joined #openstack-nova | 18:57 | |
| *** ekudryashova has quit IRC | 18:59 | |
| *** sbfox has joined #openstack-nova | 19:00 | |
| *** VW_ has quit IRC | 19:00 | |
| jogo | mriedem: looks like your patch may be working https://review.openstack.org/#/c/169520/1 | 19:00 |
| jogo | the latest failure http://logs.openstack.org/20/169520/1/check/check-tempest-dsvm-aiopcpu-full/6b28161/logs/testr_results.html.gz isn't live migration based | 19:01 |
| mriedem | jogo: i'm still trying to see a run where the debug logging in that loop shows pu | 19:02 |
| mriedem | *up | 19:02 |
| jogo | mriedem: ahh | 19:02 |
| jogo | looks like we have some other errors too | 19:02 |
| jogo | such as shelving | 19:02 |
| *** VW_ has joined #openstack-nova | 19:02 | |
| jogo | http://logs.openstack.org/27/169827/1/check/check-tempest-dsvm-aiopcpu-full/e4312c3/logs/testr_results.html.gz | 19:02 |
| *** iamjarvo has quit IRC | 19:02 | |
| mriedem | jogo: and test volume boot pattern | 19:03 |
| jogo | mriedem: right, is volume boot pattern failing on other jobs or just this one? | 19:04 |
| flaper87 | dansmith: I'm now | 19:04 |
| dansmith | flaper87: was trying to pin some failure on glance earlier, but you weren't around, so it turns out it's not glance :) | 19:04 |
| dansmith | at least, as far as we can tell at the moment | 19:04 |
| flaper87 | dansmith: ah ok, phew! Let me know if I can help | 19:05 |
| dansmith | thanks | 19:05 |
| mriedem | jogo: yeah, in aiocpu it's network problems in both fails | 19:06 |
| *** ijw has quit IRC | 19:06 | |
| jogo | mriedem: have you checked previous logs for Failed to undefine network filter? | 19:06 |
| jogo | mriedem: just checked the most recent one and no luck so rechecked | 19:07 |
| *** tellesnobrega_ has joined #openstack-nova | 19:07 | |
| jogo | mriedem: hmm maybe we should open up another bug for those failures to track it | 19:08 |
| mriedem | jogo: i checked the previous runs and didn't see the debug log yet | 19:08 |
| *** tellesnobrega_ has quit IRC | 19:08 | |
| mriedem | jogo: i wouldn't consider that a safe patch until we see that debug log show up | 19:08 |
| jogo | mriedem: ack | 19:08 |
| jogo | mriedem: areed | 19:08 |
| jogo | agreed | 19:08 |
| * jogo thinks it would be really neat to get aiopcpu stable for Kilo | 19:08 | |
| mriedem | you could wow mordred with that on your pbc results :) | 19:09 |
| *** emagana has joined #openstack-nova | 19:09 | |
| jogo | pbc? | 19:10 |
| mriedem | personal business commitments, it's an ibm joke | 19:10 |
| mriedem | set your goals for the year in may, writeup your results in november | 19:10 |
| jogo | ohh hehe | 19:10 |
| lbragstad | PBCs! | 19:10 |
| mriedem | and use lots of fun jargon along the way | 19:10 |
| mriedem | lbragstad has a pbc keyword sniffer as you can see | 19:10 |
| mriedem | s/may/march/ | 19:11 |
| *** lifeless has quit IRC | 19:11 | |
| mriedem | if you have a good year, your results don't matter vs your goals. if you have a bad year, the goals are used to justify the low rating. | 19:11 |
| *** neelashah has joined #openstack-nova | 19:11 | |
| *** VW_ has quit IRC | 19:12 | |
| *** iamjarvo has joined #openstack-nova | 19:12 | |
| *** iamjarvo has quit IRC | 19:12 | |
| dansmith | mriedem: lol | 19:12 |
| *** VW_ has joined #openstack-nova | 19:13 | |
| *** rushil has joined #openstack-nova | 19:13 | |
| *** iamjarvo has joined #openstack-nova | 19:13 | |
| mriedem | "your goals say you were going to complete 100% of your work items on time, but it looks like you only completed 99% so, 3 rating for you!" | 19:13 |
| sdague | mriedem: awesome +A | 19:13 |
| jogo | lol | 19:13 |
| mriedem | being over age 50 is an auto 3 rating | 19:14 |
| jogo | mriedem: new tracking bug https://bugs.launchpad.net/nova/+bug/1439355 | 19:14 |
| openstack | Launchpad bug 1439355 in OpenStack Compute (nova) "tempest.scenario.test_shelve_instance.TestShelveInstance.test_shelve_instance periodically failing on aiopcpu job" [Undecided,New] | 19:14 |
| mriedem | jogo: is it just that job? | 19:14 |
| jogo | mriedem: yup | 19:14 |
| jogo | mriedem: well at least mostly | 19:14 |
| jbernard | dansmith: ok, so the boot volume created with an lvm backend | 19:14 |
| mriedem | hmm, is the shelve test volume-backed? | 19:14 |
| jbernard | dansmith: contains a root filesystem | 19:14 |
| jbernard | dansmith: and appears to be bootable by itself | 19:15 |
| dansmith | jbernard: cool, so that's what I expected | 19:15 |
| mordred | mriedem: jogo always wows me with his PBC results. he also always gets his tps reports in on time | 19:15 |
| dims_ | mriedem: want me to rebase https://review.openstack.org/#/c/160858/ on top of your review? | 19:15 |
| *** HotTopic has joined #openstack-nova | 19:16 | |
| mriedem | dims_: that might save you the rebase | 19:16 |
| dims_ | mriedem: ah, it's already +A'ed :) so i definitely will | 19:17 |
| *** HotTopic has left #openstack-nova | 19:17 | |
| jogo | mriedem: don't know if this is related but: http://logs.openstack.org/27/169827/1/check/check-tempest-dsvm-aiopcpu-full/e4312c3/logs/screen-n-cpu.txt.gz?level=TRACE | 19:18 |
| jogo | that is a lot of stacktraces | 19:18 |
| mriedem | jogo: already fixed | 19:18 |
| mriedem | jogo: see the epic commit message https://review.openstack.org/#/c/169868/ | 19:18 |
| jogo | nice | 19:19 |
| *** rushiagr is now known as rushiagr_away | 19:21 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: instance termination with update_dns_entries set fails https://review.openstack.org/160858 | 19:21 |
| *** achanda has joined #openstack-nova | 19:23 | |
| mriedem | jogo: sdague: dansmith: is also an easy fix https://review.openstack.org/#/c/160858/ | 19:23 |
| *** Marga_ has quit IRC | 19:24 | |
| casusbelli | Hi! Can somebody pls give me some hints: I'm debugging a devstack setup where VMs come up in state ACTIVE but power state PAUSED ?? Hints on what could cause this? | 19:24 |
