*** haomaiwang has quit IRC | 00:01 | |
*** haomaiwang has joined #openstack-nova | 00:01 | |
melwitt | dims: is there any other context about that, like a bug or logstash query? I'm not familiar with the py3 test flakiness | 00:02 |
---|---|---|
*** Jeffrey4l has joined #openstack-nova | 00:04 | |
*** sacharya has joined #openstack-nova | 00:05 | |
*** _diana_1 has joined #openstack-nova | 00:06 | |
*** gyee has quit IRC | 00:07 | |
*** unicell1 has joined #openstack-nova | 00:07 | |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: libvirt: Allow multiple volume attachments https://review.openstack.org/193135 | 00:08 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: comp: Allow multiple volume attachments https://review.openstack.org/193134 | 00:08 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: db and obj: Allow multiple volume attachments https://review.openstack.org/193133 | 00:08 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: WIP: REST API changes to support multi-attach https://review.openstack.org/153038 | 00:08 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: WIP: virt: Pass host when call attach to Cinder https://review.openstack.org/256273 | 00:08 |
*** ducttape_ has quit IRC | 00:09 | |
*** annegentle has quit IRC | 00:09 | |
*** penick has joined #openstack-nova | 00:09 | |
*** unicell has quit IRC | 00:09 | |
*** smatzek has quit IRC | 00:09 | |
dims | melwitt : if you check the gate-nova-python34 uncategorized rechecks page - you will see these tests - http://status.openstack.org//elastic-recheck/data/uncategorized.html | 00:09 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: libvirt: Allow multiple volume attachments https://review.openstack.org/193135 | 00:10 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: comp: Allow multiple volume attachments https://review.openstack.org/193134 | 00:10 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: db and obj: Allow multiple volume attachments https://review.openstack.org/193133 | 00:10 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: WIP: REST API changes to support multi-attach https://review.openstack.org/153038 | 00:10 |
*** sacharya has quit IRC | 00:10 | |
dims | melwitt : example http://logs.openstack.org/78/258878/3/gate/gate-nova-python34/2c43415/testr_results.html.gz | 00:11 |
melwitt | dims: ah, thanks | 00:11 |
dims | melwitt : i actually encountered some of those when i was fixing the py34 tests in https://review.openstack.org/#/c/262083/ and https://review.openstack.org/#/c/261045/ | 00:12 |
*** mdrabe has joined #openstack-nova | 00:12 | |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: libvirt: Allow multiple volume attachments https://review.openstack.org/193135 | 00:13 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: comp: Allow multiple volume attachments https://review.openstack.org/193134 | 00:13 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: db and obj: Allow multiple volume attachments https://review.openstack.org/193133 | 00:13 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: WIP: REST API changes to support multi-attach https://review.openstack.org/153038 | 00:13 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: WIP: virt: Pass host when call attach to Cinder https://review.openstack.org/256273 | 00:13 |
*** gyee has joined #openstack-nova | 00:14 | |
*** _diana_1 has quit IRC | 00:15 | |
melwitt | dims: looks like all of the others are AbstractDriverTestCase | 00:15 |
dims | melwitt : ah. let me add that one as well | 00:16 |
*** ljxiash has joined #openstack-nova | 00:16 | |
*** pratikmallya has joined #openstack-nova | 00:18 | |
melwitt | dims: cool, I'll +2 after the update | 00:18 |
*** _diana_1 has joined #openstack-nova | 00:19 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Blacklist flaky tests and add warning https://review.openstack.org/263789 | 00:19 |
dims | thanks melwitt! done | 00:20 |
*** ljxiash has quit IRC | 00:21 | |
*** Sukhdev has joined #openstack-nova | 00:22 | |
*** egarbade-- has quit IRC | 00:24 | |
*** gyee has quit IRC | 00:24 | |
*** harlowja_ has joined #openstack-nova | 00:27 | |
*** _diana_1 has quit IRC | 00:28 | |
*** harlowja has quit IRC | 00:30 | |
*** lihg has quit IRC | 00:31 | |
*** lihg_ has quit IRC | 00:31 | |
*** penick has quit IRC | 00:32 | |
*** penick has joined #openstack-nova | 00:35 | |
dims | oomichi : i did some ec2-api/objectstore cleanup (chop 10k lines off nova) - https://review.openstack.org/#/c/263368/ if you get a chance | 00:41 |
oomichi | dims: hi, happy new year. ok, now checking :) | 00:42 |
*** Jeffrey4l has quit IRC | 00:43 | |
*** otter768 has joined #openstack-nova | 00:43 | |
dims | thanks oomichi | 00:44 |
dims | wish you a very happy new year as well oomichi | 00:44 |
oomichi | dims: one question. do we need to remove ec2 impl/tests with ObjectStore together? | 00:44 |
oomichi | dims: the patch becomes huge, and it is nice to separate them if possible | 00:45 |
dims | oomichi : y they are tied together. will generate more work to split them apart | 00:45 |
*** egarbade-- has joined #openstack-nova | 00:45 | |
oomichi | dims: ok, I see | 00:46 |
*** smoriya_ has joined #openstack-nova | 00:46 | |
*** blaisebool has quit IRC | 00:47 | |
*** egarbade-- has quit IRC | 00:48 | |
*** egarbade-- has joined #openstack-nova | 00:48 | |
*** egarbade-- has quit IRC | 00:51 | |
*** egarbade-- has joined #openstack-nova | 00:51 | |
*** ljxiash has joined #openstack-nova | 00:53 | |
*** ssurana1 has quit IRC | 00:53 | |
oomichi | dims: one more question: how about s3 code(nova/image/s3.py, etc.)? The patch seems to keep the code intentionally | 00:54 |
*** egarbade-- has quit IRC | 00:54 | |
*** egarbade-- has joined #openstack-nova | 00:55 | |
*** zhangjn has joined #openstack-nova | 00:57 | |
*** haomaiwang has quit IRC | 01:01 | |
*** haomaiwang has joined #openstack-nova | 01:01 | |
*** pratikmallya has quit IRC | 01:03 | |
*** Sukhdev has quit IRC | 01:08 | |
*** suro-patz has quit IRC | 01:08 | |
*** eharney has quit IRC | 01:08 | |
*** penick has quit IRC | 01:14 | |
*** ljxiash has quit IRC | 01:17 | |
*** EinstCrazy has joined #openstack-nova | 01:17 | |
dims | oomichi : yes, there's a dependency on the data stored, you will see ec2_ids etc in Instance object. this review does not deal with those. some of that data is needed/used for example when generating the data for the config drive. i am deliberately avoiding touching any changes to data | 01:18 |
dims | and removing up just the REST API | 01:19 |
dims | thanks for the review oomichi - will reply there | 01:20 |
oomichi | dims: I see. I also prefer to avoid to see object changes in REST API patches ;) | 01:20 |
dims | :) i tried cutting it several ways, this seemed the best | 01:21 |
*** mdrabe has quit IRC | 01:22 | |
*** zhangjn has quit IRC | 01:28 | |
*** zhangjn has joined #openstack-nova | 01:29 | |
*** Sukhdev has joined #openstack-nova | 01:29 | |
*** gyee has joined #openstack-nova | 01:31 | |
*** haomaiwang has quit IRC | 01:32 | |
*** zhangyufei has joined #openstack-nova | 01:33 | |
*** ljxiash has joined #openstack-nova | 01:38 | |
*** zenoway has quit IRC | 01:40 | |
*** gongysh has joined #openstack-nova | 01:44 | |
openstackgerrit | John Hua proposed openstack/nova: xenapi: Image cache cannot be disabled https://review.openstack.org/262661 | 01:44 |
*** gongysh has quit IRC | 01:47 | |
*** andreykurilin__ has quit IRC | 02:02 | |
*** zhangyufei has quit IRC | 02:05 | |
*** mtanino has quit IRC | 02:07 | |
*** zhangyufei has joined #openstack-nova | 02:08 | |
*** annegentle has joined #openstack-nova | 02:10 | |
*** lyanchih has joined #openstack-nova | 02:10 | |
*** ducttape_ has joined #openstack-nova | 02:10 | |
*** ducttape_ has quit IRC | 02:14 | |
*** annegentle has quit IRC | 02:14 | |
*** zhangjn has quit IRC | 02:17 | |
*** zhangjn has joined #openstack-nova | 02:18 | |
*** lihg has joined #openstack-nova | 02:20 | |
*** lihg_ has joined #openstack-nova | 02:20 | |
*** zhangyufei has quit IRC | 02:20 | |
*** 14WAALZU3 has joined #openstack-nova | 02:20 | |
*** lihg has quit IRC | 02:21 | |
*** lihg_ has quit IRC | 02:21 | |
*** houming has joined #openstack-nova | 02:21 | |
*** ssurana has joined #openstack-nova | 02:22 | |
*** _diana_1 has joined #openstack-nova | 02:24 | |
*** zhangyufei has joined #openstack-nova | 02:26 | |
*** Sukhdev has quit IRC | 02:39 | |
*** vilobhmm11 has quit IRC | 02:39 | |
*** zhangjn has quit IRC | 02:39 | |
*** sacharya has joined #openstack-nova | 02:42 | |
*** zhangjn has joined #openstack-nova | 02:42 | |
*** lihg has joined #openstack-nova | 02:46 | |
*** lihg_ has joined #openstack-nova | 02:46 | |
*** sacharya has quit IRC | 02:46 | |
*** _diana_1 has quit IRC | 02:47 | |
*** _diana_1 has joined #openstack-nova | 02:48 | |
*** _diana_1 has quit IRC | 02:50 | |
*** sacharya has joined #openstack-nova | 02:51 | |
*** egarbade-- has quit IRC | 02:52 | |
*** _diana_1 has joined #openstack-nova | 02:54 | |
*** zhangjn has quit IRC | 02:54 | |
*** thorst has quit IRC | 02:56 | |
*** thorst has joined #openstack-nova | 02:57 | |
*** rushiagr_away is now known as rushiagr | 02:57 | |
*** ssurana has quit IRC | 02:57 | |
*** zhangjn has joined #openstack-nova | 02:58 | |
*** dims has quit IRC | 02:59 | |
*** 14WAALZU3 has quit IRC | 03:01 | |
*** haomaiwang has joined #openstack-nova | 03:01 | |
*** thorst has quit IRC | 03:02 | |
*** browne has quit IRC | 03:02 | |
*** browne has joined #openstack-nova | 03:02 | |
*** suro-patz has joined #openstack-nova | 03:03 | |
*** _diana_1 has quit IRC | 03:07 | |
*** niusmallnan has joined #openstack-nova | 03:07 | |
*** vilobhmm11 has joined #openstack-nova | 03:08 | |
*** niusmallnan has left #openstack-nova | 03:09 | |
*** _diana_1 has joined #openstack-nova | 03:10 | |
*** lihg has quit IRC | 03:10 | |
*** lihg_ has quit IRC | 03:10 | |
*** unicell1 has quit IRC | 03:10 | |
*** ducttape_ has joined #openstack-nova | 03:11 | |
*** gjayavelu has joined #openstack-nova | 03:13 | |
*** gjayavelu has quit IRC | 03:13 | |
*** ducttape_ has quit IRC | 03:16 | |
*** thorst has joined #openstack-nova | 03:17 | |
*** sacharya has quit IRC | 03:17 | |
*** sacharya has joined #openstack-nova | 03:18 | |
*** lyanchih has quit IRC | 03:25 | |
*** armax has joined #openstack-nova | 03:26 | |
*** vilobhmm11 has quit IRC | 03:28 | |
*** gyee has quit IRC | 03:30 | |
*** lyanchih has joined #openstack-nova | 03:35 | |
openstackgerrit | Zhenyu Zheng proposed openstack/nova: config options: Centralise floating ip options https://review.openstack.org/257087 | 03:36 |
*** annegentle has joined #openstack-nova | 03:37 | |
*** unicell has joined #openstack-nova | 03:37 | |
*** ljxiash has quit IRC | 03:38 | |
*** ljxiash has joined #openstack-nova | 03:38 | |
*** zhangjn has quit IRC | 03:38 | |
*** annegentle has quit IRC | 03:39 | |
*** ljxiash has quit IRC | 03:43 | |
*** annegentle has joined #openstack-nova | 03:44 | |
openstackgerrit | Vladik Romanovsky proposed openstack/nova: pci: do not filter out any SRIOV Physical Functions https://review.openstack.org/249016 | 03:45 |
openstackgerrit | Vladik Romanovsky proposed openstack/nova: pci: changing the claiming and allocation logic for PF/VF assignment https://review.openstack.org/249018 | 03:45 |
openstackgerrit | Vladik Romanovsky proposed openstack/nova: pci: adding support to specify a device_type in pci requests https://review.openstack.org/259254 | 03:45 |
openstackgerrit | Vladik Romanovsky proposed openstack/nova: objects: adding a parent_addr field to the PciDevice object https://review.openstack.org/249015 | 03:45 |
*** suro-patz has quit IRC | 03:45 | |
*** links has joined #openstack-nova | 03:51 | |
*** yamahata has quit IRC | 03:52 | |
*** sacharya has quit IRC | 03:52 | |
*** zhangyufei has quit IRC | 03:53 | |
*** unicell has quit IRC | 03:54 | |
*** zhangyufei has joined #openstack-nova | 03:54 | |
openstackgerrit | Zhenyu Zheng proposed openstack/nova: Expose quiesce/unquiesce API https://review.openstack.org/248989 | 03:54 |
*** sacharya has joined #openstack-nova | 03:54 | |
*** Kevin_Zheng has joined #openstack-nova | 03:57 | |
*** _diana_1 has quit IRC | 04:00 | |
*** haomaiwang has quit IRC | 04:01 | |
openstackgerrit | Eli Qiao proposed openstack/nova: Calculate block_migration in virt drivers if it's None https://review.openstack.org/254101 | 04:01 |
openstackgerrit | Eli Qiao proposed openstack/nova: Return host and block_migration in task_api._execute() https://review.openstack.org/254102 | 04:01 |
openstackgerrit | Eli Qiao proposed openstack/nova: API: Improve os-migrateLive input and output values https://review.openstack.org/259319 | 04:01 |
*** haomaiwang has joined #openstack-nova | 04:01 | |
*** thorst has quit IRC | 04:04 | |
*** thorst has joined #openstack-nova | 04:04 | |
*** davideagnello has quit IRC | 04:05 | |
*** coolsvap|away is now known as coolsvap | 04:06 | |
*** otter768 has quit IRC | 04:07 | |
*** browne has quit IRC | 04:08 | |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: virtual interface tags DB table https://review.openstack.org/264014 | 04:09 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: metadata objects https://review.openstack.org/264015 | 04:09 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: libvirt driver implementation https://review.openstack.org/264016 | 04:09 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: compute manager nic tagging support https://review.openstack.org/264017 | 04:09 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: user visible changes https://review.openstack.org/264018 | 04:09 |
*** kotra03 has joined #openstack-nova | 04:09 | |
*** ducttape_ has joined #openstack-nova | 04:12 | |
*** PsionTheory has quit IRC | 04:13 | |
*** thorst has quit IRC | 04:13 | |
*** rushiagr is now known as rushiagr_away | 04:14 | |
*** shyama has joined #openstack-nova | 04:16 | |
*** ducttape_ has quit IRC | 04:17 | |
*** shyama has quit IRC | 04:18 | |
*** suro-patz has joined #openstack-nova | 04:26 | |
*** coolsvap is now known as coolsvap|away | 04:29 | |
openstackgerrit | Augustina Ragwitz proposed openstack/nova: Make project_id optional in v2.1 urls https://review.openstack.org/233076 | 04:31 |
*** annegentle has quit IRC | 04:49 | |
*** oomichi has quit IRC | 04:49 | |
*** zhangyufei has quit IRC | 04:50 | |
*** gongysh has joined #openstack-nova | 04:51 | |
*** annegentle has joined #openstack-nova | 04:51 | |
*** ljxiash has joined #openstack-nova | 04:54 | |
*** rushiagr_away is now known as rushiagr | 04:57 | |
*** sudipto has joined #openstack-nova | 04:58 | |
*** lyanchih has quit IRC | 05:00 | |
*** haomaiwang has quit IRC | 05:01 | |
*** haomaiwang has joined #openstack-nova | 05:01 | |
*** Jeffrey4l has joined #openstack-nova | 05:02 | |
*** gongysh has quit IRC | 05:06 | |
*** davideagnello has joined #openstack-nova | 05:06 | |
*** gongysh has joined #openstack-nova | 05:10 | |
*** thorst has joined #openstack-nova | 05:10 | |
*** davideagnello has quit IRC | 05:11 | |
*** aagrawal has quit IRC | 05:13 | |
*** vilobhmm11 has joined #openstack-nova | 05:13 | |
*** _diana_1 has joined #openstack-nova | 05:13 | |
*** ducttape_ has joined #openstack-nova | 05:13 | |
*** suro-patz has quit IRC | 05:15 | |
*** fawadkhaliq has joined #openstack-nova | 05:17 | |
*** chhavi has joined #openstack-nova | 05:17 | |
*** thorst has quit IRC | 05:18 | |
*** ijw_ has quit IRC | 05:18 | |
*** gjayavelu has joined #openstack-nova | 05:18 | |
*** ducttape_ has quit IRC | 05:18 | |
*** cfriesen__ has joined #openstack-nova | 05:20 | |
*** _diana_1 has quit IRC | 05:21 | |
*** chhavi has quit IRC | 05:26 | |
*** zhangjn has joined #openstack-nova | 05:26 | |
*** chhavi has joined #openstack-nova | 05:26 | |
*** unicell has joined #openstack-nova | 05:34 | |
*** zhangjn has quit IRC | 05:35 | |
*** coolsvap|away is now known as coolsvap | 05:35 | |
*** annegentle has quit IRC | 05:36 | |
*** zhangjn has joined #openstack-nova | 05:36 | |
*** annegentle has joined #openstack-nova | 05:36 | |
*** ssurana has joined #openstack-nova | 05:42 | |
openstackgerrit | Alex Xu proposed openstack/nova: doc: completes microversion use-cases in api concept doc https://review.openstack.org/263220 | 05:43 |
eliqiao | hi alex_xu, there? | 05:45 |
alex_xu | eliqiao: yup, here | 05:45 |
eliqiao | I am trying to understand https://review.openstack.org/#/c/245921/18/nova/api/openstack/compute/extension_info.py | 05:45 |
eliqiao | alex_xu: can you help to understand what does v2_extension_suppress_list mean? | 05:45 |
alex_xu | eliqiao: it means hide some extension name in the list of extensions | 05:46 |
alex_xu | eliqiao: that is used to keep the extension api looks like freeze | 05:47 |
alex_xu | as we didn't have extension in the future any more, we said we just keep the extension api looks like freeze | 05:48 |
eliqiao | alex_xu: thanks, I will try to read the code more detail. | 05:49 |
alex_xu | eliqiao: np | 05:49 |
*** irenab has joined #openstack-nova | 05:55 | |
*** coolsvap is now known as coolsvap|away | 05:55 | |
*** moshele has joined #openstack-nova | 05:55 | |
*** gcb has joined #openstack-nova | 05:56 | |
tonyb | Gah no Ken'ichi :( I'll look for him later .... | 05:57 |
*** lyanchih has joined #openstack-nova | 05:59 | |
*** annegentle has quit IRC | 06:00 | |
*** zhangjn_ has joined #openstack-nova | 06:00 | |
*** annegentle has joined #openstack-nova | 06:00 | |
*** haomaiwang has quit IRC | 06:01 | |
*** achinm has joined #openstack-nova | 06:01 | |
*** zhangjn has quit IRC | 06:02 | |
*** haomaiwa_ has joined #openstack-nova | 06:03 | |
*** annegentle has quit IRC | 06:04 | |
*** zhangjn_ has quit IRC | 06:05 | |
*** unicell1 has joined #openstack-nova | 06:06 | |
*** unicell has quit IRC | 06:07 | |
*** zhangjn has joined #openstack-nova | 06:08 | |
*** otter768 has joined #openstack-nova | 06:08 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Imported Translations from Zanata https://review.openstack.org/252197 | 06:10 |
*** sacharya has quit IRC | 06:11 | |
*** cfriesen__ has quit IRC | 06:11 | |
*** otter768 has quit IRC | 06:13 | |
*** ducttape_ has joined #openstack-nova | 06:14 | |
*** lyanchih has quit IRC | 06:15 | |
*** zhangjn has quit IRC | 06:15 | |
*** thorst has joined #openstack-nova | 06:16 | |
*** zhangjn has joined #openstack-nova | 06:16 | |
*** coolsvap|away is now known as coolsvap | 06:18 | |
*** ducttape_ has quit IRC | 06:19 | |
*** ijw has joined #openstack-nova | 06:20 | |
*** dslevin has quit IRC | 06:20 | |
*** ijw has quit IRC | 06:20 | |
*** vilobhmm111 has joined #openstack-nova | 06:22 | |
*** ijw has joined #openstack-nova | 06:22 | |
*** ijw has quit IRC | 06:22 | |
*** ijw has joined #openstack-nova | 06:23 | |
*** thorst has quit IRC | 06:23 | |
*** vilobhmm11 has quit IRC | 06:24 | |
*** zhangyufei has joined #openstack-nova | 06:29 | |
*** gwei3 has joined #openstack-nova | 06:34 | |
*** gongysh has quit IRC | 06:36 | |
*** Sukhdev has joined #openstack-nova | 06:37 | |
*** jwcroppe has quit IRC | 06:38 | |
*** suro-patz has joined #openstack-nova | 06:40 | |
openstackgerrit | Alex Xu proposed openstack/nova: doc: add version discovery guideline in api concept doc https://review.openstack.org/264044 | 06:41 |