| mriedem | casusbelli: have you grepped the nova compute logs for the instance uuid? | 19:24 |
| dims_ | jogo: one oslo context related fix - https://review.openstack.org/#/c/168153/ when you get a chance please. | 19:25 |
| casusbelli | mriedem: no errors no stack trces | 19:25 |
| casusbelli | s/trces/traces/ | 19:25 |
| mriedem | casusbelli: not for errors, but to see what it's doing | 19:25 |
| mriedem | debug trace | 19:25 |
| casusbelli | mriedem: gave it a quick run throught but go down more detailed, mom | 19:25 |
| *** nkrinner has quit IRC | 19:28 | |
| *** alanf-mc has quit IRC | 19:28 | |
| jogo | mriedem: hmm WARN: /etc/rc3.d/S40-network failed | 19:28 |
| *** doude has quit IRC | 19:29 | |
| *** doude_ has joined #openstack-nova | 19:29 | |
| mriedem | https://www.rdoproject.org/forum/discussion/305/warn-etcrc3-ds40network-failed/p1 | 19:29 |
| *** doude_ has quit IRC | 19:29 | |
| casusbelli | mriedem: This will take a while, too much output, sry. Will do new run to get a smaller log... | 19:29 |
| *** doude has joined #openstack-nova | 19:30 | |
| mriedem | jogo: maybe related https://ask.openstack.org/en/question/2053/unable-to-ping-my-instance-after-creating-it/ | 19:30 |
| *** alanf-mc has joined #openstack-nova | 19:31 | |
| *** matrohon has joined #openstack-nova | 19:32 | |
| mriedem | jogo: it's not just that job, http://goo.gl/0DHjVP - but it is 100% failure when it hits | 19:32 |
| jogo | mriedem: because that only gets printed if tempest fails | 19:34 |
| jogo | anyway lunch time, will pick this up afterwards | 19:35 |
| mriedem | you and your logic and reasons | 19:35 |
| jogo | I wonder if vishy has seen this | 19:35 |
| *** vmtrooper has joined #openstack-nova | 19:35 | |
| vishy | seen what? | 19:35 |
| dansmith | sdague: just to be clear, you'd rather us go the devstack/tempest route to using the right image before we get that test green again, instead of putting it back to skip status so that we catch other fails, right? | 19:36 |
| *** VW_ has quit IRC | 19:36 | |
| dansmith | since it's been unskipped for only two days, I figured I'd ask | 19:36 |
| vishy | jogo: ?? | 19:37 |
| sdague | dansmith: correct, I'd much rather have lvm actually tested there | 19:37 |
| sdague | because actually testing we can boot from volume in the lvm case ensures a bunch of stuff is working | 19:37 |
| mriedem | vishy: http://logs.openstack.org/27/169827/1/check/check-tempest-dsvm-aiopcpu-full/e4312c3/console.html#_2015-04-01_17_07_53_895 | 19:37 |
| dansmith | okay | 19:37 |
| sdague | mtreinish: now that we have idempotent ids in tempest, have you considered a skip_tests in tempest that could include idempotent ids | 19:38 |
| sdague | for a case like this | 19:38 |
| dansmith | what does that mean? | 19:38 |
| sdague | dansmith: every test now has a uuid | 19:38 |
| sdague | for defcore tracking | 19:38 |
| vishy | have not seen that before | 19:38 |
| vishy | it is odd that metadata fails 6 times before working | 19:39 |
| sdague | so tempest could probably more easily allow for skipping tests by uuid | 19:39 |
| mtreinish | sdague: did jogo's unskip break things? | 19:39 |
| dansmith | sdague: we'd still need it skipped per job | 19:39 |
| sdague | dansmith: yeh, you'd pass that through in the job | 19:39 |
| dansmith | okay | 19:39 |
| mtreinish | sdague: so the uuid becomes part of the test_id so we can exclude it based on that easily | 19:39 |
| sdague | mtreinish: well, it exposed bugs in the ceph path | 19:39 |
| mtreinish | sdague: in os-testr I have a mechanism for an exclude file | 19:40 |
| *** lifeless has joined #openstack-nova | 19:40 | |
| sdague | mtreinish: right, we could change that decorator in tempest to also emit a skip | 19:40 |
| mtreinish | so when we switch to that we can exclude running tests by a file of uuids somewhere | 19:40 |
| *** htruta has quit IRC | 19:40 | |
| sdague | yeh, well you could do it independent of test runner | 19:40 |
| *** vmtrooper has quit IRC | 19:40 | |
| sdague | which would be kind of nice | 19:40 |
| mtreinish | sdague: you want the skip decorator to check against a list of uuids and skip based on that? | 19:40 |
| sdague | mtreinish: yeh | 19:40 |
| mtreinish | sdague: and where would the list be kept? | 19:41 |
| sdague | well, it's an idea | 19:41 |
| sdague | in tempest.conf | 19:41 |
| mtreinish | I guess we could do that, although normally we've said that's a thing for the test runner to do | 19:41 |
| casusbelli | mriedem: This took a while. Paste: http://paste.openstack.org/show/197822/ The VM seems to spawn and then is '...paused unexpectedly. Ignore.' . I'm not sure what to make of this. | 19:42 |
| sdague | yeh, it's an idea, I'm not convinced it's a fully formed one | 19:42 |
| sdague | but it would allow for what we are seeing in that tempest skips are based on a reference config | 19:42 |
| mriedem | casusbelli: libvirt is sending an event saying the instance is paused | 19:42 |
| casusbelli | mriedem: After that paused unexp. it goes into a loop 1/min and repeats that message from time to time (about 1min) | 19:42 |
| sdague | and it would allow getting rid of the regex | 19:42 |
| dansmith | sdague: so is getting a disk image installed something I should do in devstack or somewhere more specific like d-g or whatever? | 19:44 |
| mtreinish | sdague: yeah I see where you're coming from, I'm just wondering if we have an exclude file in how we invoke the test run whether we really need to do it tempest too | 19:44 |
| dansmith | sdague: kinda seems like we might want the DEFAULT_IMAGE_NAME to be set properly, which maybe implies something that knows what kind of job we're doing | 19:44 |
| mtreinish | sdague: btw I can start working on using ostestr in tempest, I just need https://review.openstack.org/#/c/167243/ +A'd | 19:46 |