openstackgerrit | Alex Xu proposed openstack/nova: doc: add client interactive guideline for microversions https://review.openstack.org/264045 | 06:41 |
*** nelsnels_ has quit IRC | 06:42 | |
*** paul-carlton2 has joined #openstack-nova | 06:43 | |
*** cfriesen__ has joined #openstack-nova | 06:47 | |
*** shuquan has joined #openstack-nova | 06:49 | |
*** nelsnelson has joined #openstack-nova | 06:49 | |
*** paul-carlton2 has quit IRC | 06:50 | |
*** paul-carlton2 has joined #openstack-nova | 06:52 | |
*** dslevin has joined #openstack-nova | 06:54 | |
*** belmoreira has joined #openstack-nova | 06:59 | |
openstackgerrit | Yingxin Cheng proposed openstack/nova: Refactor claim code to eliminate duplication https://review.openstack.org/262939 | 06:59 |
openstackgerrit | Yingxin Cheng proposed openstack/nova: Scheduler use Claim to check resource consumption https://review.openstack.org/262938 | 06:59 |
*** gwei31 has joined #openstack-nova | 07:00 | |
*** haomaiwa_ has quit IRC | 07:01 | |
*** gwei31 has quit IRC | 07:01 | |
*** sahid has joined #openstack-nova | 07:01 | |
*** gwei3 has quit IRC | 07:01 | |
*** haomaiwang has joined #openstack-nova | 07:01 | |
*** armax has quit IRC | 07:07 | |
*** davideagnello has joined #openstack-nova | 07:07 | |
*** xianghui has quit IRC | 07:10 | |
*** zhangjn has quit IRC | 07:11 | |
*** sacharya has joined #openstack-nova | 07:12 | |
*** davideagnello has quit IRC | 07:13 | |
*** ljxiash has quit IRC | 07:14 | |
*** ljxiash has joined #openstack-nova | 07:15 | |
*** ducttape_ has joined #openstack-nova | 07:15 | |
*** zhangjn has joined #openstack-nova | 07:15 | |
*** sacharya has quit IRC | 07:16 | |
*** xianghui has joined #openstack-nova | 07:16 | |
*** suro-patz has quit IRC | 07:18 | |
*** ljxiash_ has joined #openstack-nova | 07:18 | |
*** yangyapeng has joined #openstack-nova | 07:19 | |
*** klkumar has joined #openstack-nova | 07:19 | |
*** ljxiash has quit IRC | 07:19 | |
*** ducttape_ has quit IRC | 07:20 | |
*** ljxiash_ has quit IRC | 07:21 | |
*** ljxiash has joined #openstack-nova | 07:21 | |
*** thorst has joined #openstack-nova | 07:21 | |
*** ljxiash has quit IRC | 07:21 | |
*** ljxiash has joined #openstack-nova | 07:22 | |
*** e0ne has joined #openstack-nova | 07:25 | |
*** ljxiash has quit IRC | 07:26 | |
*** artom has quit IRC | 07:28 | |
*** thorst has quit IRC | 07:28 | |
*** ljxiash has joined #openstack-nova | 07:29 | |
*** ijw has quit IRC | 07:31 | |
*** sahid has quit IRC | 07:31 | |
*** yangyapeng has quit IRC | 07:33 | |
*** ljxiash has quit IRC | 07:34 | |
*** fawadkhaliq has quit IRC | 07:37 | |
*** ljxiash has joined #openstack-nova | 07:39 | |
*** fawadkhaliq has joined #openstack-nova | 07:40 | |
*** zhangjn has quit IRC | 07:45 | |
*** Sukhdev has quit IRC | 07:46 | |
*** wuhg has joined #openstack-nova | 07:47 | |
*** fawadkhaliq has quit IRC | 07:48 | |
*** fawadkhaliq has joined #openstack-nova | 07:49 | |
*** Sukhdev has joined #openstack-nova | 07:49 | |
*** shuquan has quit IRC | 07:52 | |
*** ljxiash has quit IRC | 07:53 | |
*** shuquan has joined #openstack-nova | 07:53 | |
*** yarkot has quit IRC | 07:55 | |
*** lpetrut has joined #openstack-nova | 07:56 | |
openstackgerrit | Eli Qiao proposed openstack/nova: API: Improve os-migrateLive input and output values https://review.openstack.org/259319 | 07:56 |
*** Sukhdev has quit IRC | 07:57 | |
*** ijw has joined #openstack-nova | 07:58 | |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: libvirt: Allow multiple volume attachments https://review.openstack.org/193135 | 07:58 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: comp: Allow multiple volume attachments https://review.openstack.org/193134 | 07:58 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: db and obj: Allow multiple volume attachments https://review.openstack.org/193133 | 07:58 |
*** ljxiash has joined #openstack-nova | 07:58 | |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: WIP: REST API changes to support multi-attach https://review.openstack.org/153038 | 07:58 |
openstackgerrit | Ildiko Vancsa proposed openstack/nova: WIP: virt: Pass host when call attach to Cinder https://review.openstack.org/256273 | 07:58 |
*** coolsvap is now known as coolsvap|away | 08:00 | |
*** ljxiash has quit IRC | 08:00 | |
*** yamahata has joined #openstack-nova | 08:00 | |
*** haomaiwang has quit IRC | 08:01 | |
*** rushiagr is now known as rushiagr_away | 08:01 | |
*** haomaiwang has joined #openstack-nova | 08:01 | |
*** ljxiash has joined #openstack-nova | 08:01 | |
*** lyanchih has joined #openstack-nova | 08:02 | |
*** ljxiash_ has joined #openstack-nova | 08:02 | |
*** ljxiash__ has joined #openstack-nova | 08:04 | |
*** ljxiash has quit IRC | 08:05 | |
*** coolsvap|away is now known as coolsvap | 08:06 | |
*** ljxiash_ has quit IRC | 08:07 | |
*** otter768 has joined #openstack-nova | 08:09 | |
*** zakora has joined #openstack-nova | 08:10 | |
*** rdopiera has joined #openstack-nova | 08:10 | |
*** otter768 has quit IRC | 08:14 | |
*** ducttape_ has joined #openstack-nova | 08:16 | |
*** vilobhmm11 has joined #openstack-nova | 08:17 | |
*** vilobhmm112 has joined #openstack-nova | 08:19 | |
*** vilobhmm111 has quit IRC | 08:19 | |
*** apoorvad has joined #openstack-nova | 08:20 | |
*** ducttape_ has quit IRC | 08:21 | |
*** vilobhmm11 has quit IRC | 08:21 | |
*** shuquan has quit IRC | 08:21 | |
openstackgerrit | Chaozhe Chen proposed openstack/python-novaclient: Suppress 'get_client_class' deprecated warnings in test_client https://review.openstack.org/259401 | 08:21 |
*** rushiagr_away is now known as rushiagr | 08:22 | |
*** jichen has joined #openstack-nova | 08:25 | |
*** lyanchih has quit IRC | 08:25 | |
*** thorst has joined #openstack-nova | 08:26 | |
*** rushiagr is now known as rushiagr_away | 08:27 | |
*** gjayavelu has quit IRC | 08:28 | |
*** rushiagr_away is now known as rushiagr | 08:28 | |
*** gongysh has joined #openstack-nova | 08:30 | |
*** sahid has joined #openstack-nova | 08:31 | |
*** EinstCra_ has joined #openstack-nova | 08:32 | |
*** matrohon has joined #openstack-nova | 08:33 | |
*** thorst has quit IRC | 08:33 | |
bauzas | morning Nova | 08:33 |
openstackgerrit | Jason proposed openstack/nova: Correct the code description https://review.openstack.org/264071 | 08:33 |
*** rushiagr is now known as rushiagr_away | 08:34 | |
*** EinstCrazy has quit IRC | 08:35 | |
*** rushiagr_away is now known as rushiagr | 08:35 | |
*** rushiagr is now known as rushiagr_away | 08:41 | |
*** gongysh has quit IRC | 08:41 | |
*** rushiagr_away is now known as rushiagr | 08:42 | |
*** rushiagr is now known as rushiagr_away | 08:43 | |
*** jlanoux has joined #openstack-nova | 08:43 | |
*** fawadkhaliq has quit IRC | 08:46 | |
openstackgerrit | Tang Chen proposed openstack/nova: Add a REST API to trigger crash dump in an instance. https://review.openstack.org/202617 | 08:47 |
*** rushiagr_away is now known as rushiagr | 08:49 | |
*** shyama has joined #openstack-nova | 08:53 | |
*** rotbeard has joined #openstack-nova | 08:53 | |
*** rushiagr is now known as rushiagr_away | 08:54 | |
*** sacharya has joined #openstack-nova | 08:56 | |
*** fawadkhaliq has joined #openstack-nova | 08:56 | |
*** lpetrut has quit IRC | 08:56 | |
*** rushiagr_away is now known as rushiagr | 08:57 | |
*** lyanchih has joined #openstack-nova | 09:00 | |
*** sacharya has quit IRC | 09:00 | |
*** haomaiwang has quit IRC | 09:01 | |
*** haomaiwang has joined #openstack-nova | 09:01 | |
*** rushiagr is now known as rushiagr_away | 09:02 | |
*** rushiagr_away is now known as rushiagr | 09:04 | |
*** shyama has quit IRC | 09:04 | |
*** shyama has joined #openstack-nova | 09:04 | |
*** fawadkhaliq has quit IRC | 09:07 | |
*** davideagnello has joined #openstack-nova | 09:09 | |
openstackgerrit | Stephen Finucane proposed openstack/nova: Add 'CPUThreadAllocationPolicy' enum field https://review.openstack.org/202647 | 09:09 |
openstackgerrit | Stephen Finucane proposed openstack/nova: trivial: Add additional logs for NUMA scheduling https://review.openstack.org/254783 | 09:09 |
openstackgerrit | Stephen Finucane proposed openstack/nova: Make use of 'InstanceNUMACell.cpu_policy' field https://review.openstack.org/202653 | 09:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova: Add 'hw:cpu_threads_policy=isolate' scheduling https://review.openstack.org/202655 | 09:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova: Add 'cpu_policy' and 'cpu_thread_policy' fields https://review.openstack.org/202648 | 09:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova: Add 'hw:cpu_thread_policy=require' scheduling https://review.openstack.org/202651 | 09:10 |
*** ijw has quit IRC | 09:10 | |
*** zhangjn has joined #openstack-nova | 09:12 | |
*** sfinucan has joined #openstack-nova | 09:12 | |
*** rushiagr is now known as rushiagr_away | 09:13 | |
*** rushiagr_away is now known as rushiagr | 09:13 | |
*** shyama_ has joined #openstack-nova | 09:14 | |
openstackgerrit | Chaozhe Chen proposed openstack/python-novaclient: Suppress 'get_client_class' deprecated warnings in test_client https://review.openstack.org/259401 | 09:14 |
*** davideagnello has quit IRC | 09:14 | |
*** derekh has joined #openstack-nova | 09:16 | |
sfinucan | alaski: Could I get a workflow +1 again for a trivial rebase? https://review.openstack.org/#/c/202647/26 | 09:16 |
openstackgerrit | Jason proposed openstack/nova: Correct the manager.py code description https://review.openstack.org/264084 | 09:16 |
chaochin | alex_xu: hi | 09:16 |
*** ducttape_ has joined #openstack-nova | 09:16 | |
*** ankit_ag has joined #openstack-nova | 09:17 | |
chaochin | alex_xu: do you have time to have a look at my https://review.openstack.org/#/c/262165/ | 09:17 |
*** shyama has quit IRC | 09:17 | |
sfinucan | (Gerrit/Jenkins appeared to have broken for said change and refused to merge...) | 09:17 |
*** shyama_ is now known as shyama | 09:17 | |
*** jistr has joined #openstack-nova | 09:17 | |
alex_xu | chaochin: yea, will check now | 09:17 |
chaochin | alex_xu: you request me to create a spec for the bug. Not sure if that spec statisfy your request | 09:17 |
chaochin | alex_xu: thank you very much | 09:17 |
*** ccard has joined #openstack-nova | 09:18 | |
*** ssurana has quit IRC | 09:19 | |
*** yassine has joined #openstack-nova | 09:19 | |
alex_xu | chaochin: np, thanks for the spec | 09:19 |
*** ssurana has joined #openstack-nova | 09:19 | |
ankit_ag | Hi all, Can someone help reviewing https://review.openstack.org/#/c/247519/ | 09:21 |
*** ducttape_ has quit IRC | 09:21 | |
*** jlanoux has quit IRC | 09:22 | |
*** aix has joined #openstack-nova | 09:23 | |
*** jlanoux has joined #openstack-nova | 09:24 | |
*** ssurana has quit IRC | 09:24 | |
*** apoorvad has quit IRC | 09:24 | |
*** qianli has joined #openstack-nova | 09:25 | |
openstackgerrit | Stephen Finucane proposed openstack/nova: Forcing a configdrive in cfg should populate DB https://review.openstack.org/92899 | 09:26 |
*** thorst has joined #openstack-nova | 09:31 | |
*** ociuhandu has quit IRC | 09:34 | |
*** fawadkhaliq has joined #openstack-nova | 09:35 | |
*** ociuhandu has joined #openstack-nova | 09:35 | |
*** zhangjn has quit IRC | 09:35 | |
*** fawadkhaliq has quit IRC | 09:36 | |
*** thorst has quit IRC | 09:38 | |
*** jwcroppe has joined #openstack-nova | 09:39 | |
*** vilobhmm112 has quit IRC | 09:40 | |
*** EinstCra_ has quit IRC | 09:40 | |
*** EinstCrazy has joined #openstack-nova | 09:42 | |
*** fawadkhaliq has joined #openstack-nova | 09:43 | |
*** jwcroppe has quit IRC | 09:43 | |
*** ociuhandu has quit IRC | 09:44 | |
*** ptm_away is now known as PaulMurray | 09:44 | |
*** ijw has joined #openstack-nova | 09:46 | |
*** dedery has joined #openstack-nova | 09:50 | |
*** zhangjn has joined #openstack-nova | 09:51 | |
*** shyama has quit IRC | 09:54 | |
*** shyama has joined #openstack-nova | 09:54 | |
openstackgerrit | sahid proposed openstack/nova: libvirt: add policy check to avoid using qga in realtime context https://review.openstack.org/247585 | 09:57 |
openstackgerrit | sahid proposed openstack/nova: conductor: initialize request_spec to handle expected exception https://review.openstack.org/258011 | 09:57 |
openstackgerrit | sahid proposed openstack/nova: libvirt: add realtime support https://review.openstack.org/197569 | 09:58 |
openstackgerrit | Thomas Bechtold proposed openstack/nova: xen: Allow custom libvirt kernel in PV mode https://review.openstack.org/264101 | 09:58 |
*** dedery_ has joined #openstack-nova | 10:00 | |
*** haomaiwang has quit IRC | 10:01 | |
*** dedery__ has joined #openstack-nova | 10:01 | |
*** haomaiwang has joined #openstack-nova | 10:01 | |
*** dedery__ has quit IRC | 10:02 | |
*** zhangyufei has quit IRC | 10:02 | |
*** EinstCrazy has quit IRC | 10:04 | |
*** dedery has quit IRC | 10:04 | |
*** dedery__ has joined #openstack-nova | 10:04 | |
*** dedery_ has quit IRC | 10:04 | |
*** ndipanov has joined #openstack-nova | 10:05 | |
*** chhavi has quit IRC | 10:08 | |
ildikov | johnthetubaguy: good morning | 10:09 |
johnthetubaguy | ildikov: hi | 10:09 |
ildikov | johnthetubaguy: I changed the rpcapi to pass attachment_id, when we can | 10:09 |
*** fawadkhaliq has quit IRC | 10:09 | |
*** achinm has quit IRC | 10:09 | |
ildikov | johnthetubaguy: https://review.openstack.org/#/c/193134/57 | 10:10 |
*** otter768 has joined #openstack-nova | 10:10 | |
*** ijw has quit IRC | 10:10 | |
ildikov | johnthetubaguy: is this what you have in mind earlier? | 10:10 |
johnthetubaguy | ildikov: that sounds correct, to try and avoid the extra call to cinder, where possible. | 10:11 |
ildikov | johnthetubaguy: yeah, this is what I tried to do, it calls out only in that case if the attachment_id is missing | 10:12 |
*** ndipanov has quit IRC | 10:12 | |
ildikov | johnthetubaguy: the check is right before the call to Cinder so it's kinda centralized | 10:12 |
*** otter768 has quit IRC | 10:14 | |
johnthetubaguy | ildikov: I added a quick comment around leaving the compat code in the compute manger, so its eaiser to remove at the point where we always pass the attachement_id (when we next bump the major version) | 10:15 |
*** slunkad has quit IRC | 10:15 | |
*** zhangjn has quit IRC | 10:17 | |
*** chhavi has joined #openstack-nova | 10:17 | |
*** ducttape_ has joined #openstack-nova | 10:17 | |
ildikov | johnthetubaguy: we will need to call out to Cinder anyway, I mean in case of removing an instance etc you usually don't get the volume from the API | 10:18 |
*** zhangjn has joined #openstack-nova | 10:18 | |
*** ijw has joined #openstack-nova | 10:18 | |
*** lpetrut has joined #openstack-nova | 10:19 | |
ildikov | johnthetubaguy: the volume_api.detach call happens in multiple places and it looks messier to me to retrieve the attachment_id in each and every place we don't have it | 10:19 |
johnthetubaguy | ildikov: ah, true, that makes sense | 10:20 |
johnthetubaguy | not read the full patch, trying to work on some testing stuff this morning | 10:20 |
*** andreykurilin__ has joined #openstack-nova | 10:20 | |
ildikov | johnthetubaguy: I also wonder whether to pass the multiattach flag too | 10:20 |
ildikov | and if it's true AND we don't have attachment_id then call out to Cinder | 10:20 |
ildikov | johnthetubaguy: sorry to disturb | 10:20 |
johnthetubaguy | ildikov: no worries, PTL is a life of interruptions :) | 10:21 |
ildikov | I just wanted to double check the direction before doing further changes | 10:21 |
johnthetubaguy | totally | 10:21 |
ildikov | :) | 10:21 |
*** ducttape_ has quit IRC | 10:22 | |
johnthetubaguy | ildikov: I am just wondering if we should really store the attachment id in the BDM? not its not so clear how to detach with just the information in the BDM | 10:22 |
*** zhangjn has quit IRC | 10:22 | |
johnthetubaguy | ildikov: also maybe worth storing multi-attach in the BDM too, given how much we need that info | 10:23 |
ildikov | johnthetubaguy: I don't know about live migration and other cases where Cinder should be the ultimate source of truth | 10:23 |
*** ijw_ has joined #openstack-nova | 10:23 | |
johnthetubaguy | multi-attach is the less clear one for sure, but the attachment id feels like something cinder tells us, that we are meant to remember so we perform the correct detach | 10:24 |
ildikov | multi-attach should not change, I don't know what's the case with the attachment_id | 10:24 |
johnthetubaguy | I am not sure about these, just thinking out loud really | 10:24 |
ildikov | I can pass these two args now to detach and especially in the early phase I think it will keep the round trips low | 10:25 |
*** dguitarbite has joined #openstack-nova | 10:25 | |
ildikov | I mean I don't think many users would create volumes with the multiattach flag set to true nowadays for instance | 10:25 |
*** ijw has quit IRC | 10:26 | |
*** slunkad has joined #openstack-nova | 10:27 | |
*** cdent has joined #openstack-nova | 10:28 | |
ildikov | johnthetubaguy: later we can still add this info to the BDM as an optimization step, it can be included in the API spec for multiattach | 10:28 |
ildikov | anyway, I go back and deal with the cases we discussed yesterday | 10:28 |
ildikov | thanks for the quick check | 10:28 |
johnthetubaguy | ildikov: OK, its more the API level checks later on that worry me. The API shouldn't have to call out to cinder before it can return its response to the user | 10:31 |
*** ociuhandu has joined #openstack-nova | 10:31 | |
johnthetubaguy | ildikov: but for now, we can always fail at the compute layer anyways, so its not a massive issues, just I don't feel I can see the whole picture yet | 10:31 |
johnthetubaguy | no seeing the whole picture is my issue here, not anyone elses | 10:32 |
*** gjayavelu has joined #openstack-nova | 10:32 | |
*** gjayavelu has quit IRC | 10:32 | |
*** gjayavelu has joined #openstack-nova | 10:32 | |
ildikov | johnthetubaguy: I don't really see how the API is affected here | 10:33 |
johnthetubaguy | ildikov: we were talking about disabling certain actions if the volume is multi-attach, but actually it doesn't need the BDMs in those cases, I got it back to front | 10:34 |
ildikov | johnthetubaguy: a-ha, ok, you meant the not supported things with multi-attach | 10:34 |
johnthetubaguy | yeah | 10:34 |
*** ndipanov has joined #openstack-nova | 10:34 | |
*** gjayavelu has quit IRC | 10:35 | |
ildikov | the multiattach info is needed either from the volume info or from the BDM if we will store it there | 10:36 |
*** thorst has joined #openstack-nova | 10:36 | |
*** dedery has joined #openstack-nova | 10:36 | |
openstackgerrit | lyanchih proposed openstack/nova-specs: Spec to add resources-links in resource list API Response https://review.openstack.org/264114 | 10:37 |
ildikov | but for the actions we block that should be the only call | 10:37 |
*** dedery__ has quit IRC | 10:40 | |
*** thorst has quit IRC | 10:43 | |
*** fawadkhaliq has joined #openstack-nova | 10:44 | |
*** gjayavelu has joined #openstack-nova | 10:45 | |
*** fawadkhaliq has quit IRC | 10:45 | |
*** femnad has joined #openstack-nova | 10:45 | |
*** ljxiash__ has quit IRC | 10:47 | |
*** dims has joined #openstack-nova | 10:47 | |
*** sudipto has quit IRC | 10:47 | |
openstackgerrit | Andrea Rosa proposed openstack/nova: Wrong URL reported by the run_tests.sh message https://review.openstack.org/264119 | 10:48 |
*** lyanchih has quit IRC | 10:48 | |
dims | sdague : easy one from yesterday (flaky tests in py34) - https://review.openstack.org/#/c/263789/ - please peek when you get a chance | 10:49 |
*** BobBall has quit IRC | 10:50 | |
*** gcb has quit IRC | 10:52 | |
*** zhangyufei has joined #openstack-nova | 10:52 | |
*** houming_ has joined #openstack-nova | 10:52 | |
*** houming has quit IRC | 10:52 | |
*** houming_ is now known as houming | 10:52 | |
dims | thanks johnthetubaguy ! | 10:53 |
*** gcb has joined #openstack-nova | 10:54 | |
johnthetubaguy | dims: no worries, good to identify the broken ones like that, a nice middle ground | 10:54 |
*** fawadkhaliq has joined #openstack-nova | 10:54 | |
johnthetubaguy | dims: I kinda hope we track down a bug in mox3 if we find out whats common between those tests | 10:55 |
*** yassine has quit IRC | 10:55 | |
*** BobBall has joined #openstack-nova | 10:56 | |
openstackgerrit | Alex Xu proposed openstack/nova: doc: add version discovery guideline in api concept doc https://review.openstack.org/264044 | 10:56 |
openstackgerrit | Alex Xu proposed openstack/nova: doc: add client interactive guideline for microversions https://review.openstack.org/264045 | 10:56 |
openstackgerrit | Alex Xu proposed openstack/nova: doc: completes microversion use-cases in api concept doc https://review.openstack.org/263220 | 10:56 |
*** fawadkhaliq has quit IRC | 10:56 | |
*** klkumar has quit IRC | 10:59 | |
openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/nova: Keep py3.X compatibility for urllib https://review.openstack.org/261034 | 11:00 |
*** haomaiwang has quit IRC | 11:01 | |
*** sudipto has joined #openstack-nova | 11:01 | |
*** haomaiwang has joined #openstack-nova | 11:01 | |
*** gcb has quit IRC | 11:01 | |
*** ljxiash has joined #openstack-nova | 11:03 | |
shyama | ildikov: please look at https://review.openstack.org/#/c/264059/ when you have time | 11:03 |
ildikov | shyama: hi, tnx, just answered with further questions | 11:04 |
shyama | ildikov: yes. we need a check similar to what hemna is doing for remove export in terminate connection as well | 11:06 |
*** lucas-dinner is now known as lucasagomes | 11:06 | |
shyama | the problem is not all cinder drivers treat remove export and terminate_connection same way. since we pass the attached_host to the attachement table I think the same check in terminate_connection should be good enough | 11:07 |
ildikov | shyama: but you mean in Cinder or on Nova side? | 11:07 |
ildikov | shyama: yeah, I'm learning it on the way that all the drivers are slightly different... :S | 11:08 |
shyama | ildikov: if it goes into cinder it will be good. or we might need to think of a solution in the nova side which would not be the nicest thing to do | 11:08 |
lxsli | ndipanov, bauzas: either of you feel like +A https://review.openstack.org/#/c/206070/ please? | 11:08 |
ildikov | shyama: nah, I would like to keep everything in Cinder that belongs there if possible | 11:08 |
bauzas | lxsli: sure, lemme look | 11:09 |
shyama | we would need to discuss with hemna on this one | 11:09 |
ildikov | shyama: there's a topic for that one for the Cinder meeting today | 11:09 |
*** rfolco has joined #openstack-nova | 11:09 | |
ildikov | I think we can bring this up there too | 11:10 |
*** fawadkhaliq has joined #openstack-nova | 11:10 | |
ildikov | shyama: will you attend the meeting? | 11:10 |
shyama | ildikov: ok when is the meeting. I usually cannot make it due to time zone differences I will make it to the one today | 11:10 |
*** davideagnello has joined #openstack-nova | 11:10 | |
*** gjayavelu has quit IRC | 11:11 | |
*** klkumar has joined #openstack-nova | 11:12 | |
ildikov | shyama: 16:00 UTC | 11:15 |
*** davideagnello has quit IRC | 11:15 | |
*** smoriya_ has quit IRC | 11:15 | |
ildikov | shyama: which time zone are you in? | 11:15 |
shyama | ildikov: thanks! I work from India | 11:15 |
dims | johnthetubaguy : yep, figuring out mox is the right thing to do | 11:16 |
*** zhangjn has joined #openstack-nova | 11:16 | |
*** haomaiwang has quit IRC | 11:17 | |
ildikov | shyama: yeah, it's kinda late there at the time of the meeting | 11:17 |
ildikov | thanks for joining today! | 11:17 |
*** ducttape_ has joined #openstack-nova | 11:18 | |
shyama | ildikov: one other thing my colleague put up this blueprint proposal to edit the multiattach property of a volume after it is created. https://review.openstack.org/#/c/264059/ | 11:19 |
shyama | do take a look. | 11:19 |
*** gcb has joined #openstack-nova | 11:21 | |
dims | johnthetubaguy : for dropping the ec2/objectstore REST API review, i steered clear of any data model changes and still able to nuke about 10k worth of code - https://review.openstack.org/#/c/263368/ | 11:21 |
*** ducttape_ has quit IRC | 11:22 | |
ildikov | shyama: well, from Nova perspective it's not that good in the sense of there will be actions we cannot support with multiattach at least not in the first round | 11:23 |
dims | johnthetubaguy : filed a mox3 bug - https://bugs.launchpad.net/python-mox3/+bug/1531464 | 11:23 |
openstack | Launchpad bug 1531464 in python-mox3 "Fix flaky Nova tests" [Undecided,New] | 11:23 |
johnthetubaguy | dims: cool stuff | 11:24 |
johnthetubaguy | dims: did we hear back from the CERN folks about EC2 at all? | 11:24 |
shyama | ildikov: any specifc actions? | 11:24 |
ildikov | shyama: so if this makes it then we need to be sure in every case that we have the latest and greatest info about the volume from Cinder | 11:24 |
*** jwcroppe has joined #openstack-nova | 11:24 | |
ildikov | shyama: boot_from_volume, volume_snapshot_create | 11:24 |
*** fawadkhaliq has quit IRC | 11:24 | |
ildikov | and I'm sure there'll be more | 11:25 |
shyama | ildikov: do you have code check in the manager layer for multiattach volumes here? | 11:25 |
ildikov | I also don't know how live migration will work, especially with the Cinder bug with detach | 11:25 |
ildikov | that's what I will start working on today to add checks | 11:26 |
kashyap | markmc: When you're around, did you set any explicit flags in nova.conf in your testing (from comment#3 here)? - https://bugs.launchpad.net/nova/+bug/1441054 | 11:26 |
openstack | Launchpad bug 1441054 in OpenStack Compute (nova) "live-migration --block-migrate fails with default libvirt flags" [Medium,In progress] - Assigned to Mathieu Rohon (mathieu-rohon) | 11:26 |
dims | johnthetubaguy : searching email archives, no news from them for 10+ months | 11:26 |
ildikov | shyama: what's possible to check on the API layer that should happen there otherwise where possible | 11:27 |
openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/nova: Replace deprecated LOG.warn with LOG.warning https://review.openstack.org/264133 | 11:27 |
*** mriedem has joined #openstack-nova | 11:28 | |
shyama | ildikov: ok so when a volume is multiattach then you would not allow boot from volume etc., may be i missed this discussion in the review. can you brief me what would be the reason | 11:28 |
kashyap | markmc: I ask because, from my previous testing (a couple of months ago), I see something simliar to what you see, migration falling back to old implementation (QMP "inc":true approach), tested on Fedora-22 -- https://kashyapc.fedorapeople.org/virt/openstack/successful-logs-nova-live-block-migration/src-libv-log-of-nova-live-block-mig.txt | 11:28 |
*** cdent has quit IRC | 11:28 | |
ildikov | shyama: we discussed it yesterday here, could not get there yet to capture it in the review | 11:29 |
shyama | ildikov: ok let me go though the irc logs | 11:29 |
kashyap | markmc: (You'd see the same "Falling back to previous implementation" string, if you grep for it.) Anyhow - I'm testing afresh on Fedora-23 with latest git, and will add a comment on the bug from what I find. | 11:29 |
*** chhavi has quit IRC | 11:30 | |
*** dguitarbite has quit IRC | 11:31 | |
*** vladikr has quit IRC | 11:31 | |
ildikov | shyama: http://eavesdrop.openstack.org/irclogs/%23openstack-nova/%23openstack-nova.2016-01-05.log.html | 11:32 |
ildikov | shyama: it's around 15:00 when we get to those cases | 11:33 |
ildikov | shyama: like delete_on_termination, etc. | 11:33 |
*** sahid has quit IRC | 11:34 | |
shyama | ildikov: i do not think delete_on_termination should be a concern because the volume will not be deleted but the instance should be. | 11:35 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [python3] byte/string conversions and enable more tests https://review.openstack.org/261045 | 11:35 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [python3] Port more code to python3 and enable tests https://review.openstack.org/262083 | 11:35 |
*** ljxiash has quit IRC | 11:35 | |
ildikov | ildikov: I don't know how that works with Cinder today | 11:35 |
ildikov | shyama: | 11:35 |
*** ljxiash has joined #openstack-nova | 11:36 | |
ildikov | so basically the concept is to block everything we're not sure will work | 11:36 |
shyama | ildikov: from some testing we have done i think the instance delete will proceed. let me confirm which part of the code | 11:36 |
openstackgerrit | Stephen Finucane proposed openstack/nova: Config options: centralize "configdrive" options https://review.openstack.org/254092 | 11:37 |
shyama | ildikov: https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L2408 | 11:39 |
*** ljxiash has quit IRC | 11:40 | |
*** zhangyufei has quit IRC | 11:40 | |
*** thorst has joined #openstack-nova | 11:41 | |
*** jwcroppe has quit IRC | 11:42 | |
ildikov | shyama: do you mean that this will not delete the volume on Cinder side in case of multiattach: https://github.com/openstack/nova/blob/master/nova/compute/api.py#L1770 ? | 11:44 |
ildikov | shyama: I mean this: https://github.com/openstack/nova/blob/master/nova/volume/cinder.py#L425 | 11:45 |
shyama | ildikov: cinder will throw an exception because the volume is still attached but nova ignores that and moves on as the instance is already deleted | 11:45 |
*** chhavi has joined #openstack-nova | 11:46 | |
shyama | ildikov: i am pretty sure that all cinder drivers do the same in this case :) | 11:46 |
ildikov | shyama: are you sure there never be a case which moves the instance to error state? | 11:46 |
ildikov | shyama: at least one consistent case :) | 11:47 |
*** klkumar has quit IRC | 11:47 | |
*** coolsvap is now known as coolsvap|away | 11:47 | |
shyama | ildikov: https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L2353-L2365 if delete volume fails we do not throw an exception in the delete instance flow | 11:48 |
*** thorst has quit IRC | 11:48 | |
openstackgerrit | sandeep nandal proposed openstack/nova: Replacing range() n xrange() by six.moves.range() https://review.openstack.org/261006 | 11:50 |
openstackgerrit | sandeep nandal proposed openstack/nova: Replacing range() n xrange() by six.moves.range() https://review.openstack.org/264143 | 11:50 |
ildikov | shyama: yeah, it looks the case | 11:51 |
ildikov | shyama: I will also need to check that IRC log further for what and why to block first | 11:52 |
shyama | ildikov: I see boot from volume discussed not sure about others | 11:52 |
ildikov | unfortunately the spec does not cover all the cases and the deadline here is Jan 21 to get the code landed | 11:52 |
*** klkumar has joined #openstack-nova | 11:53 | |
ildikov | shyama: create_snapshot allso seemed to be problematic | 11:53 |
ildikov | so it would be good to have a sane first version landed before that pretty close deadline and then evolve it further | 11:53 |
*** renatoarmani has joined #openstack-nova | 11:54 | |
openstackgerrit | Marian Horban proposed openstack/nova: Stop Libvirt event threads https://review.openstack.org/259066 | 11:55 |
*** apoorvad has joined #openstack-nova | 11:57 | |
shyama | ildikov: I agree with create snapshot one, but boot from volume is not needed | 11:58 |
*** yassine has joined #openstack-nova | 11:59 | |
*** doug-fish has joined #openstack-nova | 11:59 | |
ildikov | shyama: you mean to be blocked? | 11:59 |
shyama | ildikov: yes | 12:01 |
ildikov | shyama: ok, I will re-open the discussion on that one, when I get there, tnx for the pointers | 12:01 |
shyama | ildikov: thank you! | 12:02 |
*** klkumar has quit IRC | 12:02 | |
*** klkumar has joined #openstack-nova | 12:02 | |
*** apoorvad has quit IRC | 12:06 | |
*** atuvenie has joined #openstack-nova | 12:06 | |
*** klkumar has quit IRC | 12:08 | |
openstackgerrit | Alexis Lee proposed openstack/nova: Catch errors while creating veth pair https://review.openstack.org/264146 | 12:09 |
*** cdent has joined #openstack-nova | 12:09 | |
*** jlanoux has quit IRC | 12:10 | |
*** otter768 has joined #openstack-nova | 12:11 | |
*** klkumar has joined #openstack-nova | 12:13 | |
*** pkholkin has quit IRC | 12:15 | |
*** otter768 has quit IRC | 12:15 | |
*** smatzek has joined #openstack-nova | 12:17 | |
*** obondarev_ has quit IRC | 12:17 | |
*** wuhg has quit IRC | 12:18 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [python3] byte/string conversions and enable more tests https://review.openstack.org/261045 | 12:19 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [python3] Port more code to python3 and enable tests https://review.openstack.org/262083 | 12:20 |
*** dstepanenko_ has quit IRC | 12:20 | |
*** zhangjn has quit IRC | 12:26 | |
openstackgerrit | Balazs Gibizer proposed openstack/nova: doc: add devref about versioned notifications https://review.openstack.org/263788 | 12:27 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Add service status notification https://review.openstack.org/245678 | 12:27 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Add infra for versioned notifications https://review.openstack.org/247024 | 12:27 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Adds json sample for the versioned notifications https://review.openstack.org/248167 | 12:27 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Make emitting versioned notifications configurable https://review.openstack.org/247564 | 12:27 |
openstackgerrit | ChangBo Guo(gcb) proposed openstack/nova: virt: refactor method compute_driver_matches https://review.openstack.org/263243 | 12:36 |
*** fawadkhaliq has joined #openstack-nova | 12:38 | |
*** fawadkhaliq has quit IRC | 12:39 | |
*** fawadkhaliq has joined #openstack-nova | 12:39 | |
*** jaypipes has joined #openstack-nova | 12:42 | |
*** claudiub has joined #openstack-nova | 12:42 | |
openstackgerrit | jichenjc proposed openstack/nova: Remove mox usage of test_quota https://review.openstack.org/264154 | 12:42 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: doc: add devref about versioned notifications https://review.openstack.org/263788 | 12:44 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Add service status notification https://review.openstack.org/245678 | 12:44 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Add infra for versioned notifications https://review.openstack.org/247024 | 12:44 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Adds json sample for the versioned notifications https://review.openstack.org/248167 | 12:44 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Make emitting versioned notifications configurable https://review.openstack.org/247564 | 12:44 |
*** pradk has quit IRC | 12:45 | |
*** zhangjn has joined #openstack-nova | 12:45 | |
*** zhangjn has quit IRC | 12:45 | |
*** moshele has quit IRC | 12:45 | |
*** zhangjn has joined #openstack-nova | 12:46 | |
*** zhangjn has quit IRC | 12:46 | |
sdague | johnthetubaguy / alex_xu / et al: can we get this landed - https://review.openstack.org/#/c/263798/ because it will stop throwing a bunch of gorp into logstash | 12:46 |
*** zhangjn has joined #openstack-nova | 12:47 | |
*** w_verdugo has joined #openstack-nova | 12:47 | |
*** moshele has joined #openstack-nova | 12:49 | |
openstackgerrit | Jialiang proposed openstack/nova: Update the file for IPv4-only or IPv6-only network https://review.openstack.org/264157 | 12:52 |
*** cdelatte has joined #openstack-nova | 12:53 | |
*** EinstCrazy has joined #openstack-nova | 12:53 | |
*** sahid has joined #openstack-nova | 12:54 | |
*** houming has quit IRC | 12:54 | |
*** haomaiwang has joined #openstack-nova | 12:56 | |
*** sacharya has joined #openstack-nova | 12:57 | |
bauzas | lxsli: johnthetubaguy: made a few comments on https://review.openstack.org/#/c/206070/11 | 12:58 |
*** klkumar has quit IRC | 12:59 | |
bauzas | unless I'm blind, we are mising some hypertext plus I have a problem when we give an etherpad link w/ a line number | 12:59 |
bauzas | that said, we can merge that very soon | 12:59 |
alex_xu | sdague: just +w for it | 13:00 |
sdague | alex_xu: thanks | 13:00 |
alex_xu | sdague: np | 13:00 |
*** haomaiwang has quit IRC | 13:01 | |
*** sacharya has quit IRC | 13:02 | |
*** thorst has joined #openstack-nova | 13:02 | |
markmc | kashyap, right, I just tested with TUNNELLED and NON_SHARED_INC | 13:05 |
markmc | kashyap, you're saying you did see this work fine too, then? | 13:05 |
*** baoli has joined #openstack-nova | 13:05 | |
kashyap | markmc: 1 moment | 13:05 |
*** zz_dimtruck is now known as dimtruck | 13:06 | |
*** baoli has quit IRC | 13:06 | |
kashyap | markmc: When I tested, I had these flags explicitly set in nova.conf: | 13:07 |
kashyap | live_migration_flag = VIR_MIGRATE_UNDEFINE_SOURCE, | 13:07 |
kashyap | VIR_MIGRATE_PEER2PEER, VIR_MIGRATE_LIVE | 13:07 |
kashyap | block_migration_flag = VIR_MIGRATE_UNDEFINE_SOURCE, | 13:07 |
kashyap | VIR_MIGRATE_PEER2PEER, VIR_MIGRATE_LIVE, VIR_MIGRATE_NON_SHARED_INC | 13:07 |
*** baoli has joined #openstack-nova | 13:07 | |
kashyap | Thus, overriding the default (which _does_ have TUNNELLED, as you know). | 13:07 |
kashyap | markmc: I'm about to setup a fresh env. on F23 with default flags and re-test, but just shaving a bunch of yaks with DevStack. | 13:08 |
dims | markmc : nice to see you again! | 13:08 |
*** EinstCrazy has quit IRC | 13:09 | |
*** EinstCrazy has joined #openstack-nova | 13:11 | |
markmc | dims, thanks :) | 13:11 |
*** baoli has quit IRC | 13:11 | |
*** EinstCrazy has quit IRC | 13:12 | |
markmc | kashyap, right, that bug is specifically about whether NON_SHARED_INC works with TUNNELLED | 13:12 |
johnthetubaguy | bauzas: honestly, with etherpads, the line number is always wrong, I am really just thinking about getting this in ASAP, and evolving it fast | 13:12 |
markmc | kashyap, which it does :) | 13:12 |
bauzas | johnthetubaguy: sure, hence me saying we should just give the section name, not the line number | 13:12 |
johnthetubaguy | bauzas: ah, totally makes sense | 13:13 |
bauzas | the other comments are mostly nits and rendering problems, we can quickly move forward, honestly | 13:14 |
*** gjayavelu has joined #openstack-nova | 13:14 | |
kashyap | markmc: Yeah. ETOOMANYFLAGS, sorry if I confused you :-) | 13:14 |
*** baoli has joined #openstack-nova | 13:16 | |
openstackgerrit | sahid proposed openstack/nova: libvirt: add policy check to avoid using qga in realtime context https://review.openstack.org/247585 | 13:16 |
openstackgerrit | sahid proposed openstack/nova: conductor: initialize request_spec to handle expected exception https://review.openstack.org/258011 | 13:16 |
openstackgerrit | sahid proposed openstack/nova: libvirt: add realtime support https://review.openstack.org/197569 | 13:16 |