| sdague | dansmith: so I think you should figure out how to add the .img to the IMAGE_URLS list - https://github.com/openstack-dev/devstack/blob/master/stackrc#L543 | 19:48 |
| sdague | I think that will automatically get it into glance | 19:48 |
| dansmith | sdague: okay, that's the direction I was headed, | 19:48 |
| dansmith | but if so, | 19:48 |
| sdague | then it's just a matter of getting the right things set in lib/tempest | 19:48 |
| dansmith | we'll need to just set the default in the job I guess? | 19:48 |
| dansmith | okay | 19:49 |
| dansmith | I didn't think lib/tempest would have the data to know which it should choose | 19:49 |
| *** Nikolay_St has quit IRC | 19:49 | |
| *** Nikolay_St has joined #openstack-nova | 19:50 | |
| sdague | dansmith: it has to know what image to use for tests | 19:50 |
| sdague | so it's in there somewhere | 19:50 |
| sdague | https://github.com/openstack-infra/devstack-gate/blob/master/devstack-vm-gate.sh#L370-L372 - we have a generic passthrough mechanism that can be used in jobs | 19:50 |
| mtreinish | yeah there is an image ref option that is set there | 19:51 |
| dansmith | sdague: what I mean is, how is it going to know if it should choose the uec (for normal) or img (for tempest | 19:51 |
| sdague | so the thing to do is figure out if you can do it entirely with localrc | 19:51 |
| dansmith | s/tempest/ceph | 19:51 |
| sdague | you'd set it in the job | 19:51 |
| dansmith | right, okay, I thought you were saying I could do it all in tempest | 19:51 |
| sdague | with DEVSTACK_LOCAL_CONFIG | 19:51 |
| sdague | dansmith: oh, sorry | 19:51 |
| dansmith | just setting DEFAULT_IMAGE_NAME in the job to the image one would be enough I think, | 19:52 |
| sdague | yeh, so yuo *might* have enough knobs in devstack to just use DEVSTACK_LOCAL_CONFIG to get you home | 19:52 |
| dansmith | as long as it gets into the IMAGE_URLS | 19:52 |
| dansmith | I think I can | 19:52 |
| sdague | ok, yeh, I think that's the right approach to try | 19:52 |
| dansmith | okay cool | 19:52 |
| *** alanf-mc has quit IRC | 19:53 | |
| bauzas | alaski: I will be able to attend today's cells meeting, could you just bug me when you're opening it ? | 19:54 |
| casusbelli | mriedem: think i found something, there's a type in the nova libvirt driver setting. :-/ | 19:54 |
| alaski | bauzas: sure. writing up some specs now to bring up | 19:54 |
| casusbelli | mriedem: This smells like stupid, grrh | 19:54 |
| bauzas | alaski: cool, sounds like I'll have homework | 19:55 |
| mriedem | i love stupid smells | 19:55 |
| casusbelli | mriedem: more 'types': s/type/typo/ I should have more sleep, typing weird stuff in the configs, as it seems.... :) | 19:56 |
| mriedem | casusbelli: so you mean you made a typo in nova.conf? | 19:57 |
| casusbelli | mriedem: sigh, yep | 19:57 |
| *** Marga_ has joined #openstack-nova | 19:57 | |
| casusbelli | mriedem: in the libvirt driver settings, so he's not using my nova driver which in turn means he is most likely unable to connect to the cinder volume.... | 19:57 |
| casusbelli | mriedem: I'm currently re-stacking... | 19:58 |
| *** VW_ has joined #openstack-nova | 19:58 | |
| *** alanf-mc has joined #openstack-nova | 19:59 | |
| *** annashen has quit IRC | 20:00 | |
| *** alanf-mc has quit IRC | 20:01 | |
| dansmith | jbernard: got the bug filed yet? | 20:04 |
| jbernard | dansmith: about to click 'submit' | 20:04 |
| openstackgerrit | Andrew Laski proposed openstack/nova-specs: Proposal for cells v2 scheduling https://review.openstack.org/141486 | 20:04 |
| openstackgerrit | Andrew Laski proposed openstack/nova-specs: Cells: Store build request information before scheduling https://review.openstack.org/169901 | 20:04 |
| *** alanf-mc has joined #openstack-nova | 20:04 | |
| dansmith | jbernard: cool, just want to mention it in a commit message | 20:04 |
| jbernard | dansmith: https://bugs.launchpad.net/cinder/+bug/1439371 | 20:04 |
| openstack | Launchpad bug 1439371 in Cinder "Volume creation from image fails for UEC+Ceph" [Undecided,New] - Assigned to Jon Bernard (jbernard) | 20:04 |
| dansmith | jbernard: thanks | 20:04 |
| jbernard | dansmith: np | 20:06 |
| dansmith | jbernard: sdague: https://review.openstack.org/#/c/169903/1/jenkins/jobs/devstack-gate.yaml,cm | 20:06 |
| *** annegent_ has joined #openstack-nova | 20:06 | |
| *** busterswt has quit IRC | 20:07 | |
| dansmith | sdague: hmm, that's not going to run much. can I get a ceph run against that patch somehow? | 20:08 |
| *** tjones1 has quit IRC | 20:08 | |
| mtreinish | dansmith: I don't think you can short of merging it blindly | 20:09 |
| dansmith | mtreinish: gah | 20:09 |
| casusbelli | mriedem: Hmm, issue is staying, typo was not the reason. My network config seems to be broken? the last steps prior to pause are lazy loading flavor, updated the network info cache for instance and then it is paused in the next msg. | 20:09 |
| mtreinish | and spinning up a run to test after it gets picked up | 20:09 |
| dansmith | mtreinish: well, at least it's a non-voting job now, so impact should be minimal | 20:10 |
| dansmith | mtreinish: would a depends-on change to d-g pick it up? | 20:10 |
| mtreinish | dansmith: oh, you know you can do a devstack-gate job which exports those vars the same way, that'll test it | 20:10 |
| dansmith | mtreinish: you mean a throw-away patch that just hacks them into place? | 20:10 |
| mtreinish | dansmith: no, because it needs to pick up the change in job definition for it to be run | 20:10 |
| mtreinish | dansmith: yeah | 20:10 |
| *** moshele has quit IRC | 20:11 | |
| mtreinish | and if the ceph job doesn't run on d-g you'd need to have a nova patch (or whatever) depend on the throwaway d-g patch | 20:11 |
| dansmith | hah | 20:11 |
| dansmith | mtreinish: does that stuff go into some sort of localrc for d-g? | 20:14 |
| *** alanf-mc has quit IRC | 20:16 | |
| *** alanf-mc has joined #openstack-nova | 20:16 | |
| *** vilobhmm11 has quit IRC | 20:16 | |
| *** vilobhmm1 has joined #openstack-nova | 20:16 | |
| *** vilobhmm1 has quit IRC | 20:17 | |
| *** emagana has quit IRC | 20:17 | |
| *** vilobhmm1 has joined #openstack-nova | 20:17 | |
| *** vilobhmm1 has quit IRC | 20:18 | |
| *** vilobhmm1 has joined #openstack-nova | 20:18 | |
| *** vilobhmm1 has quit IRC | 20:18 | |
| *** vilobhmm1 has joined #openstack-nova | 20:18 | |
| *** vladikr has quit IRC | 20:19 | |
| openstackgerrit | Dan Smith proposed openstack/nova: April fools! https://review.openstack.org/169909 | 20:20 |
| dansmith | ah dammit | 20:21 |
| openstackgerrit | Dan Smith proposed openstack/nova: April fools! https://review.openstack.org/169909 | 20:22 |
| *** suresh12 has quit IRC | 20:24 | |
| *** ZZelle has joined #openstack-nova | 20:26 | |
| *** tjones1 has joined #openstack-nova | 20:26 | |
| *** vilobhmm1 has quit IRC | 20:27 | |
| mtreinish | dansmith: well you could do it that way in devstack-gate.sh it basically echo the localrc into a file | 20:28 |
| *** VW__ has joined #openstack-nova | 20:28 | |
| dansmith | mtreinish: did this: https://review.openstack.org/#/c/169907/ | 20:28 |
| mtreinish | dansmith: but you could just export the lines like you do in the project-config patch at the beginning of devstack-gate-wrap.sh | 20:28 |
| dansmith | oh shit | 20:28 |
| dansmith | mine isn't going to work | 20:28 |
| dansmith | because I put it into the if | 20:29 |
| *** neelashah has quit IRC | 20:29 | |
| *** claudiub has joined #openstack-nova | 20:29 | |
| mtreinish | dansmith: heh, yeah I was just going to say | 20:29 |
| jogo | sdague: which jobs is it failing on? | 20:29 |
| dansmith | jogo: ceph jobs | 20:30 |
| jogo | dansmith: lol, is it because ceph has a bug or it doesn't support something? | 20:30 |
| *** emagana has joined #openstack-nova | 20:30 | |
| *** rfolco has quit IRC | 20:31 | |
| dansmith | jogo: it's because the cinder ceph driver is broken for uec images | 20:31 |
| jogo | dansmith: ohh interesting | 20:31 |
| *** annegent_ has quit IRC | 20:31 | |
| mtreinish | jogo: is it really? | 20:31 |
| mtreinish | :) | 20:31 |
| dansmith | heh | 20:31 |
| *** VW_ has quit IRC | 20:32 | |
| *** lpetrut has quit IRC | 20:32 | |
| jogo | mtreinish: yes in as I didn't expect ceph to fail for specific image types | 20:32 |
| jogo | but not interesting enough for me to look into it atall | 20:32 |
| dansmith | we're on it anyway | 20:32 |
| jogo | FWIW I tend to agree with dansmith on that making sure ceph works is important for upstream | 20:33 |
| dansmith | jogo: for the moment -> https://review.openstack.org/#/c/169903/ | 20:33 |
| *** neelashah has joined #openstack-nova | 20:33 | |
| jbernard | dansmith: thanks for that one, btw | 20:33 |
| openstackgerrit | Ed Leafe proposed openstack/nova: Optimize periodic call to get_by_host https://review.openstack.org/165849 | 20:33 |
| jbernard | dansmith: that will give me some time to get things sorted | 20:33 |
| *** adalbas has quit IRC | 20:34 | |
| dansmith | jbernard: np | 20:34 |
| dansmith | jogo: sounds like jbernard will fix the cinder driver, and hopefully get that into L | 20:34 |
| dansmith | but I'd like to see the ceph job go voting sooner than that to make sure we don't regress further/again | 20:35 |
| jbernard | dansmith: it's tecnically a bugfix, so the window on K isn't quite closed (i think) | 20:35 |
| jbernard | dansmith: but i cant promise anything | 20:35 |
| *** suresh12 has joined #openstack-nova | 20:35 | |
| jbernard | dansmith: though, ill try my best | 20:35 |
| *** annashen has joined #openstack-nova | 20:35 | |
| dansmith | jbernard: well, I'm expecting it's going to be a little heavy for a fix at this point, but... sure :) | 20:35 |
| jogo | dansmith: makes sense | 20:37 |
| jogo | dansmith: so crazy idea for L | 20:37 |
| jogo | make aiopcpu use ceph | 20:37 |
| bauzas | sdague: why couldn't we fencing for Neutron giving us -2? | 20:37 |
| jogo | and make that the default job ? | 20:37 |
| bauzas | re: https://review.openstack.org/#/c/169605/ | 20:37 |
| *** Sukhdev has quit IRC | 20:37 | |
| dansmith | jogo: I think that would be good at some point, but maybe we wait a bit until aiopcpu is proven stable/ | 20:37 |
| dansmith | jogo: I expect we'll shake out a new color of bugs with that for a while | 20:38 |
| sdague | bauzas: because that seems like total craziness out of the neutron api | 20:38 |
| bauzas | sdague: preventing bugs from other projets going to Nova sounds like okay to me | 20:38 |
| sdague | like actual clown car craziness | 20:38 |
| jogo | dansmith: yeah at best this would be 2 to 3 months out | 20:38 |
| dansmith | jogo: yeah | 20:38 |
| dansmith | sdague: what is "actual clown car craziness" ? :P | 20:38 |
| bauzas | sdague: agreed on the point that Neutron could perhaps give us something bad | 20:38 |
| sdague | bauzas: actually fixing those projects instead of providing layers of work arounds is much better | 20:38 |
| *** vilobhmm1 has joined #openstack-nova | 20:39 | |
| * jogo goes back to debugging nova-net | 20:39 | |
| bauzas | sdague: yeah, but I honestly thing the bugfix was worth it because it was preventing something bad from Neutron | 20:39 |
| bauzas | for a very little impact | 20:39 |
| *** fawadkhaliq has quit IRC | 20:39 | |
| bauzas | by saying "eh, that's Neutron fault", we leave the bug while providing any workaround | 20:40 |
| bauzas | while *not | 20:40 |
| mriedem | dansmith: literal clown car | 20:40 |
| dansmith | bauzas: but we can actually fix neutron, right? | 20:41 |
| bauzas | sdague: and saying that it should be deferred to Neutron makes the bugfix merged probably only for Libverty | 20:41 |