openstackgerrit | sahid proposed openstack/nova: hardware: check whether realtime capable in API https://review.openstack.org/264165 | 13:16 |
*** kotra03 has quit IRC | 13:17 | |
*** baoli_ has joined #openstack-nova | 13:17 | |
johnthetubaguy | bauzas: the links are working I think, that is a reference to the __ above and named href below, I think | 13:17 |
*** ducttape_ has joined #openstack-nova | 13:18 | |
bauzas | johnthetubaguy: okay, that's probably me who is not reading correctly | 13:18 |
openstackgerrit | Adelina Tuvenie proposed openstack/nova: Moves OVS related code to a common location https://review.openstack.org/179727 | 13:18 |
openstackgerrit | Adelina Tuvenie proposed openstack/nova: Adds Hyper-V OVS ViF driver https://review.openstack.org/140045 | 13:18 |
bauzas | johnthetubaguy: but IIRC, the "nova people" tag references the wikipage link at the end of the doc, right? | 13:18 |
bauzas | johnthetubaguy: if so, it should show the URL nope? | 13:19 |
*** ducttape_ has quit IRC | 13:19 | |
*** gjayavelu has quit IRC | 13:19 | |
*** ducttape_ has joined #openstack-nova | 13:19 | |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: deprecate live/block_migration_flag opts https://review.openstack.org/263436 | 13:19 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: remove the live/block_migration_flag opts https://review.openstack.org/263437 | 13:19 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: add a new live_migration_tunneled config https://review.openstack.org/263434 | 13:19 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: add live_migration_extra_flags config option https://review.openstack.org/263435 | 13:19 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: force use of direct vs p2p migration https://review.openstack.org/263432 | 13:19 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: force config related migration flags https://review.openstack.org/263433 | 13:19 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: parse live migration flags at startup https://review.openstack.org/263430 | 13:19 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: force use/non-use of NON_SHARED_INC flag https://review.openstack.org/263431 | 13:19 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: better error for bad live migration flag https://review.openstack.org/264167 | 13:19 |
*** baoli has quit IRC | 13:20 | |
*** dslev has joined #openstack-nova | 13:22 | |
*** lyanchih has joined #openstack-nova | 13:23 | |
eliqiao | hi markmc, there ?one question about your patch https://review.openstack.org/#/c/263430/2/nova/virt/libvirt/driver.py | 13:24 |
openstackgerrit | Harshada Mangesh Kakad proposed openstack/nova: Replace deprecated library function os.popen() with subprocess https://review.openstack.org/262570 | 13:24 |
*** mdrabe has joined #openstack-nova | 13:26 | |
johnthetubaguy | bauzas: it does I think, look at the mentoring czar, it correctly links to the wiki | 13:27 |
*** dane-fichter has joined #openstack-nova | 13:28 | |
dane-fichter | If I'm incrementing the the version number of ImageMetaProps from 1.9, does it go to 1.10 or to 2.0 ? | 13:29 |
*** dimtruck is now known as zz_dimtruck | 13:31 | |
bauzas | johnthetubaguy: indeed, but the `Nova People` doesn't if you see my comments | 13:32 |
bauzas | that said, it's a nit | 13:32 |
johnthetubaguy | bauzas: let me re-read them, I must be missing something | 13:32 |
*** links has quit IRC | 13:32 | |
*** jlanoux has joined #openstack-nova | 13:33 | |
*** zhangjn has quit IRC | 13:33 | |
johnthetubaguy | bauzas: that __`Nova People`_ bit is to link `Mentoring Czar`__. to .. _`Nova people`: https://wiki.openstack.org/wiki/Nova#People I belive | 13:33 |
bauzas | johnthetubaguy: if lxsli changes the etherpad line number to a section name, we can just land it and discuss later | 13:33 |
*** edmondsw has joined #openstack-nova | 13:34 | |
bauzas | I don't want to cut hair :) | 13:34 |
markmc | eliqiao, yes? | 13:34 |
*** mdrabe has quit IRC | 13:37 | |
eliqiao | markmc: just comments in lines. | 13:37 |
eliqiao | markmc: https://review.openstack.org/#/c/263430/2/nova/virt/libvirt/driver.py | 13:38 |
*** jwcroppe has joined #openstack-nova | 13:38 | |
*** zhangjn has joined #openstack-nova | 13:38 | |
*** dslev has quit IRC | 13:38 | |
*** zhangjn has quit IRC | 13:39 | |
*** ducttape_ has quit IRC | 13:40 | |
*** rushiagr is now known as rushiagr_away | 13:40 | |
*** eharney has joined #openstack-nova | 13:43 | |
*** dave-mccowan has joined #openstack-nova | 13:45 | |
lxsli | bauzas: thanks for the review, responded | 13:45 |
*** ankit_ag has quit IRC | 13:48 | |
johnthetubaguy | lxsli: are we good to fix that line number, then we can try get it in I think? | 13:49 |
lxsli | I'd rather not, as the point of the patch is to be a forklift | 13:49 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: Handle race in allocate_for_instance https://review.openstack.org/221803 | 13:50 |
bauzas | lxsli: sure, but I also took the opportunity to just modify that | 13:50 |
*** jwcroppe has quit IRC | 13:50 | |
lxsli | I'd put up a followup but I'm unsure it's good advice, for something to make it onto the trivial bug list it must have a +1 | 13:50 |
bauzas | lxsli: tbh, I just saw that and I can modify the wiki now | 13:50 |
lxsli | that doesn't help >.< | 13:50 |
lxsli | Afterwards someone else will come along and point out another content error and the patch never lands | 13:51 |
bauzas | meh | 13:51 |
lxsli | we should land it as a forklift, THEN tinker with content | 13:51 |
*** vladikr has joined #openstack-nova | 13:51 | |
lxsli | I do totally agree it needs to be fixed, but how to fix it is up for debate right now, it's not a 20 second fix | 13:51 |
bauzas | it's not a 20 sec fix ? orly? | 13:52 |
bauzas | lemme try to play with the Gerrit UI | 13:52 |
lxsli | Well you want to put a section name there and I want to remove the advice | 13:52 |
lxsli | and when have our convos ever taken 20sec? :D | 13:52 |
bauzas | meh, you're good anyway :) | 13:52 |
lxsli | ahh sdague appears out of left field :) | 13:53 |
sdague | johnthetubaguy: do you want to give https://review.openstack.org/#/c/199205/42 a glance again, I just went through it and it seems fine to me | 13:53 |
openstackgerrit | dane-fichter proposed openstack/nova: Add image signature verification https://review.openstack.org/189843 | 13:54 |
lxsli | so johnthetubaguy - do you agree asking new contributors to +1 trivial patches (which already have +1) is a bad use of their time? | 13:54 |
johnthetubaguy | sdague: ah, yes, looking now | 13:54 |
lxsli | they're unlikely to learn all that much from trivial patches either imo | 13:54 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: api: add soft-affinity policies for server groups https://review.openstack.org/147520 | 13:55 |
johnthetubaguy | lxsli: whats this in reference too, that doc? | 13:55 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: scheduler: fill RequestSpec.instance_group.members https://review.openstack.org/148277 | 13:55 |
openstackgerrit | Balazs Gibizer proposed openstack/nova: scheduler: add soft-(anti-)affinity weighers https://review.openstack.org/147516 | 13:55 |
*** zhangjn has joined #openstack-nova | 13:56 | |
lxsli | yep, the line # issue | 13:56 |
lxsli | bauzas suggested changing the line # to a section name, but on reading it, the advice seems just bad | 13:56 |
bauzas | so that's all the problem about having a persistent link for an etherpad | 13:57 |
bauzas | I'm not really sold either | 13:57 |
sdague | lxsli / bauzas on docs patches I prefer we merge improvements more quickly. English is never perfect. And random walking towards better is better than the docs being really bad until we have a perfect replacement. | 13:57 |
bauzas | I just feel it's less errorprone | 13:57 |
lxsli | sdague: +1 | 13:57 |
bauzas | sdague: I certainly share your views on that, hence me not -1ing | 13:58 |
lxsli | bauzas: section name or line # is irrelevant if we're going to delete the paragraph | 13:58 |
*** mgoddard_ has joined #openstack-nova | 13:58 | |
lxsli | although I do prefer section name | 13:58 |
bauzas | lxsli: agreed, that just sucks less than providing a line number, that's it :( | 13:58 |
sdague | johnthetubaguy: did we agree we're dropping the ec2 stack right now? - https://review.openstack.org/#/c/263368/ ? | 13:59 |
bauzas | but my goal wasn't to split hairs, just to mention something I was feeling wrong :) | 13:59 |
*** jwcroppe has joined #openstack-nova | 13:59 | |
lxsli | yes thank you, I'm working on a followup now | 13:59 |
sdague | yeh, follow up seems fine | 13:59 |
johnthetubaguy | sdague: I don't remember us deciding, I thought we were waiting for CERN to say it worked for them, if possible | 14:00 |
sdague | johnthetubaguy: ok, well dropping the tests isn't really an option if we aren't going to drop the code | 14:00 |
sdague | I'm going to -2 that patch for now | 14:00 |
*** annegentle has joined #openstack-nova | 14:01 | |
*** haomaiwang has joined #openstack-nova | 14:01 | |
*** klkumar has joined #openstack-nova | 14:01 | |
*** mgoddard has quit IRC | 14:01 | |
johnthetubaguy | sdague: yeah, I think we said remove before the end of the release if there is no news, but it seems a touch early, good call | 14:02 |
sdague | docs are also one of those places where we don't get enough contributions, so I'd rather say "thanks" and merge them fast unless they are definitively wrong | 14:02 |
*** zenoway has joined #openstack-nova | 14:02 | |
bauzas | sdague: see https://review.openstack.org/#/c/229250/6/doc/source/filter_scheduler.rst then :) | 14:03 |
johnthetubaguy | lxsli: yeah, I see what you mean on that now, I suspect the other things on that etherpad would be more interesting, find an sub team you are interested in, and help them out | 14:03 |
*** rlrossit has joined #openstack-nova | 14:03 | |
johnthetubaguy | lxsli: the trivial ones have got a bit trivial to be useful, in many ways | 14:03 |
bauzas | I'm not totally sold on that | 14:04 |
sdague | bauzas: done | 14:04 |
bauzas | for newcomers, it can be interesting to review a very easy patch if they take the opportunity to expand the code and read the other bits | 14:05 |
lxsli | johnthetubaguy: part of that is that markus_z + I have been a bit too busy to spend much time looking for fixes | 14:05 |
lxsli | The new Gerrit will help, I need to put some time aside to fill the list back up | 14:05 |
bauzas | of course, +1ing for the sake of gamifying doesn't really help yourself, but you can have good pointers for learning code | 14:05 |
*** atuvenie has quit IRC | 14:05 | |
ndipanov | bauzas, I disagree - we should make it possible to do actual work in Nova even if you don't spend years on it | 14:05 |
ndipanov | fsck typos | 14:05 |
*** zenoway has quit IRC | 14:06 | |
lxsli | I was going to suggest looking for -1 and -2d reviews and reading the problems | 14:06 |
lxsli | that should give some idea of the traps not to fall into and things to question while reviewing | 14:06 |
*** zenoway has joined #openstack-nova | 14:06 | |
bauzas | ndipanov: I don't see where we disagree, my point is that people can come to Nova by reviewing easy tiny changes and leave their vote | 14:06 |
bauzas | that would help them to read some code sections | 14:07 |
dims | sdague : only tests that were removed were testing the REST API that we removed (https://review.openstack.org/#/c/263368/) | 14:07 |
bauzas | if the change itself isn't trivial, that's extra overhead in mind to understand | 14:07 |
sdague | dims: we didn't remove the REST API | 14:07 |
kashyap | ndipanov: Sorry for randomly jumping in, while at it, look at this response from Greg Kroah Hartman about new contributors and spelling fixes - https://lkml.org/lkml/2015/9/18/94 | 14:07 |
*** pkholkin has joined #openstack-nova | 14:07 | |
sdague | we disabled it by default | 14:07 |
sdague | there is a difference | 14:08 |
sdague | hmmm... wait, we did remove it? | 14:08 |
bauzas | kashyap: here, I'm talking about the Trivial Bug list | 14:08 |
bauzas | kashyap: not about cosmetic changes | 14:08 |
*** femnad has quit IRC | 14:09 | |
bauzas | kashyap: ie. there is a real bug behind the patch | 14:09 |
bauzas | that can be easily solved | 14:09 |
kashyap | ndipanov: But I don't disagree with your subtext of your comment though. | 14:09 |
sdague | dims: ok, so I didn't realize that ec2 was full on deleted from the tree | 14:09 |
dims | sdague : https://review.openstack.org/#/c/232325/ | 14:09 |
sdague | yep, I'm looking now | 14:09 |
sdague | I guess I had assumed we had just removed the paste.ini entries | 14:10 |
ndipanov | well apart from "your full of crap" I don't see how that guy disagrees with the mail he's responding to | 14:10 |
dims | sdague : y we did that in liberty | 14:10 |
kashyap | bauzas: Yeah, I agree that those who just go only for pimping stats ought to be ignored like the plague | 14:10 |
*** sudipto has quit IRC | 14:10 | |
*** zenoway has quit IRC | 14:11 | |
bauzas | honestly, I don't really feel it's a problem | 14:11 |
ndipanov | kashyap, my point is that we should not make typos a thing more than it should be eventually you should move on, and that's where we should focus our effort on making it easy to do meaningful work | 14:12 |
*** otter768 has joined #openstack-nova | 14:12 | |
ndipanov | should be _comma_ eventually * | 14:12 |
*** dslevin_ has joined #openstack-nova | 14:12 | |
bauzas | here, I'm talking about the opportunity to review an easy change, not to implement it | 14:12 |
bauzas | for a newcomer | 14:12 |
*** mdrabe has joined #openstack-nova | 14:12 | |
johnthetubaguy | sdague: I guess we did decide then | 14:12 |
kashyap | ndipanov: Sure, I'm with you on that, maybe I didn't fully digest your response and made a knee-jerk comment. | 14:13 |
sdague | johnthetubaguy: ok, is nova meeting tomorrow early or late? | 14:13 |
sdague | I kind of feel like we should at least huddle up there and acknowledge this is what we are doing | 14:13 |
johnthetubaguy | sdague: outlook tells me early, I guess | 14:13 |
sdague | dims: mind if I wait until post nova meeting to remove my -2? | 14:14 |
sdague | just so we're all sure on this | 14:14 |
dims | sdague : no worries at all | 14:14 |
johnthetubaguy | sdague: lets just make sure we are all on board, thats a good plan | 14:14 |
bauzas | the ICS cal has been fixed | 14:14 |
*** kotra03 has joined #openstack-nova | 14:14 | |
bauzas | that's an odd week | 14:14 |
johnthetubaguy | bauzas: hmm, I guess I have just been doing alternating, rather than odd vs even | 14:15 |
bauzas | johnthetubaguy: ttx told about that in -dev | 14:15 |
bauzas | lemme find it | 14:15 |
johnthetubaguy | bauzas: yeah, I remember someone mentioning that also | 14:15 |
*** zz_dimtruck is now known as dimtruck | 14:15 | |
bauzas | http://lists.openstack.org/pipermail/openstack-dev/2016-January/083216.html | 14:15 |
bauzas | johnthetubaguy: ^ | 14:16 |
*** otter768 has quit IRC | 14:16 | |
bauzas | so it's a late nova meeting for us | 14:16 |
johnthetubaguy | bauzas: oh, my, OK | 14:18 |
*** dstepanenko_ has joined #openstack-nova | 14:18 | |
bauzas | johnthetubaguy: we should probably publicize that, nope? | 14:19 |
*** changbl has quit IRC | 14:19 | |
johnthetubaguy | bauzas: yeah, was just updating the wiki and sending an ML note, unless you fancy sorting that out for me? | 14:19 |
*** ducttape_ has joined #openstack-nova | 14:20 | |
*** zhangjn has quit IRC | 14:20 | |
bauzas | your call | 14:21 |
bauzas | I mean, it's okay | 14:21 |
bauzas | doing that now | 14:21 |
*** chhavi has quit IRC | 14:21 | |
*** ljxiash has joined #openstack-nova | 14:22 | |
*** links has joined #openstack-nova | 14:22 | |
johnthetubaguy | bauzas: sweet, appreciate a hand with that, I updated the wiki, but not sent the email yet | 14:22 |
bauzas | ack | 14:23 |
johnthetubaguy | bauzas: digging on the feature classification stuff right now, trying to get something more for that | 14:23 |
*** whenry has joined #openstack-nova | 14:23 | |
bauzas | yep | 14:23 |
*** chhavi has joined #openstack-nova | 14:23 | |
*** cdlhh has joined #openstack-nova | 14:26 | |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: deprecate live/block_migration_flag opts https://review.openstack.org/263436 | 14:26 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: remove the live/block_migration_flag opts https://review.openstack.org/263437 | 14:26 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: add a new live_migration_tunneled config https://review.openstack.org/263434 | 14:26 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: add live_migration_extra_flags config option https://review.openstack.org/263435 | 14:26 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: force use of direct vs p2p migration https://review.openstack.org/263432 | 14:26 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: force config related migration flags https://review.openstack.org/263433 | 14:26 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: parse live migration flags at startup https://review.openstack.org/263430 | 14:26 |
openstackgerrit | Mark McLoughlin proposed openstack/nova: libvirt: force use/non-use of NON_SHARED_INC flag https://review.openstack.org/263431 | 14:26 |
cdlhh | hello | 14:26 |
cdlhh | 我是一个新人 | 14:27 |
openstackgerrit | Alexis Lee proposed openstack/nova: Fix advice for new contribs https://review.openstack.org/264195 | 14:27 |
*** jichen has quit IRC | 14:27 | |
cdlhh | hello everyone! | 14:27 |
*** ducttape_ has quit IRC | 14:28 | |
*** links has quit IRC | 14:28 | |
*** cdent has quit IRC | 14:29 | |
*** dimtruck is now known as zz_dimtruck | 14:29 | |
*** lucasagomes is now known as lucas-hungry | 14:30 | |
lxsli | hi cdlhh | 14:32 |
cdlhh | hi lxsli | 14:32 |
cdlhh | where is from? | 14:32 |
cdlhh | lxsli | 14:33 |
lxsli | England | 14:33 |
cdlhh | oh,I'm from china | 14:34 |
lxsli | something gave it away :) | 14:34 |
cdlhh | can you speak chinesse? | 14:35 |
cdlhh | haha | 14:35 |
cdlhh | chinese? | 14:36 |
eliqiao | ä½ å¥½æ–°äºº cdlhh :) | 14:39 |
*** rk4n has joined #openstack-nova | 14:39 | |
cdlhh | 哈哈 eliqiao | 14:40 |
cdlhh | ä½ æ˜¯å“ªé‡Œï¼Ÿ | 14:40 |
*** mgoddard has joined #openstack-nova | 14:41 | |
*** mgoddard_ has quit IRC | 14:41 | |
*** smatzek has quit IRC | 14:42 | |
openstackgerrit | Alexis Lee proposed openstack/nova: Catch errors while creating veth pair https://review.openstack.org/264146 | 14:42 |
*** pradk has joined #openstack-nova | 14:42 | |
*** zenoway has joined #openstack-nova | 14:42 | |
eliqiao | cdlhh: am from BJ, but please talk with English for other openstacker who can not understand Chinese :) | 14:43 |
cdlhh | yes thank you tip | 14:43 |
dansmith | mriedem: sdague: These are rebased, feedback addressed, passing jenkins: https://review.openstack.org/#/c/247853/18 | 14:44 |
dansmith | just three more to go :) | 14:44 |
mriedem | ok | 14:44 |