| *** VW__ has quit IRC | 20:41 | |
| bauzas | dansmith: sure, if Neutron is wrong, Neutron has to fix it | 20:41 |
| dansmith | bauzas: no, *you* can go fix it :) | 20:41 |
| dansmith | is this in some way hard to fix in neutron? | 20:41 |
| dansmith | seems like a one-liner that could be done in less time than discussing it, no? | 20:42 |
| bauzas | dansmith: but if the workaround is worth it (https://review.openstack.org/#/c/169605/) I don't see much impact of merging that change which could help operators for Kilo | 20:42 |
| dansmith | bauzas: kilo isn't out yet | 20:42 |
| bauzas | dansmith: honestly, I can't bet on the level of change necessary | 20:42 |
| bauzas | in Neutron I mean | 20:42 |
| dansmith | mestery: around for a question? | 20:42 |
| mestery | danieru: you know it :) | 20:42 |
| dansmith | mestery: https://review.openstack.org/#/c/169605/ | 20:43 |
| mestery | dansmith: You know it too | 20:43 |
| * mestery looks | 20:43 | |
| *** melwitt has joined #openstack-nova | 20:43 | |
| dansmith | mestery: any reason that's not a trivial one-liner in neutron to fix? | 20:43 |
| mestery | dansmith: It seems like it should be, yes. salv-orl_ : ^^^^ | 20:43 |
| *** eglynn has quit IRC | 20:43 | |
| bauzas | cool then | 20:44 |
| *** VW_ has joined #openstack-nova | 20:44 | |
| *** mrda-away is now known as mrda | 20:44 | |
| dansmith | mestery: thanks | 20:45 |
| salv-orl_ | mestery: yeah that is ;) it's part of my blueprint that we deferred, but I guess this should go in as a simple bug fix | 20:45 |
| mestery | salv-orl_: That's what I was thinking too :) | 20:45 |
| bauzas | nice, so we can just -1 it for Nova | 20:45 |
| salv-orl_ | I'll try to fix it before bed time. | 20:45 |
| bauzas | awesome | 20:45 |
| bauzas | salv-orl_: ty | 20:46 |
| *** tonytan4ever has quit IRC | 20:46 | |
| *** tonytan4ever has joined #openstack-nova | 20:47 | |
| *** VW_ has quit IRC | 20:47 | |
| *** VW_ has joined #openstack-nova | 20:48 | |
| sdague | bauzas: right, that's why I was -1, neutron wasn't even added to the bug report, there was no attempt by anyone to fix the root cause, which looked like it would be quick. | 20:48 |
| sdague | salv-orl_: thanks much for running with it | 20:48 |
| bauzas | sdague: yeah, you're right | 20:48 |
| bauzas | sdague: it should at least have been tagged for Neutron | 20:49 |
| bauzas | speaking of the bug report of course | 20:49 |
| *** VW_ has quit IRC | 20:51 | |
| *** sdake_ has joined #openstack-nova | 20:52 | |
| *** flwang1 has joined #openstack-nova | 20:54 | |
| dansmith | sdague: https://jenkins02.openstack.org/job/gate-project-config-layout/624/console | 20:54 |
| dansmith | sdague: it doesn't like my variables I think | 20:54 |
| *** ferest has joined #openstack-nova | 20:54 | |
| dansmith | maybe it thinks anything in {} is a yaml variable? | 20:55 |
| openstackgerrit | Andrew Laski proposed openstack/nova-specs: Cells instance migration https://review.openstack.org/136490 | 20:55 |
| *** sdake has quit IRC | 20:55 | |
| mikal | Morning | 20:57 |
| *** bkopilov has joined #openstack-nova | 20:57 | |
| *** dprince has quit IRC | 20:57 | |
| sdague | dansmith: hmmm... don't know, and I have to run in 5. I would ask clarkb or jeblair about how to pull that off. | 20:57 |
| mriedem | mikal: hi! | 20:57 |
| dansmith | sdague: okay | 20:57 |
| mriedem | mikal: how was camping? | 20:57 |
| dansmith | don't you mean the wiggles "concert" ? | 20:57 |
| sdague | dansmith: devstack-virtual-ironic seems to have it | 20:58 |
| mikal | mriedem: twas good, if cold. | 20:58 |
| mriedem | really? | 20:58 |
| mikal | dansmith: that was a separate even | 20:58 |
| sdague | you may need to use " " strings and escape the $ a bunch | 20:58 |
| mikal | event even | 20:58 |
| mriedem | are the wiggles still big in australia? | 20:58 |
| mikal | mriedem: yeah, they're a $45 million a year business | 20:58 |
| *** rushil has quit IRC | 20:59 | |
| bauzas | alaski: cells meeting in 2 mins, right ? | 20:59 |
| *** ferest has quit IRC | 20:59 | |
| alaski | bauzas: yes | 20:59 |
| bauzas | help me save from watching Grey's Anatomy | 20:59 |
| *** VW_ has joined #openstack-nova | 20:59 | |
| *** emagana has quit IRC | 21:00 | |
| sdague | mikal: ok, I have flipped the auto-expire bit since you chimed in | 21:00 |
| mikal | sdague: good man | 21:00 |
| *** emagana has joined #openstack-nova | 21:00 | |
| mikal | Do you know if that will auto-expire stuff that is really old? | 21:00 |
| mikal | Or only new things entering that state? | 21:00 |
| *** sbfox has quit IRC | 21:00 | |
| sdague | auto expire expires incomplete bugs without response by author after 60 days, iirc | 21:01 |
| sdague | unless they have an owner or are targetted to a milestone | 21:01 |
| mikal | sdague: sure, but is that only on entering that state? Or will the first run capture all the really old ones? | 21:01 |
| sdague | that I don't know | 21:01 |
| sdague | it must be a sweep | 21:01 |
| mikal | Well, we're about to find out! | 21:01 |
| sdague | because it's about bugs that have been Incomplete for > 60 days | 21:02 |
| sdague | our Incomplete bug count had grown to about 250 | 21:02 |
| *** Nikolay_St has quit IRC | 21:03 | |
| sdague | anyway, need to run to lug meeting here, night all | 21:03 |
| mikal | sdague: good luck, don't let the nerds lick you | 21:03 |
| cfriesen | odd question...if I have a host aggregate that specifies an availability zone, is there a way to use "nova aggregate-update" to remove the availability zone from the aggregate? | 21:04 |
| *** emagana_ has joined #openstack-nova | 21:04 | |
| *** emagana has quit IRC | 21:04 | |
| *** achanda has quit IRC | 21:06 | |
| *** oro has quit IRC | 21:07 | |
| *** thangp has quit IRC | 21:08 | |
| *** doude has quit IRC | 21:09 | |
| *** aysyd has quit IRC | 21:09 | |
| *** devlaps1 has joined #openstack-nova | 21:10 | |
| *** devlaps has quit IRC | 21:10 | |
| *** diegows has quit IRC | 21:10 | |
| *** doude has joined #openstack-nova | 21:10 | |
| melwitt | bauzas: in my patch I used the real context, if it helps https://review.openstack.org/#/c/169165/ | 21:12 |
| bauzas | melwitt: yeah, I saw that when reviewing | 21:12 |
| melwitt | bauzas: oh, cool | 21:12 |
| bauzas | melwitt: the fix is super easy but I wanted to avoid that because it was modifying the signature, hence overcomplicating the review | 21:13 |
| mriedem | jogo: should the aiocpu job be running on stable? | 21:13 |
| bauzas | melwitt: but it was a fair point | 21:13 |
| mriedem | probably not right? | 21:13 |
| melwitt | bauzas: I understand | 21:13 |
| jogo | mriedem: why not? | 21:13 |
| jogo | mriedem: prob not | 21:13 |
| jogo | mriedem: I saw its failing on juno | 21:13 |
| mriedem | jogo: i thought it required a bunch of changes to devstack on master | 21:13 |
| *** doude has quit IRC | 21:13 | |
| mriedem | yeah | 21:13 |
| * jogo pushes a patch | 21:13 | |
| mriedem | cool | 21:13 |
| *** Marga_ has quit IRC | 21:14 | |
| *** sdake_ has quit IRC | 21:15 | |
| *** Marga_ has joined #openstack-nova | 21:15 | |
| melwitt | bauzas: unrelated, I'm seeing the paperclip on gerrit now too | 21:15 |
| jogo | mriedem: have a example if it failing on juno on hand? | 21:15 |
| mriedem | jogo: yeah, sec | 21:15 |
| mriedem | jogo: http://logs.openstack.org/46/168646/1/check/check-tempest-dsvm-aiopcpu-full/ca47536/ | 21:16 |
| mriedem | http://logs.openstack.org/46/168646/1/check/check-tempest-dsvm-aiopcpu-full/ca47536/console.html#_2015-04-01_12_08_59_572 | 21:16 |
| jogo | hmm it fails on tempest | 21:16 |
| jogo | so not a devstack issue | 21:16 |
| jogo | anyway https://review.openstack.org/169920 | 21:17 |
| mriedem | jogo: well did you have some devstack changes on master? are those all on stable? | 21:17 |
| *** achanda has joined #openstack-nova | 21:18 | |
| jogo | mriedem: don't remember actually :/ | 21:19 |
| mriedem | you'll never get a pbc 1 rating that way | 21:19 |
| *** dims_ has quit IRC | 21:19 | |
| mriedem | jogo: btw, gate-grenade-dsvm-partial-ncpu is spiking http://jogo.github.io/gate/ | 21:19 |
| mriedem | i'm seeing build timeouts | 21:19 |
| mriedem | http://logs.openstack.org/54/164754/1/gate/gate-grenade-dsvm-partial-ncpu/aeb991f/console.html | 21:20 |
| bauzas | melwitt: apologies for seeing clippy, that guy is very noisy | 21:20 |
| jogo | mriedem: no devstack changes needed | 21:20 |
| jogo | only devstack-gate | 21:20 |
| mriedem | hmm http://logs.openstack.org/54/164754/1/gate/gate-grenade-dsvm-partial-ncpu/aeb991f/logs/old/screen-n-cpu.txt.gz?level=TRACE#_2015-04-01_18_22_47_681 | 21:21 |
| mriedem | aren't we supposed to not be running stuff while doing the upgrade? | 21:21 |
| *** dboik_ has joined #openstack-nova | 21:22 | |
| mriedem | i guess that's normal | 21:22 |
| dansmith | n-cpu should be running | 21:22 |
| dansmith | and complaining a lot | 21:22 |
| mriedem | yeah it is | 21:22 |
| jogo | right | 21:22 |
| mriedem | i meant not running tempest jobs while upgrading conductor | 21:22 |
| jogo | large ops spiked too | 21:23 |
| *** tjones1 has quit IRC | 21:23 | |
| jogo | err wrong blue | 21:24 |
| *** vmtrooper has joined #openstack-nova | 21:24 | |
| *** sbfox has joined #openstack-nova | 21:24 | |
| jogo | aha http://logs.openstack.org/54/164754/1/gate/gate-grenade-dsvm-partial-ncpu/aeb991f/logs/new/screen-n-cond.txt.gz#_2015-04-01_18_27_53_327 | 21:25 |
| jogo | everyone knows 2 values is not enough to unpack | 21:25 |
| *** dboik has quit IRC | 21:25 | |
| jogo | dansmith mriedem:^ | 21:25 |
| mriedem | oh snap | 21:25 |
| mriedem | yeah, so that's probably due to the same thing that blew up the deallocate_for_instance fixed_ips thing today | 21:26 |
| *** hdd has quit IRC | 21:26 | |
| *** dboik_ has quit IRC | 21:26 | |
| *** hdd has joined #openstack-nova | 21:26 | |
| mriedem | http://goo.gl/zvLqo7 | 21:26 |
| jogo | this was working as of 20 minutes ago or so | 21:27 |
| mriedem | it looks to be sporadic in logstash | 21:27 |
| *** boris-42 has quit IRC | 21:28 | |
| jogo | no hits in last hour or so | 21:28 |
| *** whenry has joined #openstack-nova | 21:28 | |
| mriedem | there were no hits on 3/30 either, it's just gappy | 21:29 |
| mriedem | but it's there | 21:29 |
| mriedem | https://bugs.launchpad.net/nova/+bug/1439402 | 21:29 |
| openstack | Launchpad bug 1439402 in OpenStack Compute (nova) "gate-grenade-dsvm-partial-ncpu fails with "ValueError: need more than 2 values to unpack" on requested_networks in conductor" [Undecided,New] | 21:29 |
| dansmith | mriedem: we just merged a patch to fix that | 21:29 |
| mriedem | dansmith: which one? | 21:29 |
| dansmith | those should all be icehouse->juno upgrades | 21:29 |
| *** vmtrooper has quit IRC | 21:29 | |
| *** cbader has quit IRC | 21:30 | |
| dansmith | I don't think it's merged in juno yet | 21:30 |
| jogo | dansmith: yup that is all icehouse -> juno | 21:30 |
| dansmith | https://review.openstack.org/#/c/169538/ | 21:30 |
| dansmith | merged in master | 21:30 |
| jogo | no hits for this bug on master | 21:30 |
| dansmith | you wouldn't hit it on master | 21:30 |
| dansmith | and, it required a devstack force merge first | 21:31 |
| mriedem | ah i see, and that same tempest change triggered it | 21:31 |
| *** gtt116_ has joined #openstack-nova | 21:31 | |
| *** gtt116_ has left #openstack-nova | 21:32 | |
| *** tjones1 has joined #openstack-nova | 21:33 | |
| *** whenry has quit IRC | 21:34 | |
| *** mwagner_lap has quit IRC | 21:36 | |
| *** sdake has joined #openstack-nova | 21:37 | |
| *** Mike_D_laptop has quit IRC | 21:37 | |
| *** tonytan_brb has joined #openstack-nova | 21:40 | |