*** rushiagr_away is now known as rushiagr | 14:44 | |
*** burt has joined #openstack-nova | 14:48 | |
*** zenoway has quit IRC | 14:49 | |
*** kotra03 has quit IRC | 14:50 | |
*** ducttape_ has joined #openstack-nova | 14:50 | |
*** bklei has joined #openstack-nova | 14:51 | |
*** annegentle has quit IRC | 14:55 | |
*** annegentle has joined #openstack-nova | 14:55 | |
*** ccarmack has joined #openstack-nova | 14:56 | |
*** lmtaylor has joined #openstack-nova | 14:58 | |
openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/nova: Replace deprecated LOG.warn with LOG.warning https://review.openstack.org/264133 | 14:59 |
*** haomaiwang has quit IRC | 15:01 | |
*** 77CAAHDEK has joined #openstack-nova | 15:02 | |
*** ctrath has joined #openstack-nova | 15:04 | |
*** smatzek has joined #openstack-nova | 15:04 | |
*** cdent has joined #openstack-nova | 15:07 | |
openstackgerrit | Nikola Dipanov proposed openstack/nova: pci: do not filter out any SRIOV Physical Functions https://review.openstack.org/249016 | 15:08 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: pci: changing the claiming and allocation logic for PF/VF assignment https://review.openstack.org/249018 | 15:08 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: pci: adding support to specify a device_type in pci requests https://review.openstack.org/259254 | 15:08 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: objects: adding a parent_addr field to the PciDevice object https://review.openstack.org/249015 | 15:08 |
ndipanov | dansmith, ^ objects data mig | 15:08 |
ndipanov | would be awesome if you could take a look pls | 15:08 |
dansmith | okay | 15:08 |
openstackgerrit | dane-fichter proposed openstack/nova: Replaces xrange() with six.moves.range() to provide py3 compatibility https://review.openstack.org/205307 | 15:08 |
*** zenoway has joined #openstack-nova | 15:10 | |
*** zz_dimtruck is now known as dimtruck | 15:10 | |
edleafe | bauzas: I have a question from another project that wants to be able to call the scheduler's select_destinations() | 15:11 |
edleafe | bauzas: is RPC from an external project acceptable? | 15:11 |
dansmith | edleafe: no | 15:11 |
edleafe | dansmith: yeah, I didn't think so either | 15:11 |
bauzas | what dansmith said | 15:11 |
*** davideagnello has joined #openstack-nova | 15:12 | |
*** annegentle has quit IRC | 15:12 | |
edleafe | bauzas: what they really need is more like check_desination() | 15:12 |
ndipanov | which is almost ironic seeing that our rpc versioning is far superior to our API versioning... | 15:12 |
edleafe | how can an external project access the nova scheduler for stuff like that? | 15:12 |
edleafe | ndipanov: :) | 15:12 |
ndipanov | edleafe, :) | 15:13 |
bauzas | edleafe: heh, you know there is a BP about that ? | 15:13 |
*** mrkz has joined #openstack-nova | 15:14 | |
edleafe | bauzas: I know there was *talk* about that, but have no idea if it was decided | 15:14 |
bauzas | edleafe: http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/check-destination-on-migrations.html | 15:14 |
bauzas | which is basically, get the Spec object when the migration is called, and ask the scheduler anywat | 15:15 |
bauzas | anyway | 15:15 |
edleafe | bauzas: that doesn't add any external interface, though, right? | 15:15 |
bauzas | but for many cloudy reasons, just queying the scheduler is not good | 15:15 |
edleafe | bauzas: IOW, it's all internal to noca | 15:15 |
edleafe | nova even | 15:15 |
*** shyama has quit IRC | 15:15 | |
bauzas | yup | 15:16 |
alaski | edleafe: there's been talk of eventually having an HTTP interface for the scheduler, but as far as I know there's no movement on that | 15:16 |
bauzas | right | 15:16 |
cdent | the boundaries between projects seems rather artificial or arbitrary | 15:16 |
edleafe | alaski: yeah, I haven't heard anything about that since Gantt died | 15:16 |
cdent | hashtag we are all openstack, or some such | 15:16 |
*** davideagnello has quit IRC | 15:16 | |
bauzas | it was discussed to have a separate endpoint for calling the scheduler, but that needs a lot of discussion | 15:16 |
edleafe | cdent: hashtag we are all running the same queue | 15:17 |
cdent | hashtag totes | 15:17 |
bauzas | just because giving a response doesn't mean you'll be able to have your instance being created on that destination | 15:17 |
*** zenoway has quit IRC | 15:17 | |
edleafe | bauzas: sure, that's always the case | 15:17 |
bauzas | it's more a capacity problem IMHO | 15:17 |
ndipanov | bauzas, that logic should be part of the client lib | 15:17 |
alaski | edleafe: I still hear the idea tossed around, but not often anymore given the pressing work that's already in progress | 15:17 |
edleafe | bauzas: but it's more for ruling out hosts that could not accept an instance | 15:17 |
edleafe | bauzas: capacity?? | 15:18 |
bauzas | indeed | 15:18 |
bauzas | I mean | 15:18 |
bauzas | you want to get a destination | 15:18 |
*** Raji_ has joined #openstack-nova | 15:19 | |
bauzas | it's asynchronous, which means that this destination could be fine at *this* moment, not later | 15:19 |
*** xyang2 has joined #openstack-nova | 15:19 | |
openstackgerrit | Nikola Dipanov proposed openstack/nova: neutron: pci_request logic considers 'direct-physical' vnic type https://review.openstack.org/262584 | 15:20 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: pci: Add utility method for getting the MAC addr https://review.openstack.org/262585 | 15:20 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: neutron: Update the port with a MAC address for PFs https://review.openstack.org/262586 | 15:20 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: Add a vnic type for PF passthrough and a new libvirt vif driver https://review.openstack.org/262583 | 15:20 |
edleafe | bauzas: oh, the optimistic stuff. I thought you were talking about the capacity of the scheduler | 15:20 |
*** kotra03 has joined #openstack-nova | 15:20 | |
bauzas | no, about the cloud capacity | 15:20 |
openstackgerrit | Ryan Moore proposed openstack/nova: Add better help to rdp options https://review.openstack.org/253556 | 15:20 |
bauzas | so, having a client calling the scheduler for getting a destination is exactly like the conductor doing that | 15:21 |
bauzas | but the retry logic isn't there | 15:21 |
ndipanov | bauzas, it could be | 15:21 |
edleafe | bauzas: understood | 15:21 |
bauzas | rather in the compute manager | 15:21 |
ndipanov | retry logic is 2 lines of python | 15:21 |
bauzas | which calls again the conductor is something is excepted | 15:21 |
bauzas | so the conductor is dumb of the retry logic | 15:22 |
bauzas | although it's the scheduler client | 15:22 |
edleafe | bauzas: so what is that status on the migration's checking the destination? | 15:22 |
edleafe | s/that/the | 15:22 |
*** thangp has joined #openstack-nova | 15:22 | |
bauzas | there are lots of things to implement | 15:22 |
bauzas | I feel it's as big as implementing the RequestSpec objectification itself | 15:23 |
bauzas | some bits can also be mutualized for other reasons, mostly cells v2 | 15:23 |
bauzas | altough we don't care about migrating a running instance in the cells v2 world | 15:24 |
edleafe | bauzas: because of the persisting of the ReqSpec? | 15:24 |
bauzas | correct | 15:24 |
*** ccarmack has left #openstack-nova | 15:24 | |
*** Raji_ has quit IRC | 15:24 | |
*** ccarmack has joined #openstack-nova | 15:24 | |
edleafe | bauzas: what is the status of that? | 15:24 |
bauzas | the object now has a save method | 15:25 |
bauzas | thanks to alaski | 15:25 |
bauzas | (and a getter) | 15:25 |
*** ssurana has joined #openstack-nova | 15:25 | |
bauzas | now, we need to use those crud operators in the conductor | 15:25 |
alaski | edleafe: WIP https://review.openstack.org/#/c/258628/ | 15:25 |
*** rajalakshmig has joined #openstack-nova | 15:25 | |
edleafe | so for new instances, the reqspec is persisted? The problem is with the existing instances? | 15:25 |
bauzas | well, C and R of course | 15:25 |
bauzas | for the moment, it isn't | 15:26 |
*** yarkot has joined #openstack-nova | 15:26 | |
bauzas | but it will be | 15:26 |
edleafe | alaski: thanks for the link | 15:26 |
*** ccarmack has quit IRC | 15:26 | |
bauzas | cells v2 shares the same concern | 15:26 |
*** yamahata has quit IRC | 15:26 | |
*** mtanino has joined #openstack-nova | 15:26 | |
bauzas | because we need to somehow persist the request information between the cells | 15:26 |
*** ccarmack has joined #openstack-nova | 15:26 | |
*** yamahata has joined #openstack-nova | 15:26 | |
bauzas | how we use that persisted information is where we diverge w/ cells v2 | 15:26 |
doffm | alaski: Are we persisting the request information so that if a cell cannot handle the request it can be rescehduled to another? | 15:27 |
bauzas | edleafe: but back to your original question, I know that carl_baldwin was keen to discuss on that for neutron | 15:28 |
edleafe | bauzas: ok, that's good info. So if an external project wants to call a migration, they can pick a host, and (eventually) the scheduler will be called to veto it if it isn't valid per the original reqspec? | 15:28 |
*** yarkot has quit IRC | 15:28 | |
edleafe | but for now, they have no insight into the scheduler? | 15:28 |
bauzas | edleafe: I wasn't there at the last scheduler meeting because $holidays but I know that's something he wanted to discuss during the midcycle and johnthetubaguy told about a backlog spec | 15:29 |
* carl_baldwin reading back... | 15:29 | |
alaski | doffm: not exactly. cells doesn't specifically need the request spec persisted, what it needs is enough of the instance info persisted to be able to hydrate an instance object before data has been written to the instance table | 15:29 |
bauzas | edleafe: that workflow hasn't yet been agreed | 15:29 |
bauzas | neither discssed :) | 15:29 |
edleafe | bauzas: ack | 15:29 |
*** breitz has quit IRC | 15:30 | |
alaski | doffm: because the instance won't be written to an instance table until it has been scheduled to a cell. but the API needs to return a response before that, so we shove the data into the request spec | 15:30 |
*** breitz has joined #openstack-nova | 15:30 | |
bauzas | and that's why we have a purgatory cell :p | 15:30 |
alaski | doffm: and a list/show will hydrate an instance from the req spec, and another object/table called buildrequest | 15:30 |
mriedem | dansmith: in https://review.openstack.org/#/c/247853/18/nova/virt/xenapi/vmops.py - i didn't think x in obj worked if obj didn't have the dict compat mixin | 15:30 |
dansmith | mriedem: we moved it in nova to our base, right? | 15:31 |
mriedem | dik | 15:32 |
mriedem | *idk | 15:32 |
mriedem | looking | 15:32 |
dansmith | I thought we did to bridge the gap to o.vo having it | 15:32 |
dansmith | it's not in there anymore, | 15:32 |
dansmith | and your __contains__ merged to o.vo 3 months ago, | 15:32 |
dansmith | so I think we have it now | 15:33 |
mriedem | nope https://github.com/openstack/nova/blob/master/nova/objects/base.py#L61 | 15:33 |
*** lucas-hungry is now known as lucasagomes | 15:33 | |
dansmith | right ^ | 15:33 |
dansmith | mriedem: https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/base.py#L316 | 15:34 |
*** whenry has quit IRC | 15:34 | |
dansmith | mriedem: and in 0.13: https://github.com/openstack/oslo.versionedobjects/blob/0.13.0/oslo_versionedobjects/base.py | 15:34 |
dansmith | which is in our requirements | 15:34 |
doffm | alaski: Got it. Thanks. | 15:34 |
mriedem | dansmith: ahhhh | 15:34 |
* mriedem thanks past version of myself for doing that | 15:34 | |
doffm | Using the request spec like a 'pre instance' table. | 15:35 |
*** dougs has quit IRC | 15:35 | |
*** cfriesen__ has quit IRC | 15:35 | |
alaski | doffm: exactly | 15:35 |
*** annegentle has joined #openstack-nova | 15:36 | |
ndipanov | dansmith, is there an easy way to access service version data | 15:38 |
ndipanov | dansmith, it's a single db call right? | 15:39 |
mriedem | dansmith: ok, +2 now | 15:39 |
dansmith | ndipanov: it is.. I have a method for getting the computes | 15:39 |
ndipanov | but I need others | 15:39 |
dansmith | I know, turns out I'm smarter than I remember: | 15:39 |
dansmith | ndipanov: https://github.com/openstack/nova/blob/master/nova/objects/service.py#L283 | 15:40 |
*** Sukhdev has joined #openstack-nova | 15:40 | |
*** dougs has joined #openstack-nova | 15:40 | |
*** tonytan4ever has joined #openstack-nova | 15:40 | |
ndipanov | dansmith, nice - how do I know which version of the object this relates to? | 15:40 |
mriedem | https://github.com/openstack/nova/blob/master/nova/objects/service.py#L87 | 15:43 |
*** _diana_1 has joined #openstack-nova | 15:43 | |
dansmith | it's not the object version, but that global counter | 15:43 |
dansmith | so you can just define the version in the history and then check to see that we're >= that | 15:43 |
ndipanov | dansmith, but that means nothing to me when deciding whether to migrate data or not | 15:43 |
ndipanov | dansmith, ok that makes sens | 15:43 |
ndipanov | e | 15:43 |
ndipanov | I think... | 15:44 |
dansmith | you can either add the pci device version to the things that are considered, or not | 15:45 |
dansmith | even just adding another entry to that list with the same compute rpc version will give you the sentinel you need | 15:45 |
dansmith | i.e.: | 15:45 |
dansmith | # Version 5: Online migration of parent_addr in PciDevice | 15:46 |
dansmith | {'compute_rpc': $same_as_before} | 15:46 |
openstackgerrit | Andrew Laski proposed openstack/nova: Clean up filter_properties/availability_zone in boot https://review.openstack.org/254434 | 15:46 |
dansmith | then only do your migration if the minimum service version for conductor and api is >= 5 | 15:47 |
ndipanov | dansmith, got it | 15:47 |
wznoinsk | claudiub, are you around? | 15:47 |
claudiub | wznoinsk: hello | 15:48 |
dansmith | ndipanov: and I already have a patch up that claims version 4, so hopefully that can get in the queue soon so we're not competing for it | 15:48 |
*** zenoway has joined #openstack-nova | 15:48 | |
*** _diana_1 has quit IRC | 15:48 | |
*** devananda has quit IRC | 15:48 | |
*** mragupat has joined #openstack-nova | 15:48 | |
dansmith | i.e. this one: https://review.openstack.org/#/c/251588/19 | 15:48 |
wznoinsk | claudiub, wanted to check what's the story with instance live resize, I'm trying to get the gerrit code working at the moment to do some libvirt virt driver tests but I wanted to check is it under RFE or N? | 15:49 |
*** klkumar has quit IRC | 15:50 | |
openstackgerrit | dane-fichter proposed openstack/nova: Add image signature verification https://review.openstack.org/189843 | 15:51 |
* wznoinsk hopes it's still on the table | 15:51 | |
*** klkumar has joined #openstack-nova | 15:51 | |
mriedem | vponomaryov: i don't see anything in the cinder change history around 1/3 that looks related to the volume attach/detach issues you're seeing | 15:52 |
claudiub | wznoinsk: eehh... it is unfortunate, even if it has 2x +2s and 11x +1, it wasn't approved. :( johnthetubaguy said that it's going to get approved for N though. | 15:52 |
vponomaryov | mriedem: right, code and packages are the same | 15:52 |
vponomaryov | mriedem: python and system packages are completely the same | 15:52 |
*** zenoway has quit IRC | 15:53 | |
claudiub | wznoinsk: as for the code, pretty much all of it should be there. ofc, it needs rebase, as the microversion version has changed since then, but last time I did nova live-resize <server> <new-flavor-id>, it was working, up to the driver's live_resize call. | 15:54 |
claudiub | wznoinsk: do you have any issues? | 15:54 |
vponomaryov | mriedem: I think would be useful to send commit with creation of only one Vm and attach of one volume and see its timings | 15:55 |
wznoinsk | claudiub, I didn't see libvirt implementation, checking again | 15:55 |
claudiub | wznoinsk: libvirt implementation wasn't added, afaik. | 15:55 |
vponomaryov | mriedem: to get to know weak point | 15:55 |
wznoinsk | claudiub, getting endpoint not supporthing the method but it maybe my merging of the code | 15:55 |
wznoinsk | claudiub, ok so for libvirt there's still some work I suppose | 15:55 |
mriedem | vponomaryov: there are several tempest tests that attach a single volume to an instance in the normal tempest full job, we should be able to figure that out | 15:55 |
*** cdlhh has quit IRC | 15:56 | |
mriedem | vponomaryov: like http://logs.openstack.org/85/263385/1/check/gate-tempest-dsvm-full/e2f519a/console.html#_2016-01-04_22_08_21_522 here | 15:56 |
claudiub | wznoinsk: makes sense, IMO, the python-novaclient microversion version is behind. it is still 2.13, it should be 2.15 | 15:57 |
vponomaryov | mriedem: 1.5 sec? )) | 15:57 |
claudiub | wznoinsk: in this patch: https://review.openstack.org/#/c/248611/4 | 15:57 |
*** shyama has joined #openstack-nova | 15:57 | |
vponomaryov | mriedem: looks like main time were spent in "setup_resource" method | 15:57 |
*** mragupat has quit IRC | 15:58 | |
*** krtaylor has quit IRC | 15:58 | |
*** mragupat has joined #openstack-nova | 15:58 | |
wznoinsk | claudiub, I went away with curl POST and it was fine: curl -i -X POST -H "Content-Type: application/json" -H "X-OpenStack-Nova-API-Version: 2.15" -H "X-Auth-Token:fc323d6cb2d040fbb845273d9a5ff3b1" -d '{"live-resize": {"flavorRef": "3"}}' http://127.0.0.1:8774/v2.1/0bf245bfd4da437d92ba2b58fb9a7d8f/servers/9093e818-5a41-4e1a-a9d6-71e883990bb2/action | 15:59 |
wznoinsk | needed to incloude the header for the proper api version | 15:59 |
vponomaryov | mriedem: and big difference, this test uses explciit device path | 15:59 |
*** _diana_1 has joined #openstack-nova | 15:59 | |
vponomaryov | mriedem: that could be the reason | 15:59 |
wznoinsk | claudiub, I'm not sure whether not using the pythonclient is maybe the cause of my further problems (but unlikely) | 15:59 |
mriedem | vponomaryov: yeah most of the tempest tests pass in a device name on attach/boot from volume | 16:00 |
cdent | rlrossit: thanks for the review, I've responded, hoping you've got an example for the right way to do versioned objects backed by the api database, I was just following what I could find | 16:00 |
mriedem | vponomaryov: however, with the libvirt driver in nova, the device name passed in is ignored anyway | 16:00 |
claudiub | wznoinsk: yeah. you need the python-novaclient patch, which itself is out of date, and it sends 2.13 as the nova-api version. :) | 16:00 |
vponomaryov | mriedem: then, problem is in other things, not "attach/detach" | 16:00 |
*** 77CAAHDEK has quit IRC | 16:01 | |
*** devananda has joined #openstack-nova | 16:01 | |
*** haomaiwang has joined #openstack-nova | 16:01 | |
*** rushiagr is now known as rushiagr_away | 16:01 | |
claudiub | wznoinsk: anyways, I'll have to rebase everything once the spec is approved. Until then, the microversion version will keep changing, so rebasing over and over again is pointless. :) | 16:01 |
*** bklei has quit IRC | 16:03 | |
wznoinsk | claudiub, sure, understood | 16:03 |
*** artom has joined #openstack-nova | 16:03 | |
*** unicell has joined #openstack-nova | 16:06 | |
*** haomaiwang has quit IRC | 16:07 | |
*** unicell1 has quit IRC | 16:08 | |
*** zenoway has joined #openstack-nova | 16:08 | |
openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/nova: Keep py3.X compatibility for urllib https://review.openstack.org/261034 | 16:08 |
*** sahid has quit IRC | 16:09 | |
*** krtaylor has joined #openstack-nova | 16:10 | |
*** salv-orlando has joined #openstack-nova | 16:12 | |
*** davideagnello has joined #openstack-nova | 16:12 | |
*** otter768 has joined #openstack-nova | 16:12 | |
openstackgerrit | Stephen Finucane proposed openstack/nova: test_fields: Remove all 'Enum' subclass tests https://review.openstack.org/244372 | 16:13 |
*** zenoway has quit IRC | 16:13 | |
*** zenoway has joined #openstack-nova | 16:14 | |
*** davideagnello has quit IRC | 16:14 | |
*** davideagnello has joined #openstack-nova | 16:14 | |
openstackgerrit | Stephen Finucane proposed openstack/nova: tests_fields: Remove most enum-type field tests https://review.openstack.org/244370 | 16:15 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: Move config options from nova/api directory https://review.openstack.org/258180 | 16:15 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: Better document the osapi_max_limit config option https://review.openstack.org/233212 | 16:15 |
artom | Ah crap, that didn't do what I wanted :( | 16:15 |
*** lyanchih has quit IRC | 16:16 | |
artom | I mean, sort of, but I just barged into edleafe's work - while all I wanted to do was depend on it. | 16:17 |
*** otter768 has quit IRC | 16:17 | |
*** belmoreira has quit IRC | 16:18 | |
artom | edleafe, I accidentally uploaded a patch to your move config options review here: https://review.openstack.org/#/c/258180/ | 16:19 |
edleafe | artom: no worries - I've done that myself | 16:20 |
artom | edleafe, so how do we fix this? You just re-submit a new (unchanged) patchset yourself? | 16:20 |
edleafe | artom: let me look over the changes. | 16:21 |
edleafe | artom: did you modify https://review.openstack.org/#/c/258180 at all? | 16:21 |
artom | edleafe, I think I did - but I had no intention to | 16:21 |
artom | My purpose was to rebase by patch on top of yours and submit. | 16:21 |
openstackgerrit | Stephen Finucane proposed openstack/nova: config options: Make xvp proxy into vnc group https://review.openstack.org/235398 | 16:22 |
artom | I think I failed the rebase and introduced changes to your patch. | 16:22 |
edleafe | artom: ah, you probably rebased and then modified the branch | 16:22 |
edleafe | artom: yep | 16:22 |
edleafe | let me push a refresh | 16:22 |
artom | So we should pretend patchset 6 never existed ;) | 16:22 |
*** ctrath has quit IRC | 16:23 | |
edleafe | artom: sweep it under the rug, so to speak :) | 16:24 |
artom | edleafe, you missed the chance to make a 'turn a new leaf' pun :) | 16:24 |
*** zenoway has quit IRC | 16:25 | |
*** zenoway has joined #openstack-nova | 16:25 | |
edleafe | artom: I've made a ton of those in my life. Now leaf me alone. | 16:25 |
openstackgerrit | Ed Leafe proposed openstack/nova: Move config options from nova/api directory https://review.openstack.org/258180 | 16:26 |
*** yarkot has joined #openstack-nova | 16:26 | |
edleafe | artom: ok, should be reverted now | 16:27 |
artom | edleafe, awesome. Sorry for the trouble. | 16:27 |
*** blaisebool has joined #openstack-nova | 16:28 | |
edleafe | artom: heh, it happens | 16:28 |
*** zenoway has quit IRC | 16:30 | |
*** wznoinsk has quit IRC | 16:31 | |
*** wznoinsk has joined #openstack-nova | 16:32 | |
*** mdrabe has quit IRC | 16:32 | |
*** matrohon has quit IRC | 16:33 | |
*** cgascoig has quit IRC | 16:33 | |
*** cgascoig_ has joined #openstack-nova | 16:34 | |
*** cgascoig_ is now known as cgascoig | 16:34 | |
*** dslevin_ has quit IRC | 16:34 | |
*** kylek3h_ has joined #openstack-nova | 16:37 | |
*** kylek3h has quit IRC | 16:37 | |
*** davideagnello has quit IRC | 16:39 | |
*** sacharya has joined #openstack-nova | 16:41 | |
*** kotra03 has quit IRC | 16:41 | |
*** vishwanathj has quit IRC | 16:42 | |
*** armax has joined #openstack-nova | 16:42 | |
*** dslevin_ has joined #openstack-nova | 16:44 | |
*** mgoddard_ has joined #openstack-nova | 16:45 | |
mriedem | vponomaryov: does it seem odd that the failure only happens on 2 manila changes in the check queue? | 16:45 |
*** rfolco has quit IRC | 16:45 | |
vponomaryov | mriedem: tempest broke us too | 16:45 |
vponomaryov | mriedem: )) | 16:45 |
vponomaryov | mriedem: https://bugs.launchpad.net/designate/+bug/1531049 | 16:46 |
openstack | Launchpad bug 1531049 in Manila "Devstack stop supporting Tempest deprecated cred options" [Critical,In progress] - Assigned to Valeriy Ponomaryov (vponomaryov) | 16:46 |
*** gjayavelu has joined #openstack-nova | 16:46 | |
vponomaryov | mriedem: so, those two have workaround | 16:46 |
mriedem | oh right https://review.openstack.org/#/c/263664/ | 16:46 |
mriedem | and those fail b/c of the timeout | 16:46 |
vponomaryov | mriedem: right | 16:47 |
*** gjayavel_ has joined #openstack-nova | 16:47 | |
*** chhavi has quit IRC | 16:47 | |
*** zakora has quit IRC | 16:48 | |
*** mgoddard has quit IRC | 16:48 | |
*** jwcroppe has quit IRC | 16:48 | |
mriedem | vponomaryov: so, i'm wondering if the tempest creds thing is the cause - are the tests all running under the same tenant? | 16:48 |
*** dslevin_ has quit IRC | 16:49 | |
*** rfolco has joined #openstack-nova | 16:49 | |
vponomaryov | mriedem: manila requires preconfigured tenant/user | 16:49 |
vponomaryov | mriedem: because creates VM in Nova for shares | 16:49 |
vponomaryov | mriedem: and if start use new creds | 16:50 |
vponomaryov | mriedem: we will create huge amount of redundant resources | 16:50 |
*** mdrabe has joined #openstack-nova | 16:50 | |
vponomaryov | mriedem: we without it excceed job timeouts on OVH nodes | 16:50 |
*** ctrath has joined #openstack-nova | 16:50 | |
*** _diana_1 has quit IRC | 16:50 | |
vponomaryov | mriedem: in 50 % cases | 16:50 |
openstackgerrit | Pawel Koniszewski proposed openstack/nova: Add way to pause VM during live migration https://review.openstack.org/245921 | 16:50 |
openstackgerrit | Pawel Koniszewski proposed openstack/nova: Add new DB API method to retrieve migration for instance https://review.openstack.org/257270 | 16:50 |
mriedem | w/o preconfigured credentials you timeout on the ovh nodes? | 16:51 |
*** gjayavelu has quit IRC | 16:51 | |
*** _diana_1 has joined #openstack-nova | 16:51 | |
vponomaryov | mriedem: I suspect, we will on all | 16:51 |
vponomaryov | mriedem: with preconfigured only on OVH | 16:51 |
vponomaryov | mriedem: sometimes on HP nodes | 16:51 |
*** erlon has joined #openstack-nova | 16:51 | |
*** whenry has joined #openstack-nova | 16:51 | |
vponomaryov | mriedem: on those nodes, where devstack installation takes more than 40 minutes | 16:52 |
vponomaryov | mriedem: RAX nodes take about 25 | 16:52 |
*** ijw_ has quit IRC | 16:52 | |
bauzas | should we call pilgrimstack ? | 16:54 |
mriedem | mtreinish: so if https://bugs.launchpad.net/bugs/1524717 is a critical bug that's not fixed and is breaking a bunch of gate jobs per https://bugs.launchpad.net/designate/+bug/1531049 shouldn't we temporarily revert https://review.openstack.org/#/c/256164/ ? | 16:54 |
openstack | Launchpad bug 1524717 in tempest "ObjectACL tempest tests are flaky with preprov creds" [Critical,Confirmed] - Assigned to Andrea Frittoli (andrea-frittoli) | 16:54 |
openstack | Launchpad bug 1531049 in Manila "Devstack stop supporting Tempest deprecated cred options" [Critical,In progress] - Assigned to Valeriy Ponomaryov (vponomaryov) | 16:54 |
vponomaryov | mriedem: according to bug https://bugs.launchpad.net/bugs/1524717 all devstack plugin for manila backends are affected too - all third-party CIs | 16:55 |
vponomaryov | mriedem: in summary, more than 10 repos require urgent fix | 16:55 |
mtreinish | mriedem: I don't think it's critical, and it shouldn't block anyone for that second bug | 16:56 |
mtreinish | it's only a scheduling issue when running in parallel | 16:56 |
mtreinish | if projects were only working using hard coded config credentials then they should only be running serially anyway | 16:56 |
openstackgerrit | Chuck Carmack proposed openstack/nova: config options: centralize 'spice' options https://review.openstack.org/264271 | 16:56 |
mtreinish | and they couldn't have specified specific roles which is where that's coming up | 16:57 |
*** garyk has quit IRC | 16:57 | |
mriedem | vponomaryov: are the manila tests run in serial or parallel? | 16:57 |
vponomaryov | mtreinish: now it is impossible to use preconfigured creds in latest tempest and devstack | 16:57 |
vponomaryov | mriedem: parallel | 16:57 |
*** salv-orlando has quit IRC | 16:57 | |
mriedem | vponomaryov: that's dangerous | 16:58 |
mriedem | that's why we have tenant isolation in the gate | 16:58 |
mriedem | for running in parallel | 16:58 |
mtreinish | vponomaryov: umm, it always was. Using the legacy credential options from the config file precludes running in parallel | 16:58 |
mriedem | otherwise you can hit over quota type issues | 16:58 |
mtreinish | like mriedem said | 16:58 |
*** salv-orlando has joined #openstack-nova | 16:58 | |
mtreinish | the test accounts file enables running in parallel with multiple prevprovisioned accounts | 16:58 |
mtreinish | but you don't have to run in parallel with it | 16:59 |
vponomaryov | mriedem, mtreinish: we handle quotas not excceed them | 16:59 |
vponomaryov | s/excceed/exceeding/ | 16:59 |
vponomaryov | mriedem, mtreinish: so, quotas is out of issues | 16:59 |
mriedem | well quotas is only one thing | 17:00 |
mriedem | there are also physical resources on the test node | 17:00 |
mriedem | i'm not sure how you're handling quotas | 17:00 |
vponomaryov | thin provisioned LVM volumes | 17:00 |
vponomaryov | 4mb each | 17:00 |
mtreinish | mriedem: the issue here is there were a number of projects that didn't understand how the plugins where using credential information and were relying on something they shouldnt have been | 17:00 |
*** chhavi has joined #openstack-nova | 17:00 | |
*** dslevin_ has joined #openstack-nova | 17:00 | |
mtreinish | they either need to migrate to the abstract credential provider that tempest is using | 17:00 |
*** whenry has quit IRC | 17:00 | |
mtreinish | or in the case of projects that have weird requirements (which don't make sense to me) they should assert control over their own configuration | 17:01 |
*** zenoway has joined #openstack-nova | 17:01 | |
alaski | nova cells meeting in #openstack-meeting-3 | 17:01 |
*** moshele has quit IRC | 17:01 | |
*** Sukhdev has quit IRC | 17:01 | |
mtreinish | mriedem: what I'd really like to see is for someone to step up and help andreaf migrate the credential providers to tempest-lib which has been moving very slowly | 17:02 |
mriedem | ok, i'm in the cells meeting now, | 17:02 |
vponomaryov | mtreinish: we are ok with any approach for having "single" set of tenant/user | 17:02 |
mriedem | i guess i don't understand the manila requirements or the tempest config issues, | 17:02 |
mriedem | but i'm trying to figure out how much i need to investigate the actual nova failures that manila is hitting, or if their config is invalid somehow | 17:02 |
mriedem | but the only thing i can think that caused their failures since 1/3 is removing those deprecation options | 17:03 |
vponomaryov | mriedem: options removed 24 hours after it | 17:03 |
vponomaryov | mriedem:no relation | 17:03 |
vponomaryov | mriedem:100% for sure | 17:03 |
vponomaryov | mriedem: tempest and nova problems are not related | 17:04 |
mriedem | vponomaryov: there were no changes in nova on 1/3 that i can see that would be related to what you're seeing | 17:04 |
*** rajalakshmig has quit IRC | 17:04 | |
vponomaryov | mriedem: tempest/devstack change was merged on 1/4 4PM | 17:04 |
*** shyama has quit IRC | 17:04 | |
mtreinish | vponomaryov: for executing serially you should be fine with 1 set of credentials | 17:04 |
*** mgoddard_ has quit IRC | 17:05 | |
*** mgoddard has joined #openstack-nova | 17:05 | |
mtreinish | but running parallely with a single set of creds breaks underlying assumptions in the provider mechanisms | 17:05 |
vponomaryov | mtreinish: serially not suitable because of time reasons | 17:05 |
mriedem | yeah i guess historically when we ran tempest internally with a hard-coded set of tenant creds, we couldn't run concurrently | 17:05 |
*** zenoway has quit IRC | 17:05 | |
*** jlanoux has quit IRC | 17:05 | |
vponomaryov | mtreinish: if so, then I can say that creds provider is not flexible enough | 17:06 |
mtreinish | vponomaryov: well the internal assumption of tempest is one set of creds for each test class | 17:06 |
mtreinish | that's always been the case because tempest needs to provide isolation between each class | 17:06 |
mtreinish | it's not a matter of flexibility | 17:06 |
*** yassine has quit IRC | 17:06 | |
vponomaryov | mtreinish: we separate tests by classes not because of creds, but functionality | 17:07 |
vponomaryov | mtreinish: so, should be mechanism to use same creds for several classes then | 17:07 |
vponomaryov | mtreinish: and completely no problems about usage of same creds | 17:08 |
vponomaryov | mtreinish: in our case | 17:08 |
mtreinish | vponomaryov: there isn't and can't be for tempest proper, because it doesn't work in parallel | 17:08 |
mtreinish | vponomaryov: there's nothing stopping you from just owning your own configuration for whatever weird use case you think you need | 17:08 |
*** tongli has joined #openstack-nova | 17:09 | |
vponomaryov | mtreinish: it does not work in parallel because it is done so or fails? | 17:10 |
mtreinish | tempest will not work if you try to run the tests in parallel with the same credentials | 17:11 |
openstackgerrit | dane-fichter proposed openstack/nova: Add image signature verification https://review.openstack.org/189843 | 17:11 |
mtreinish | the tests would walk all over each other and things will fail randomly | 17:11 |
vponomaryov | mtreinish: manila runs tests in parallel for long time and tests designed to be aware about concurrent things | 17:12 |
vponomaryov | mtreinish: so, it is strange to me to hear that it does nor work | 17:12 |
vponomaryov | mtreinish: then, maybe, tests designed worse than it could be if that happens | 17:13 |
mtreinish | vponomaryov: that's not how tempest is designed. The assumption provided by tempest is each class will be provided an isolated set of credentials | 17:13 |
mtreinish | it's a design point going back a long time | 17:13 |
vponomaryov | mtreinish: and reason is in "potential interfere by tests"? | 17:14 |
*** ctrath1 has joined #openstack-nova | 17:14 | |
*** gjayavel_ has quit IRC | 17:15 | |
*** jgriffith is now known as jgriffith_away | 17:15 | |
vponomaryov | I mean "mutual interference" | 17:15 |
*** ctrath has quit IRC | 17:17 | |
openstackgerrit | Rushi Agrawal proposed openstack/nova: Remove mox and Stubs from tests/unit/pci/test_manager.py https://review.openstack.org/263362 | 17:18 |
rlrossit | cdent: you have uncovered a rabbit hole I never knew existed with this api database stuff... | 17:19 |
cdent | impressive isn't it? | 17:19 |
*** salv-orlando has quit IRC | 17:19 | |
*** rk4n has quit IRC | 17:20 | |
*** salv-orlando has joined #openstack-nova | 17:20 | |
rlrossit | cdent: so now I understand that you're doing things the same way as everyone else, but now I don't get why everyone else is doing it this way | 17:20 |
rlrossit | all of those sqlalchemy queries could very easily live in nova.db.sqlalchemy instead of nova.objects | 17:21 |
cdent | the stuff in nova/db/api is configured for only one database | 17:21 |
cdent | so I guess people are working around that | 17:21 |
cdent | I agree that it is a strange way to go | 17:21 |
* rlrossit hates databases, but is glad ORMs exist at least | 17:22 | |
artom | So... how *do* I submit a patch for review that depends on someone else's patch in review? | 17:22 |
* edleafe feels exactly the opposit of rlrossit | 17:22 | |
artom | Without submitting a new patchset in their review as wlel? | 17:22 |
*** ljxiash has quit IRC | 17:22 | |
rlrossit | edleafe: let's agree to disagree ;) | 17:22 |
*** ljxiash has joined #openstack-nova | 17:23 | |
*** renatoarmani has quit IRC | 17:23 | |
edleafe | rlrossit: too many years of being a DBA to ever like working with ORMs :) | 17:23 |
edleafe | artom: 1) grab the ps from gerrit with git review -d NNNNNN | 17:23 |
edleafe | 2) run 'git co -b my_dependent_branch' | 17:23 |
*** acabot has joined #openstack-nova | 17:24 | |
dansmith | rlrossit: we said that for the api database stuff, we were going to do the DB bits in the objects directly and avoid the multiple levels of indirection to a single implementation | 17:24 |
edleafe | 3) do all your changes in that branch | 17:24 |
ndipanov | artom, rebase on the patch you want | 17:24 |
*** nic has joined #openstack-nova | 17:24 | |
ndipanov | and then instead of doing git review | 17:24 |
ndipanov | do: git push gerrit HEAD:refs/for/master | 17:24 |
edleafe | 4) push changes with 'git review -y'. Unless you changed anything in the original patch, it should be good. | 17:24 |
*** sfinucan has quit IRC | 17:24 | |
edleafe | artom: or you could follow ndipanov's method | 17:25 |
dansmith | edleafe: it'll still try to rebase the whole set on master unless you use -R I think | 17:25 |
artom | edleafe, ndipanov's way makes more sense to my brain, sorry :) | 17:25 |
rlrossit | dansmith: totally devil's advocate but... isn't all of nova already doing that anyways with only having sqlalchemy? | 17:25 |
edleafe | dansmith: oops, forgot the rebase step | 17:26 |
ndipanov | edleafe, no I think it will still create a new PS in that casae... | 17:26 |
dansmith | ndipanov: it won't if it doesn't rebase | 17:26 |
edleafe | 2a) git rebase -i original_branch | 17:26 |
dansmith | rlrossit: yes, but we have two layers of indirection required for any new call for no real abstraction reason | 17:26 |
*** garyk has joined #openstack-nova | 17:27 | |
rlrossit | dansmith: wait maybe I was not saying things right then... I was thinking it live in nova.db.sqlalchemy.api (not nova.db.api) to at least put all of the queries in one place | 17:29 |
artom | ndipanov, wait, when you write refs/for/master...? | 17:30 |
dansmith | rlrossit: yeah, and we could (continue to) do that | 17:30 |
dansmith | rlrossit: it's just that when we set off down the api database path, we didn't see the reason for it, and decided that we'd start putting things directly in the object layer to avoid all the extra hoops | 17:30 |
*** jgriffith_away is now known as jgriffith | 17:30 | |
dansmith | rlrossit: I'm not sure that having all the queries in one place really helps anything, and definitely makes it harder to optimize some things | 17:30 |
*** e0ne has quit IRC | 17:32 | |
rlrossit | dansmith: well, at least I understand the reasoning for it now. Initially I was super confused to see this stuff living in the objects | 17:32 |
cdent | rlrossit, dansmith: I rather like having it in the object itself | 17:32 |
dansmith | I have to run to a meeting, back in a bit | 17:32 |
cdent | It makes the objects...uh... feel more real | 17:33 |
rlrossit | hey | 17:33 |
rlrossit | objects are real :( | 17:33 |
rlrossit | at least I think they are... | 17:33 |
cdent | I just mean in the sense of the the VersionedObjects managing their own persistence | 17:34 |
rlrossit | eh, I wonder if they should actually handle their persistence, or just let the DB handle those operations | 17:35 |
*** renatoarmani has joined #openstack-nova | 17:35 | |
rlrossit | either way, it's just a "where do I want the code to live" decision | 17:36 |