| openstackgerrit | Dan Smith proposed openstack/nova: April fools! https://review.openstack.org/169909 | 21:40 |
| *** sdake_ has joined #openstack-nova | 21:41 | |
| jogo | lol | 21:41 |
| *** burt has quit IRC | 21:43 | |
| *** tonytan4ever has quit IRC | 21:44 | |
| *** sdake has quit IRC | 21:45 | |
| *** oro has joined #openstack-nova | 21:45 | |
| *** sdake has joined #openstack-nova | 21:45 | |
| *** xyang1 has joined #openstack-nova | 21:47 | |
| *** marun has quit IRC | 21:49 | |
| *** sdake_ has quit IRC | 21:49 | |
| * tonyb was expect soo much more from that review :/ | 21:49 | |
| *** asettle has joined #openstack-nova | 21:51 | |
| mtreinish | dansmith: you should have added a picture of yourself wearing a fake mustache to the nova tree for that review:) | 21:51 |
| jogo | I have a nova request ID that I cannot find the n-api log for | 21:51 |
| *** mriedem is now known as mriedem_away | 21:51 | |
| mtreinish | jogo: link? | 21:51 |
| dansmith | mtreinish: or maybe, jaypipes in a pink beret | 21:51 |
| mtreinish | heh, that works too | 21:51 |
| dansmith | I actually have one of those pictures | 21:51 |
| *** tonytan_brb has quit IRC | 21:52 | |
| mriedem_away | https://lh5.googleusercontent.com/-CmDk1YhICWs/VF8adz4vRNI/AAAAAAAACmA/rMfzfqsjSFg/w703-h937-n-o/20141107_210403.jpg | 21:52 |
| mriedem_away | as do i | 21:52 |
| dansmith | heh | 21:52 |
| dansmith | yeah, mine is from a different angle | 21:52 |
| dansmith | glad we have that covered | 21:53 |
| mriedem_away | indeed | 21:53 |
| mriedem_away | too bad he's not here | 21:53 |
| dansmith | I'll put it on twitter | 21:53 |
| jogo | any idea what is causing http://logs.openstack.org/73/169873/1/check/check-tempest-dsvm-aiopcpu-full/ea5db7b/logs/screen-n-net.txt.gz#_2015-04-01_19_33_06_091 | 21:54 |
| jogo | the instance in question is not even on that server | 21:54 |
| *** mjturek1 has quit IRC | 21:54 | |
| mriedem_away | jogo: that should be refresh nw info cache | 21:54 |
| *** kaufer has quit IRC | 21:55 | |
| mriedem_away | jogo: maybe the _heal_instance_info_cache periodic task? | 21:55 |
| dansmith | mriedem_away: https://twitter.com/get_offmylawn/status/583387061790515200 | 21:55 |
| mriedem_away | s/needs/needs to/. | 21:55 |
| *** ferest has joined #openstack-nova | 21:55 | |
| dansmith | shit | 21:55 |
| mriedem_away | i would like and retweet if i could | 21:56 |
| jogo | unrelated bug http://logs.openstack.org/73/169873/1/check/check-tempest-dsvm-aiopcpu-full/ea5db7b/logs/screen-n-cpu.txt.gz#_2015-04-01_19_02_55_274 | 21:56 |
| jogo | how does that have a request ID | 21:56 |
| jogo | that is not right | 21:56 |
| mriedem_away | dansmith: don't worry about it, people on twitter don't expect correct grammar :) | 21:56 |
| *** iamjarvo has quit IRC | 21:56 | |
| *** vilobhmm1 has quit IRC | 21:57 | |
| dansmith | mriedem_away: https://twitter.com/get_offmylawn/status/583387590314688512 | 21:57 |
| dansmith | mriedem_away: yes, but I do | 21:57 |
| jogo | mriedem_away: but the instance doesn't live on that host | 21:57 |
| mriedem_away | dansmith: oh i didn't know you could amend | 21:57 |
| dansmith | you can delete | 21:57 |
| dansmith | it's not as good | 21:57 |
| mriedem_away | it's like it never happened... | 21:57 |
| dansmith | but I'm not willing to let that sit forever | 21:57 |
| mriedem_away | alright, time to go now that that's all settled | 21:57 |
| *** boris-42 has joined #openstack-nova | 21:57 | |
| dansmith | mriedem_away: thanks for catching it | 21:59 |
| *** ferest has quit IRC | 22:00 | |
| jogo | vishy: nova-net question if your around | 22:00 |
| *** emagana_ has quit IRC | 22:01 | |
| bauzas | dansmith: so lucky I am to follow you on Twitter... | 22:02 |
| *** setuid has quit IRC | 22:03 | |
| *** flwang1 has quit IRC | 22:03 | |
| *** ekarlso has quit IRC | 22:04 | |
| *** ekarlso has joined #openstack-nova | 22:04 | |
| vishy | jogo: shoot | 22:05 |
| *** iamjarvo has joined #openstack-nova | 22:05 | |
| *** VW_ has quit IRC | 22:06 | |
| jogo | vishy: trying to debug why this multi node nova-net multihost job is failing http://logs.openstack.org/73/169873/1/check/check-tempest-dsvm-aiopcpu-full/ea5db7b/console.html | 22:07 |
| *** vilobhmm1 has joined #openstack-nova | 22:07 | |
| jogo | we have two nodes and nova-net on the wrong node tries to access info about the instance on the other host | 22:07 |
| jogo | http://logs.openstack.org/73/169873/1/check/check-tempest-dsvm-aiopcpu-full/ea5db7b/logs/screen-n-net.txt.gz#_2015-04-01_19_33_06_091 | 22:07 |
| jogo | not sure why | 22:07 |
| jogo | looks like its a periodic task doing it though | 22:07 |
| vishy | it isn’t getting dhcp | 22:08 |
| jogo | vishy: any ideas why? | 22:09 |
| vishy | not sure | 22:10 |
| vishy | is it just that one instance that fails to dhcp? | 22:10 |
| jogo | yeah only one test failures in that case | 22:11 |
| jogo | in general we are seeing a few issues with the multi node nova-net job | 22:11 |
| *** hdd has quit IRC | 22:11 | |
| jogo | mostly DHCP related | 22:11 |
| *** ZZelle__ has quit IRC | 22:11 | |
| jogo | seeing one with shelve/unshelve across instances etc. | 22:12 |
| *** 77CAAKTHH has joined #openstack-nova | 22:13 | |
| *** mwagner_lap has joined #openstack-nova | 22:14 | |
| *** Nic has quit IRC | 22:14 | |
| *** angdraug has joined #openstack-nova | 22:15 | |
| *** neelashah has quit IRC | 22:16 | |
| dansmith | bauzas: heh | 22:16 |
| *** tjones1 has quit IRC | 22:18 | |
| *** angdraug has quit IRC | 22:18 | |
| *** ZZelle__ has joined #openstack-nova | 22:19 | |
| *** tjones1 has joined #openstack-nova | 22:19 | |
| jogo | vishy: http://logs.openstack.org/73/169873/1/check/check-tempest-dsvm-aiopcpu-full/ea5db7b/logs/10.209.96.247-subnode/ps.txt.gz | 22:20 |
| jogo | there are two duplicate copies of dnsmasq running is that expected? | 22:20 |
| *** Marga_ has quit IRC | 22:21 | |