* cdent nods | 17:36 | |
*** unicell has quit IRC | 17:36 | |
cdent | I think if there wasn't already a precedent I'd put the query code elsewhere | 17:36 |
cdent | but it doesn't bother me too much | 17:36 |
rlrossit | and I suppose if everyone else has it live in the object, might as well leave it consistent with what's there | 17:36 |
rlrossit | if I want it to change, making you do it in your patch is the wrong place to make that request | 17:37 |
* cdent nods | 17:37 | |
cdent | I think it is definitely worth discussion. You going to the mid-cycle? | 17:37 |
rlrossit | cdent: no :(. mriedem and doffm already took our spots | 17:37 |
* cdent shakes his tiny fist | 17:38 | |
*** ctrath1 has quit IRC | 17:40 | |
vponomaryov | mtreinish, mriedem: "The assumption provided by tempest is each class will be provided an isolated set of credentials" - my answer is "The assumption provided by Manila is each class should not be dependent on creds at all.". About "serial/parallel" thing, even if we run tests serially, we can not use same creds for each class anymore. But exactly this is the USE CASE of Manila project. So, Manila is doomed to workaround tempest design, becau | 17:41 |
*** yamahata has quit IRC | 17:42 | |
rlrossit | dansmith: I'm curious what "makes it harder to optimize things" means though. From how I see things it's either the code is in nova.objects or the code is in nova.db.sqlalchemy.api. Unless there's some dynamic mumbo jumbo going on in the nova.db.sqlalchemy package | 17:42 |
*** blaisebool has quit IRC | 17:42 | |
rlrossit | (but I'm about to run off to lunch, so we probably can't continue this discussion until I'm back) | 17:43 |
*** rotbeard has quit IRC | 17:44 | |
*** jistr has quit IRC | 17:45 | |
dansmith | rlrossit: just things like we end up re-using some query because it's there when really we could have tweaked one thing about it if we had access to the sqlalchemy query object | 17:45 |
openstackgerrit | Harshada Mangesh Kakad proposed openstack/nova: Replace deprecated library function os.popen() with subprocess https://review.openstack.org/262570 | 17:45 |
dansmith | rlrossit: for extreme cases, we add a new flag parameter or something to control that behavior | 17:45 |
*** ctrath has joined #openstack-nova | 17:45 | |
*** chhavi has quit IRC | 17:46 | |
dansmith | rlrossit: but it just expands the number of things we have to do in order to effect that optimization so I expect it doesn't happen as often as possible | 17:46 |
*** jwcroppe has joined #openstack-nova | 17:48 | |
rlrossit | dansmith: oh so you were talking about optimizations within sqlalchemy? | 17:48 |
dansmith | rlrossit: not within, but with how we call it | 17:48 |
*** gjayavelu has joined #openstack-nova | 17:48 | |
*** Jeffrey4l has quit IRC | 17:49 | |
*** RichardRaseley has joined #openstack-nova | 17:50 | |
*** klkumar has quit IRC | 17:51 | |
*** renatoarmani has quit IRC | 17:51 | |
*** renatoarmani has joined #openstack-nova | 17:53 | |
*** BobBall is now known as BobBall_AWOL | 17:55 | |
*** nic1 has joined #openstack-nova | 17:55 | |
*** nic has quit IRC | 17:56 | |
*** cfriesen__ has joined #openstack-nova | 17:58 | |
*** derekh has quit IRC | 17:59 | |
*** suro-patz has joined #openstack-nova | 18:03 | |
*** hemna is now known as hemnafk | 18:04 | |
*** yamahata has joined #openstack-nova | 18:05 | |
*** moshele has joined #openstack-nova | 18:05 | |
openstackgerrit | Nikola Dipanov proposed openstack/nova: pci: do not filter out any SRIOV Physical Functions https://review.openstack.org/249016 | 18:07 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: pci: changing the claiming and allocation logic for PF/VF assignment https://review.openstack.org/249018 | 18:07 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: pci: adding support to specify a device_type in pci requests https://review.openstack.org/259254 | 18:07 |
openstackgerrit | Nikola Dipanov proposed openstack/nova: objects: adding a parent_addr field to the PciDevice object https://review.openstack.org/249015 | 18:07 |
*** xyang2 has quit IRC | 18:07 | |
*** lihg has joined #openstack-nova | 18:08 | |
*** salv-orlando has quit IRC | 18:08 | |
*** salv-orlando has joined #openstack-nova | 18:09 | |
openstackgerrit | Rushi Agrawal proposed openstack/nova: [WIP]Remove mox completely from nova/tests/unit/objects/test_instance.py https://review.openstack.org/264292 | 18:09 |
*** ctrath has quit IRC | 18:10 | |
*** unicell has joined #openstack-nova | 18:10 | |
ndipanov | dansmith, I updated that patch | 18:12 |
ndipanov | see if you like it | 18:12 |
*** gjayavelu has quit IRC | 18:13 | |
*** lihg_ has joined #openstack-nova | 18:13 | |
*** otter768 has joined #openstack-nova | 18:13 | |
*** dslevin_ has quit IRC | 18:14 | |
dansmith | ndipanov: yeah, looking | 18:17 |
*** otter768 has quit IRC | 18:18 | |
*** zenoway has joined #openstack-nova | 18:19 | |
*** ctrath has joined #openstack-nova | 18:20 | |
dansmith | ndipanov: I think you still have an issue | 18:20 |
dansmith | ndipanov: it bit me hard with the flavor stuff, probably not technically an issue for pci stuff because I think that addr probably can't change (or doesn't in practice) | 18:20 |
dansmith | but ignoring that case because we don't think it'll happen is likely to be a bad move I think | 18:21 |
*** abhishekk has joined #openstack-nova | 18:21 | |
ndipanov | dansmith, let me re-read that | 18:21 |
ndipanov | dansmith, it could change... | 18:22 |
abhishekk | johnthetubaguy: hi john, could you please reply your opinion about comment in review https://review.openstack.org/#/c/163798/? | 18:22 |
*** dslevin_ has joined #openstack-nova | 18:22 | |
ndipanov | from None to something | 18:22 |
dansmith | ah, then... we better handle it | 18:22 |
abhishekk | johnthetubaguy: thank you | 18:23 |
*** PsionTheory has joined #openstack-nova | 18:24 | |
*** nelsnelson has quit IRC | 18:24 | |
*** nelsnels_ has joined #openstack-nova | 18:24 | |
johnthetubaguy | abhishekk: looking | 18:25 |
abhishekk | johnthetubaguy: thank you | 18:26 |
*** rdo has quit IRC | 18:26 | |
openstackgerrit | Diana Clarke proposed openstack/nova: Replace stubs.Set with stub_out https://review.openstack.org/264297 | 18:26 |
*** pratikmallya has joined #openstack-nova | 18:26 | |
johnthetubaguy | abhishekk: sorry, I don't really understand your comment/question in the review | 18:27 |
*** rdo has joined #openstack-nova | 18:28 | |
*** renatoarmani has quit IRC | 18:28 | |
johnthetubaguy | abhishekk: which of my comments were you asking about? | 18:28 |
abhishekk | comment on ps 10 asked by rajesh | 18:29 |
abhishekk | allow force delete only when the task state is in "deleting" status? | 18:29 |
*** _diana_1 has quit IRC | 18:31 | |
johnthetubaguy | abhishekk: the code in there makes sure if a user calls delete multiple times, we quickly reject the request, so its not a Denial of service vector, my comments on that patchset were basically saying, the current force-delete idea seems to re-introduce that. We should probably only allow one force delete call, after the initial delete, maybe? although that seems a bit odd. | 18:31 |
*** _diana_1 has joined #openstack-nova | 18:32 | |
lihg | hello! Can anybody help me? I have a question about virtual box | 18:32 |
lihg | how to use virtual box as hypervisor in nova? | 18:32 |
abhishekk | johnthetubaguy: right | 18:32 |
*** zenoway has quit IRC | 18:33 | |
*** ctrath has quit IRC | 18:34 | |
abhishekk | but as per Oct 6 2:18 PM comment, one scenario is still not handled | 18:34 |
abhishekk | johnthetubaguy: "When user sends force-delete request for first time for instance deletion, its task_state gets changed to 'force-deleting'. Now since the instance task_state is already in 'force-deleting' state and if at that moment rabbitmq-server get crashed by some reasons, then the instance task_state remain 'force-deleting' forever and instance will not be deleted." | 18:35 |
*** _diana_1 has quit IRC | 18:35 | |
*** nelsnels_ has quit IRC | 18:35 | |
*** nelsnelson has joined #openstack-nova | 18:35 | |
*** READ10 has joined #openstack-nova | 18:35 | |
johnthetubaguy | abhishekk: its a trade off, yes | 18:35 |
abhishekk | to overcome this only way is allow force delete only when the task state is in "deleting" status | 18:36 |
abhishekk | what do you suggest | 18:36 |
*** _diana_1 has joined #openstack-nova | 18:39 | |
johnthetubaguy | abhishekk: You could maybe allow admins to do unlimited force deletes, but its not something you want want to give to all users. But I feel like its getting very complicated at that point. | 18:41 |
abhishekk | johnthetubaguy | 18:41 |
abhishekk | yes, you are right,should I abondened this patch as it is not satsisfying the need? | 18:42 |
*** e0ne has joined #openstack-nova | 18:44 | |
johnthetubaguy | abhishekk: so I think there is a need, but its very much an edge case. Reset-state for the admin can help in many of the real world cases. | 18:44 |
*** vilobhmm11 has joined #openstack-nova | 18:44 | |
*** READ10 has quit IRC | 18:45 | |
*** zenoway has joined #openstack-nova | 18:45 | |
openstackgerrit | Tracy Jones proposed openstack/nova: config options: centralize section "volume" https://review.openstack.org/253747 | 18:45 |
*** armax has quit IRC | 18:46 | |
*** jichen has joined #openstack-nova | 18:46 | |
abhishekk | johnthetubaguy: ok, you mean to say admin can reset the state in this case, right? | 18:47 |
johnthetubaguy | abhishekk: yes, I know that means the end user can't fix it themselves, but thats kinda on purpose to prevent denial of service attacks, if my memory is correct | 18:47 |
abhishekk | johnthetubaguy: ok, I will have internal discussion on your suggestion and will ping you back on this | 18:49 |
*** zenoway has quit IRC | 18:49 | |
*** paul-carlton2 has quit IRC | 18:50 | |
abhishekk | johnthetubaguy: thank you for your time | 18:50 |
*** fawadkhaliq has quit IRC | 18:50 | |
*** ctrath has joined #openstack-nova | 18:51 | |
*** hemnafk is now known as hemna | 18:51 | |
*** ijw has joined #openstack-nova | 18:51 | |
johnthetubaguy | abhishekk: happy to help | 18:51 |
*** krtaylor has quit IRC | 18:53 | |
*** gjayavelu has joined #openstack-nova | 18:54 | |
*** alejandrito has joined #openstack-nova | 18:56 | |
*** browne has joined #openstack-nova | 18:57 | |
openstackgerrit | LiuNanke proposed openstack/python-novaclient: Replace assertTrue(isinstance()) by optimal assert https://review.openstack.org/264308 | 18:58 |
*** abhishekk has quit IRC | 19:00 | |
*** rlrossit has quit IRC | 19:01 | |
*** andreykurilin__ has quit IRC | 19:02 | |
*** paul-carlton2 has joined #openstack-nova | 19:03 | |
*** tonytan4ever has quit IRC | 19:04 | |
*** fawadkhaliq has joined #openstack-nova | 19:06 | |
*** krtaylor has joined #openstack-nova | 19:07 | |
*** whenry has joined #openstack-nova | 19:07 | |
*** xyang2 has joined #openstack-nova | 19:08 | |
*** rlrossit has joined #openstack-nova | 19:08 | |
*** cfriesen__ has quit IRC | 19:12 | |
*** fawadkhaliq has quit IRC | 19:13 | |
*** edmondsw has quit IRC | 19:14 | |
*** tonytan4ever has joined #openstack-nova | 19:15 | |
ccarmack | alaski: I had a question related to https://review.openstack.org/#/c/225199/ ... in cells v1, are neutron provided security groups *not supported, same as n-net provided security groups? | 19:15 |
dansmith | ccarmack: I think so | 19:16 |
*** renatoarmani has joined #openstack-nova | 19:17 | |
dansmith | ccarmack: nova/compute/cells_api.py doesn't have a SecurityGroupAPI | 19:17 |
*** edmondsw has joined #openstack-nova | 19:17 | |
*** paul-carlton2 has quit IRC | 19:17 | |
dansmith | ccarmack: and if you used the one from api.py it wouldn't know how to contact the hosts, AFAIK | 19:17 |
*** harlowja_ has quit IRC | 19:18 | |
*** cfriesen__ has joined #openstack-nova | 19:18 | |
*** harlowja has joined #openstack-nova | 19:19 | |
ccarmack | dansmith: ok, thanks, I'll answer the comments that way | 19:19 |
*** zenoway has joined #openstack-nova | 19:21 | |
*** lpetrut has quit IRC | 19:25 | |
*** zenoway has quit IRC | 19:25 | |
alaski | ccarmack: dansmith that's correct. the issue with cells and security groups is the security group api only operates at the parent cell level | 19:26 |
*** pece has joined #openstack-nova | 19:26 | |
mriedem | vponomaryov: in this log http://logs.openstack.org/64/263664/5/check/gate-manila-tempest-dsvm-neutron/835bf51/logs/screen-n-cpu.txt.gz i was tracing the volume attach and detach for instance 17d72364-a54e-4ea7-9ff2-4ff3821ab022 - the attach takes ~15 seconds and the detach takes ~3 seconds | 19:26 |
dansmith | yeah, that's what I figured | 19:26 |
dansmith | ccarmack: let it be known that I was right once | 19:26 |
ccarmack | dansmith: duly noted | 19:27 |
alaski | heh. now that you've answered something about cells correctly you are close to as knowledgeable as the rest of us on it | 19:27 |
*** dedery_ has joined #openstack-nova | 19:28 | |
mriedem | vponomaryov: looks like iscsiadm sesssions are piling up http://logs.openstack.org/64/263664/5/check/gate-manila-tempest-dsvm-neutron/835bf51/logs/screen-n-cpu.txt.gz#_2016-01-06_14_05_11_976 | 19:28 |
*** moshele has quit IRC | 19:29 | |
*** alraddarla has joined #openstack-nova | 19:29 | |
*** jidar has joined #openstack-nova | 19:29 | |
*** moshele has joined #openstack-nova | 19:30 | |
*** gjayavelu has quit IRC | 19:30 | |
jidar | anybody know why my nova hosts would flap state=up -> state=down over and over again? | 19:30 |
*** dedery has quit IRC | 19:31 | |
*** dedery_ has quit IRC | 19:32 | |
*** eil397 has joined #openstack-nova | 19:33 | |
*** alraddarla has quit IRC | 19:34 | |
*** changbl has joined #openstack-nova | 19:35 | |
*** ociuhandu has quit IRC | 19:35 | |
*** READ10 has joined #openstack-nova | 19:35 | |
*** rlrossit has left #openstack-nova | 19:35 | |
*** rlrossit has joined #openstack-nova | 19:36 | |
*** smatzek has quit IRC | 19:38 | |
mriedem | vponomaryov: i thought you said the manila job used thinly provisioned targets? http://logs.openstack.org/64/263664/5/check/gate-manila-tempest-dsvm-neutron/835bf51/logs/screen-c-vol.txt.gz#_2016-01-06_14_05_01_723 | 19:39 |
mriedem | those all say they aren't thin provisioned | 19:39 |
*** READ10 has quit IRC | 19:40 | |
*** moshele has quit IRC | 19:42 | |
rlrossit | dansmith: is there any plan to axe NovaObjectDictCompat? | 19:43 |
*** smatzek has joined #openstack-nova | 19:43 | |
dansmith | rlrossit: you fix all the places we still use it and we can axe it like a mofo | 19:43 |
rlrossit | dansmith: do we want to classify it under a bp and do it in N? Or should I just go crazy? | 19:44 |
*** gjayavelu has joined #openstack-nova | 19:45 | |
*** gjayavelu has quit IRC | 19:45 | |
dansmith | I think it's legit cleanup suitable for doing until the last phase | 19:45 |
rlrossit | dansmith: totally unlinked to anything? it scares me because it's a py3-like change but won't have any sort of tracking | 19:46 |
*** ljxiash has quit IRC | 19:46 | |
dansmith | rlrossit: no, you can create a specless bp for it if you want, or slam it under the general objects one we have | 19:46 |
*** Sukhdev has joined #openstack-nova | 19:46 | |
dansmith | mriedem: agree? | 19:46 |
*** dslevin_ has quit IRC | 19:47 | |
* mriedem is deep in the bowels of libvirtd logs | 19:47 | |
*** e0ne has quit IRC | 19:47 | |
*** andreykurilin__ has joined #openstack-nova | 19:47 | |
mriedem | i'd be fine with the existing mitaka-objects bp or a new cleanup on | 19:47 |
mriedem | dansmith: what was that other person using for cleaning up the fake uuid stuff? | 19:47 |
rlrossit | I'll get a bp started. I don't want to put it in the general one because it'll spam the heck out of it | 19:47 |
dansmith | mriedem: I don't think he was using one | 19:48 |
dansmith | mriedem: and that person would be migi | 19:48 |
* rlrossit feels like he should be given the title nova-janitor | 19:48 | |
mriedem | rm-obj-dict-compat | 19:48 |
mriedem | sure | 19:48 |
mriedem | no, make the title make sense | 19:49 |
mriedem | unlike calling a new project 'jacket' | 19:49 |
mriedem | or 'nova' | 19:49 |
* mriedem gets coffee | 19:49 | |
rlrossit | mriedem: no, *I* want that title | 19:49 |
mriedem | oh in that case, | 19:49 |
mriedem | i have some vomit and urine in my office that needs cleaning | 19:49 |
mriedem | vponomaryov: i think i'm closer to your issue | 19:49 |
mriedem | http://logs.openstack.org/64/263664/5/check/gate-manila-tempest-dsvm-neutron/835bf51/logs/libvirt/libvirtd.txt.gz | 19:50 |
mriedem | vponomaryov: there is a call to virDomainAttachDeviceFlags for each attach | 19:50 |
mriedem | but there are only 2 calls to virDomainDetachDeviceFlags | 19:50 |
mriedem | for vdb | 19:50 |
mriedem | so it looks like libvirt is not detaching the devices | 19:50 |
*** pece has quit IRC | 19:50 | |
*** dslevin_ has joined #openstack-nova | 19:52 | |
*** penick has joined #openstack-nova | 19:52 | |
*** tjones has joined #openstack-nova | 19:54 | |
openstackgerrit | Tracy Jones proposed openstack/nova: config options: centralize section "service" https://review.openstack.org/256569 | 19:55 |
*** baoli_ has quit IRC | 19:57 | |
*** smatzek_ has joined #openstack-nova | 19:59 | |
mriedem | oh and another difference in the manila jobs is they don't specify a device name on attach to the compute api | 20:00 |
mriedem | the integrated gate jobs with tempest always specify a device name | 20:01 |
mriedem | i'm not entirely sure why that would make a difference though b/c the libvirt driver ignores the device name on the request | 20:01 |
*** lpetrut has joined #openstack-nova | 20:02 | |
*** smatzek has quit IRC | 20:02 | |
*** lucasagomes is now known as lucas-dinner | 20:03 | |
suro-patz | Request for review - https://review.openstack.org/#/c/257596/ https://review.openstack.org/#/c/254423/ melwitt jaypipes jwcroppe | 20:05 |
rlrossit | dansmith: https://blueprints.launchpad.net/nova/+spec/rm-object-dict-compat if you want a look | 20:07 |
*** ndipanov has quit IRC | 20:09 | |
*** zenoway has joined #openstack-nova | 20:09 | |
mriedem | vponomaryov: there are only 2 calls to detach in the n-cpu logs | 20:10 |
mriedem | so manila keeps attaching volumes but doesn't detach them | 20:10 |
mriedem | and that's why we run out of devices | 20:10 |
*** baoli has joined #openstack-nova | 20:10 | |
*** ociuhandu has joined #openstack-nova | 20:11 | |
dansmith | rlrossit: we normally approve those during the meetings, but I'd just start cranking away if I were you and we'll hit it tomorrow | 20:11 |
rlrossit | +1 sounds good | 20:12 |
rlrossit | dansmith: should I add it to the agenda? | 20:12 |
dansmith | rlrossit: yeah | 20:12 |
*** unicell has quit IRC | 20:12 | |
*** unicell1 has joined #openstack-nova | 20:12 | |
*** krtaylor has quit IRC | 20:13 | |
*** artom has quit IRC | 20:13 | |
rlrossit | dansmith: really stupid question... I just edit the nova meetings wiki page right? | 20:14 |
dansmith | yeah | 20:14 |
*** otter768 has joined #openstack-nova | 20:14 | |
*** dedery has joined #openstack-nova | 20:17 | |
*** gtt116_ has quit IRC | 20:17 | |
*** gtt116_ has joined #openstack-nova | 20:18 | |
*** otter768 has quit IRC | 20:19 | |
*** penick_ has joined #openstack-nova | 20:20 | |
*** ctrath has quit IRC | 20:21 | |
*** pkarikh has quit IRC | 20:21 | |
*** belmoreira has joined #openstack-nova | 20:22 | |
*** hemna is now known as hemnafk | 20:22 | |
*** penick has quit IRC | 20:23 | |
*** penick_ is now known as penick | 20:23 | |
*** pkarikh has joined #openstack-nova | 20:25 | |
*** zenoway has quit IRC | 20:25 | |
cdent | rlrossit: can you point me at a good example of a nondictcompat that is iterating its fields? | 20:27 |
*** ctrath has joined #openstack-nova | 20:28 | |
*** e0ne has joined #openstack-nova | 20:28 | |
rlrossit | cdent: I'm not sure if any classes have been changed to not use NovaObjectDictCompat yet... I'll take a look, and at worst I'll send you the first one that I'm getting ready for the bp I made for cleaning it up | 20:29 |
*** rfolco has quit IRC | 20:29 | |
*** e0ne has quit IRC | 20:29 | |
cdent | cool thanks | 20:29 |
rlrossit | cdent: actually I think this does it the desirable way: https://github.com/openstack/nova/blob/master/nova/objects/agent.py#L37-L42 | 20:30 |
*** e0ne has joined #openstack-nova | 20:30 | |
cdent | ah, okay, I was hoping for something with *attr() ... | 20:31 |
smatzek_ | sdague: Thanks for the review on "glance: secure https when getting client version". This method will be called at https://github.com/openstack/nova/blob/master/nova/image/glance.py#L142 when we start making actions use Glance V2. The parameters that need to be passed downstream are already in params which is why I'm doing what I'm doing. Would you prefer me to move the code that builds the params dict to a separate method and | 20:31 |
smatzek_ | call it inside _determine_curr_major_version? | 20:31 |
cdent | rlrossit: whoops, s/with/without/ | 20:32 |
openstackgerrit | Andrew Laski proposed openstack/nova: Clean up filter_properties/availability_zone in boot https://review.openstack.org/254434 | 20:32 |
openstackgerrit | Andrew Laski proposed openstack/nova: Persist the request spec during an instance boot https://review.openstack.org/258628 | 20:32 |
rlrossit | cdent: was there something else were you thinking to use? | 20:33 |
rlrossit | like dot syntax? | 20:33 |
cdent | SpecialMagic! | 20:33 |
cdent | ;) | 20:33 |
*** e0ne has quit IRC | 20:33 | |
sdague | smatzek_: I would rather all the params be explicit in the function call | 20:33 |
rlrossit | SpecialMagic is reserved for dansmith's use only | 20:34 |
cdent | :) | 20:34 |
cdent | I'll fix it either tonight or tomorrow, thanks for the review. | 20:35 |
smatzek_ | sdague: Since the method will simply pass them down to http.HttpClient, the defaults of the parameters on _determine_cur_major_version would need to match the defaults in http.HttpClient. That would be a management issue downstream if/when they change on http.HttpClient. | 20:36 |
sdague | smatzek_: that's fine, we want to know that, if it's just **foo then it's all magic random and we have no idea what's going on. | 20:37 |
smatzek_ | sdague: OK, I was just trying to avoid tying the method's parameters and defaults to http.HttpClient and getting bit in the future when a mismatch comes along. I can do it that way. | 20:38 |
*** ctrath has quit IRC | 20:39 | |
*** armax has joined #openstack-nova | 20:40 | |
*** zhenq has joined #openstack-nova | 20:41 | |
*** krtaylor has joined #openstack-nova | 20:43 | |
*** rk4n has joined #openstack-nova | 20:44 | |
*** rfolco has joined #openstack-nova | 20:45 | |
*** ctrath has joined #openstack-nova | 20:47 | |
*** matrohon has joined #openstack-nova | 20:47 | |
*** baoli has quit IRC | 20:48 | |
ildikov | mriedem: hi. are you around? | 20:51 |
mriedem | in a meeting | 20:52 |
ildikov | mriedem: will you be available soonish today? | 20:52 |
mriedem | might be awhile | 20:52 |
mriedem | post your thing and i can try to catch up later | 20:52 |
ildikov | ok, thanks | 20:53 |
ildikov | basically I added multiattach_support to the capabilities dict for the virt drivers | 20:54 |
ildikov | and now I'm searching for the right place in bdm.attach to fail if the volume would be attached to a second instance and the virt_driver does not support multiattach | 20:55 |
*** annegentle has quit IRC | 20:55 | |
*** vishwanathj has joined #openstack-nova | 20:56 | |
*** annegentle has joined #openstack-nova | 20:56 | |
*** haomaiwang has joined #openstack-nova | 20:56 | |
ildikov | and I got confused regarding where we do check_attach and why the volume['attach_status'] == 'detached' check is here: https://github.com/openstack/nova/blob/master/nova/virt/block_device.py#L296 ? in the sense of if the volume is already attached that should be catched by check_attach at an earlier phase | 20:57 |
mriedem | ildikov: you'd have to look to see where do_check_attach=False is passed in there | 20:59 |
*** FL1SK has quit IRC | 21:00 | |
*** baoli has joined #openstack-nova | 21:00 | |
ildikov | but check_attach is called fro other functions too, so I don't really get the "rules" here | 21:00 |
*** haomaiwang has quit IRC | 21:01 | |
ildikov | mriedem: although the question on top of my head is that if the status is not detached then we just silently skip the Cinder part as everywhere else we raise an error | 21:01 |
ildikov | I mean in the case when the volume cannot be attached | 21:03 |
*** Sukhdev has quit IRC | 21:05 | |
*** raildo is now known as raildo-afk | 21:05 | |
openstackgerrit | Chris Dent proposed openstack/nova: Propagate qemu-img errors to compute manager https://review.openstack.org/264349 | 21:05 |
cdent | bauzas, n0ano: that's a fix for https://bugs.launchpad.net/nova/+bug/1436166 | 21:05 |
openstack | Launchpad bug 1436166 in OpenStack Compute (nova) "Problems with images bubble up as a simple "There are not enough hosts available"" [High,In progress] - Assigned to Chris Dent (cdent) | 21:05 |
bauzas | ack | 21:06 |
ildikov | mriedem: I assume if the virt driver does not support multi attach and the volume is already attached to another instance then I should raise an error on the top of bdm.attach, right? I wanted to do it in manager.py, but the volume is only available in bdm.attach | 21:07 |
* cdent waves goodnight | 21:08 | |
*** cdent has quit IRC | 21:08 | |
*** zenoway has joined #openstack-nova | 21:08 | |
*** baoli has quit IRC | 21:12 | |
*** baoli has joined #openstack-nova | 21:13 | |
*** mragupat has quit IRC | 21:16 | |
openstackgerrit | jichenjc proposed openstack/nova: Remove mox usage of versions, utils, wsgi https://review.openstack.org/264353 | 21:16 |
*** mragupat has joined #openstack-nova | 21:16 | |
*** smatzek_ has quit IRC | 21:16 | |
*** kragniz is now known as {^-^} | 21:17 | |
*** thangp has quit IRC | 21:18 | |
*** {^-^} is now known as Guest46353 | 21:18 | |
openstackgerrit | Tracy Jones proposed openstack/nova: config options: centralize section "guestfs" https://review.openstack.org/248963 | 21:18 |
*** Guest46353 is now known as kragniz | 21:18 | |
*** alejandrito has quit IRC | 21:20 | |
*** vipul has quit IRC | 21:21 | |
*** ssurana has quit IRC | 21:23 | |
*** lmtaylor has left #openstack-nova | 21:23 | |
*** ssurana has joined #openstack-nova | 21:23 | |
*** jichen has quit IRC | 21:23 | |
*** baoli has quit IRC | 21:24 | |
*** baoli has joined #openstack-nova | 21:25 | |
*** vladikr has quit IRC | 21:25 | |
*** hemnafk is now known as hemna | 21:26 | |
mriedem | ildikov: sorry i haven't read through the things on the driver capabilities part of the multiattach spec | 21:28 |
mriedem | i know john was bringing that up in one of the changes | 21:28 |
mriedem | in general it seems like what you're saying is correct | 21:28 |
ildikov | yeah, mainly the point is to not allow multiattach if the virt driver is not safe | 21:29 |
mriedem | check_attach assumes a single volume attach | 21:29 |
ildikov | I modified check_attach, so it knows about multiattach now | 21:29 |
*** belmoreira has quit IRC | 21:29 | |
ildikov | but it does not have the virt_driver info | 21:29 |
*** yarkot has quit IRC | 21:30 | |
ildikov | and also called from many places so I don't really know how it worth it to shoehorn this there | 21:30 |
*** rk4n has quit IRC | 21:31 | |
mriedem | like in compute api | 21:32 |
mriedem | where we don't have the virt driver | 21:32 |
ildikov | the other question is that would I need to add anything to the volume['attach_status'] == 'detached' check? I mean is it there to triple sure we don't try to attach the volume twice or is it there as the 'attaching' state should be checked twice? | 21:32 |
mriedem | so yeah i agree we probably don't want it in volume_api.check_attach | 21:32 |
mriedem | i'm not sure | 21:32 |
*** xyang2 has quit IRC | 21:34 | |
ildikov | ok, never mind, will ask around more | 21:34 |
ildikov | thanks for the help! | 21:35 |
*** baoli has quit IRC | 21:36 | |
*** artom has joined #openstack-nova | 21:36 | |
*** baoli has joined #openstack-nova | 21:37 | |
*** moshele has joined #openstack-nova | 21:38 | |
*** rk4n has joined #openstack-nova | 21:40 | |
*** ctrath has quit IRC | 21:42 | |
*** ctrath has joined #openstack-nova | 21:42 | |
*** whenry has quit IRC | 21:46 | |
*** dave-mccowan has quit IRC | 21:47 | |
*** rdopiera has quit IRC | 21:48 | |
*** baoli has quit IRC | 21:51 | |
*** rfolco has quit IRC | 21:52 | |
*** baoli has joined #openstack-nova | 21:53 | |
*** armax has quit IRC | 21:53 | |
*** matrohon has quit IRC | 21:53 | |
*** baoli has quit IRC | 21:55 | |
*** rk4n has quit IRC | 21:55 | |
openstackgerrit | Chuck Carmack proposed openstack/nova: config options: centralize 'spice' options https://review.openstack.org/264271 | 21:55 |
*** baoli has joined #openstack-nova | 21:55 | |
*** RichardRaseley has quit IRC | 21:55 | |
*** erlon has quit IRC | 21:56 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: Block more flaky py34 tests https://review.openstack.org/264368 | 21:58 |
*** _diana_1 has joined #openstack-nova | 21:59 | |
*** renatoarmani has quit IRC | 22:00 | |
*** baoli has quit IRC | 22:01 | |
*** baoli has joined #openstack-nova | 22:02 | |
*** mgoddard_ has joined #openstack-nova | 22:02 | |
*** tjones1 has joined #openstack-nova | 22:02 | |
*** gjayavelu has joined #openstack-nova | 22:02 | |
*** moshele has quit IRC | 22:03 | |
*** vishwana_ has joined #openstack-nova | 22:03 | |
*** ociuhandu_ has joined #openstack-nova | 22:04 | |
*** baoli has quit IRC | 22:04 | |
*** Kevin_Zheng_ has joined #openstack-nova | 22:04 | |
*** tjones has quit IRC | 22:04 | |
*** mrsoul` has joined #openstack-nova | 22:04 | |
*** DuncanT has quit IRC | 22:04 | |
*** Kevin_Zheng has quit IRC | 22:04 | |
*** bradjones has quit IRC | 22:04 | |
*** mrsoul has quit IRC | 22:04 | |
*** Kevin_Zheng_ is now known as Kevin_Zheng | 22:04 | |
*** ociuhandu has quit IRC | 22:04 | |
*** dougs has quit IRC | 22:04 | |
*** doffm has quit IRC | 22:04 | |
*** dimtruck has quit IRC | 22:04 | |
*** swamireddy has quit IRC | 22:04 | |
*** ociuhandu_ is now known as ociuhandu | 22:04 | |
*** aorourke_ has joined #openstack-nova | 22:04 | |
*** doffm has joined #openstack-nova | 22:04 | |
*** bradjones has joined #openstack-nova | 22:04 | |
*** bradjones has quit IRC | 22:04 | |
*** bradjones has joined #openstack-nova | 22:04 | |
*** baoli has joined #openstack-nova | 22:04 | |
*** omnipresent has joined #openstack-nova | 22:04 | |
*** vishwanathj has quit IRC | 22:05 | |
*** mgoddard has quit IRC | 22:05 | |
*** aorourke has quit IRC | 22:05 | |
*** sdague has quit IRC | 22:05 | |
*** junjie has quit IRC | 22:05 | |
*** dougwig has quit IRC | 22:05 | |
*** dougs has joined #openstack-nova | 22:05 | |
*** junjie has joined #openstack-nova | 22:07 | |
*** angdraug has joined #openstack-nova | 22:07 | |
*** zenoway has quit IRC | 22:07 | |
mriedem | py34 job is sad, which i think people know already | 22:07 |
mriedem | http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22b'%20%20%20%20raise%20exception.ImageNotFound(image_id%3Dimage_id)'%5C%22%20AND%20tags%3A%5C%22console%5C%22%20AND%20build_name%3A%5C%22gate-nova-python34%5C%22 | 22:07 |
*** salv-orl_ has joined #openstack-nova | 22:09 | |
*** baoli has quit IRC | 22:10 | |
*** sdague has joined #openstack-nova | 22:11 | |
*** dimtruck has joined #openstack-nova | 22:11 | |
*** salv-orlando has quit IRC | 22:12 | |
alaski | good to know. I was just trying to repro locally but I'm not having much luck running the py34 tests here | 22:13 |
mriedem | looks like those are all vmware api tests | 22:13 |
mriedem | http://logs.openstack.org/26/224726/16/gate/gate-nova-python34/cfa8b45/console.html | 22:13 |
*** mdrabe has quit IRC | 22:13 | |
*** swamireddy has joined #openstack-nova | 22:15 | |
dims | mriedem : i pushed out https://review.openstack.org/#/c/264368/ | 22:15 |
*** otter768 has joined #openstack-nova | 22:15 | |
dims | mriedem : for that problem after looking through all the uncategorized runs for the last days | 22:15 |
dims | s/days/24 hours/ | 22:16 |
alaski | oh, there's a helpful note from mriedem in tox.ini that describes my issue exactly | 22:16 |
mriedem | alaski: \o/ | 22:16 |
mriedem | dims:+2 at least for the vmware api test part of that | 22:16 |
*** zenoway has joined #openstack-nova | 22:16 | |
*** dougwig has joined #openstack-nova | 22:17 | |
*** DuncanT has joined #openstack-nova | 22:17 | |
mriedem | we need to fix the stubs.Set usage here https://github.com/openstack/nova/blob/master/nova/tests/unit/image/fake.py#L252 | 22:18 |
*** omnipresent has quit IRC | 22:19 | |
*** otter768 has quit IRC | 22:20 | |
*** fmccrthy has quit IRC | 22:22 | |
*** baoli has joined #openstack-nova | 22:22 | |
mriedem | but doing that with self.stub_out or even fixtures isn't going to happen w/o a reference to the test that is doing the stubbing, nice | 22:22 |
*** dims has quit IRC | 22:22 | |
*** fmccrthy has joined #openstack-nova | 22:24 | |
*** dedery has quit IRC | 22:24 | |
*** Sukhdev has joined #openstack-nova | 22:24 | |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: virtual interface tags DB table https://review.openstack.org/264014 | 22:25 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: metadata objects https://review.openstack.org/264015 | 22:25 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: user visible changes https://review.openstack.org/264018 | 22:25 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: compute manager nic tagging support https://review.openstack.org/264017 | 22:25 |
openstackgerrit | Artom Lifshitz proposed openstack/nova: vNIC tagging: libvirt driver implementation https://review.openstack.org/264016 | 22:25 |
*** tonytan4ever has quit IRC | 22:26 | |
*** mrkz has quit IRC | 22:26 | |
*** omnipresent has joined #openstack-nova | 22:28 | |
*** baoli has quit IRC | 22:28 | |
openstackgerrit | Andrew Laski proposed openstack/nova: Persist the request spec during an instance boot https://review.openstack.org/258628 | 22:29 |
*** mrkz has joined #openstack-nova | 22:30 | |
*** FL1SK has joined #openstack-nova | 22:30 | |
*** pratikmallya has quit IRC | 22:30 | |
*** dane-fichter has quit IRC | 22:31 | |
*** zenoway has quit IRC | 22:32 | |
*** _diana_1 has quit IRC | 22:33 | |
*** lihg has quit IRC | 22:34 | |
*** lihg_ has quit IRC | 22:34 | |
*** sacharya_ has joined #openstack-nova | 22:35 | |
*** cfriesen__ has quit IRC | 22:35 | |
*** _diana_1 has joined #openstack-nova | 22:36 | |
*** dedery has joined #openstack-nova | 22:38 | |
*** sacharya has quit IRC | 22:38 | |
*** sacharya_ has quit IRC | 22:39 | |
openstackgerrit | Matt Riedemann proposed openstack/nova: WIP: Convert nova.tests.unit.image.fake.stub_out_image_service to use stub_out https://review.openstack.org/264393 | 22:44 |
*** tonytan4ever has joined #openstack-nova | 22:45 | |
*** haomaiwang has joined #openstack-nova | 22:45 | |
*** mriedem has quit IRC | 22:46 | |
*** mrkz has quit IRC | 22:48 | |
*** gjayavelu has quit IRC | 22:49 | |
*** haomaiwang has quit IRC | 22:49 | |
*** mrkz has joined #openstack-nova | 22:50 | |
*** krtaylor has quit IRC | 22:51 | |
*** baoli has joined #openstack-nova | 22:52 | |
*** dims has joined #openstack-nova | 22:53 | |
*** thorst has quit IRC | 22:54 | |
*** pratikmallya has joined #openstack-nova | 22:54 | |
*** bklei has joined #openstack-nova | 22:54 | |
*** lpetrut has quit IRC | 22:55 | |
openstackgerrit | Ryan Rossiter proposed openstack/nova: Remove NovaObjectDictCompat from Aggregate https://review.openstack.org/264397 | 22:56 |
*** rlrossit has quit IRC | 22:56 | |
*** baoli has quit IRC | 22:56 | |
*** baoli has joined #openstack-nova | 22:57 | |
*** mragupat has quit IRC | 22:57 | |
*** davideagnello has joined #openstack-nova | 22:57 | |
*** pratikmallya has quit IRC | 22:59 | |
*** mrkz has quit IRC | 22:59 | |
*** mrkz has joined #openstack-nova | 23:00 | |
*** zenoway has joined #openstack-nova | 23:02 | |
*** ctrath has quit IRC | 23:04 | |
*** dims has quit IRC | 23:04 | |
*** omnipresent has quit IRC | 23:04 | |
*** tjones1 has quit IRC | 23:04 | |
*** tjones has joined #openstack-nova | 23:06 | |
*** tonytan4ever has quit IRC | 23:07 | |
*** bklei has quit IRC | 23:07 | |
*** baoli has quit IRC | 23:09 | |
*** dimtruck is now known as zz_dimtruck | 23:11 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/python-novaclient: Updated from global requirements https://review.openstack.org/264405 | 23:11 |
*** burt has quit IRC | 23:13 | |
*** penick_ has joined #openstack-nova | 23:17 | |
*** rfolco has joined #openstack-nova | 23:17 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/264441 | 23:18 |
*** penick__ has joined #openstack-nova | 23:18 | |
*** pradk has quit IRC | 23:18 | |
*** penick has quit IRC | 23:19 | |
*** penick__ is now known as penick | 23:19 | |
*** davideagnello has quit IRC | 23:20 | |
*** davideagnello has joined #openstack-nova | 23:21 | |
*** penick_ has quit IRC | 23:22 | |
*** pratikmallya has joined #openstack-nova | 23:22 | |
openstackgerrit | Diana Clarke proposed openstack/nova: Replace stubs.Set with stub_out https://review.openstack.org/264297 | 23:23 |
*** davideagnello has quit IRC | 23:25 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/264441 | 23:26 |
*** zenoway has quit IRC | 23:27 | |
*** dims_ has joined #openstack-nova | 23:29 | |
*** dave-mccowan has joined #openstack-nova | 23:31 | |
*** davideagnello has joined #openstack-nova | 23:34 | |
*** dims_ has quit IRC | 23:34 | |
*** annegentle has quit IRC | 23:36 | |
*** dims_ has joined #openstack-nova | 23:36 | |
*** baoli has joined #openstack-nova | 23:38 | |
*** zenoway has joined #openstack-nova | 23:39 | |
*** mrkz has quit IRC | 23:39 | |
*** thorst has joined #openstack-nova | 23:40 | |
*** rfolco has quit IRC | 23:41 | |
*** baoli has quit IRC | 23:43 | |
*** zenoway has quit IRC | 23:43 | |
openstackgerrit | melanie witt proposed openstack/nova: Add methods for RequestContext to switch db connection https://review.openstack.org/161906 | 23:44 |
*** mriedem has joined #openstack-nova | 23:44 | |
*** thorst has quit IRC | 23:45 | |
*** krtaylor has joined #openstack-nova | 23:45 | |
*** baoli has joined #openstack-nova | 23:49 | |
*** s-moriya has joined #openstack-nova | 23:54 | |
*** Sukhdev has quit IRC | 23:58 | |
jroll | would y'all mind doing a novaclient release somewhat soon to pick up https://review.openstack.org/#/c/262632/ ? | 23:59 |
jroll | mriedem: ^ | 23:59 |
*** oomichi has joined #openstack-nova | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!