| *** Marga_ has joined #openstack-nova | 22:21 | |
| *** nelsnelson has quit IRC | 22:23 | |
| *** 77CAAKTHH has quit IRC | 22:23 | |
| *** jecarey has quit IRC | 22:24 | |
| *** dkliban is now known as dkliban_afk | 22:24 | |
| *** asettle has quit IRC | 22:25 | |
| *** matrohon has quit IRC | 22:26 | |
| *** hdd has joined #openstack-nova | 22:26 | |
| *** dave-mccowan has quit IRC | 22:28 | |
| *** oro has quit IRC | 22:30 | |
| vishy | there are always two yes | 22:30 |
| dansmith | we just confirmed this actually fixes production, if someone can hit the +W on it: https://review.openstack.org/#/c/169446/ | 22:31 |
| vishy | one is privileged and the other drops to run the script | 22:31 |
| *** igordcard_ has joined #openstack-nova | 22:32 | |
| *** emagana has joined #openstack-nova | 22:35 | |
| *** dave-mccowan has joined #openstack-nova | 22:40 | |
| *** emagana has quit IRC | 22:43 | |
| mikal | dansmith: /me looks at the review | 22:45 |
| dansmith | mikal: we only hit it when we're generating ungodly amounts of load in the guest during the destroy, so it's not really something we can easily hit in CI | 22:46 |
| *** annegentle has joined #openstack-nova | 22:48 | |
| *** annegentle has quit IRC | 22:48 | |
| dansmith | mikal: thanks | 22:49 |
| mikal | NP | 22:50 |
| *** andreykurilin_ has joined #openstack-nova | 22:55 | |
| *** hblixt has quit IRC | 22:56 | |
| *** ferest has joined #openstack-nova | 22:56 | |
| *** annegentle has joined #openstack-nova | 22:58 | |
| *** setuid has joined #openstack-nova | 22:59 | |
| *** hemna is now known as hemnafk | 22:59 | |
| *** ferest has quit IRC | 23:00 | |
| salv-orl_ | sdague, dansmith: I have a quick patch for bug 1438738. however the author of the nova patch assigned the neutron bug to himself. Since it would be unfair to strip a fairly new contributor to a bug (which is also a low-hanging fruit), I just attached a patch to launchpad so that the assignee can decide whether to use all of it, part of it, or none of it | 23:00 |
| openstack | bug 1438738 in neutron "Nova boot fails with Error Code 500, if quota_port is < -1 in neutron.conf" [Medium,New] https://launchpad.net/bugs/1438738 - Assigned to Sudipta Biswas (sbiswas7) | 23:00 |
| *** tjones1 has quit IRC | 23:00 | |
| *** ijw has joined #openstack-nova | 23:01 | |
| dansmith | salv-orl_: as you wish, but I'd rather just have it fixed :/ | 23:02 |
| *** sbfox has quit IRC | 23:03 | |
| *** tjones1 has joined #openstack-nova | 23:03 | |
| *** pixelb has joined #openstack-nova | 23:04 | |
| dansmith | salv-orl_: looks like it's targeted, so hopefully if it's not picked up and submitted in the next couple days, it'll be fair game.. | 23:04 |
| salv-orl_ | dansmith: sure, I targeted it. If I do not see a patch by friday I will push mine | 23:05 |
| dansmith | salv-orl_: okay, sounds good | 23:06 |
| dansmith | thanks! | 23:06 |
| salv-orl_ | np | 23:06 |
| *** penick has quit IRC | 23:11 | |
| jogo | hmm vishy hrm thanks, any ideas why that job failed then? | 23:12 |
| *** vmtrooper has joined #openstack-nova | 23:13 | |
| *** andreykurilin_ has quit IRC | 23:13 | |
| *** annashen has quit IRC | 23:14 | |
| *** igordcard has quit IRC | 23:14 | |
| *** annegentle has quit IRC | 23:15 | |
| *** vmtrooper has quit IRC | 23:18 | |
| vishy | jogo: it is running nova-network on all hosts right? | 23:19 |
| vishy | is sharing the dhcp address i wonder? | 23:20 |
| vishy | or is this the one with a single nova-network running | 23:20 |
| *** asettle has joined #openstack-nova | 23:20 | |
| *** alexandra has joined #openstack-nova | 23:21 | |
| vishy | f the second case | 23:21 |
| vishy | i would suspect that somehow the network is broken between the two machines | 23:21 |
| *** asettle has quit IRC | 23:22 | |
| *** alexandra is now known as asettle | 23:22 | |
| *** ijw has quit IRC | 23:23 | |
| *** packet has quit IRC | 23:24 | |
| *** Marga_ has quit IRC | 23:29 | |
| *** annegentle has joined #openstack-nova | 23:29 | |
| *** Marga_ has joined #openstack-nova | 23:30 | |
| *** ijw has joined #openstack-nova | 23:30 | |
| *** harlowja has quit IRC | 23:31 | |
| *** harlowja has joined #openstack-nova | 23:32 | |
| *** annegentle has quit IRC | 23:37 | |
| *** devlaps1 has quit IRC | 23:37 | |
| *** salv-orlando has joined #openstack-nova | 23:40 | |
| *** claudiub has quit IRC | 23:42 | |
| *** salv-orl_ has quit IRC | 23:42 | |
| *** zzzeek has quit IRC | 23:43 | |
| *** harlowja has quit IRC | 23:49 | |
| *** harlowja has joined #openstack-nova | 23:50 | |
| jogo | vishy: yes nova-network on both hosts | 23:53 |
| vishy | and do they use the same ip? | 23:54 |
| vishy | share_dhcp_address = true either on network create or as a flag? | 23:54 |
| jogo | vishy: let me check | 23:55 |
| *** smccully has quit IRC | 23:55 | |
| jogo | vishy: http://logs.openstack.org/73/169873/1/check/check-tempest-dsvm-aiopcpu-full/ea5db7b/logs/etc/nova/nova.conf.txt.gz | 23:55 |
| *** tjones1 has quit IRC | 23:55 | |
| jogo | vishy: share_dhcp_address defaults to false and has a deprecated warning next to it | 23:56 |
| *** tjones1 has joined #openstack-nova | 23:56 | |
| jogo | set when creating the network | 23:56 |
| vishy | it can also be set during network create | 23:57 |
| vishy | it is set? | 23:57 |
| vishy | ok so in that case each node has the same ip address | 23:57 |
| *** ferest has joined #openstack-nova | 23:57 | |
| vishy | and egress traffic for dhcp is blocked by ebtables | 23:57 |
| vishy | very hard to figure out why it wouldn’t be working without a packet dump | 23:57 |
| jogo | strange thing is this only fails sometimes and not sure why | 23:58 |
| jogo | vishy: I have to go but maybe we can get a packet dump for next time | 23:58 |
| vishy | that would be nice | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
